I get asked some version of “is Google finished?” at nearly every business event I speak at. So here’s what the research actually shows, where it’s thin, and what I’d do about it.

Are people really using both Google and AI?

Yes. Similarweb data for May 2026 shows 95 per cent of ChatGPT users worldwide also use Google. People aren’t switching. They’re adding a tool and using whichever one suits the question in front of them.

That matches what we see in practice. A new client told us recently they’d found us through a mate’s recommendation, checked us on Google, then asked ChatGPT whether we were any good. Both platforms, used in sequence, to make one decision.

If people still use Google, what’s actually changing?

How much they use it, and for what.

A Bocconi University study tracked US desktop households from late 2024 to mid-2025 and compared those with access to ChatGPT Search against those without. Households with access ran 9.4 per cent fewer traditional searches, and the gap grew over time, reaching 17 per cent after 20 weeks. Google’s reach barely moved: about half of households used it either way. Same audience, fewer searches.

The drop wasn’t even across topics. It was steepest for informational questions. Referrals to academic sites fell by about a third, reference sites by a quarter, and developer and news sites by smaller but still real amounts. People took their “explain this to me” questions to AI first.

Not every study points the same way. Research by Gholami and colleagues, using data from over 4,000 desktop users, found people who started using a chatbot consistently visited more websites, while their traditional search activity rose at first and then declined. The authors describe their search estimates as noisy. Adoption seems to go through phases, which is a good reason to be wary of any single snapshot.

What happens to clicks when Google answers the question itself?

They fall, and by a fair bit. In a seven-day experiment published in August 2026, researchers recruited 1,100 US Chrome users and routed a randomly assigned group’s searches into Google’s AI Mode. Click-throughs to external websites dropped by 18.8 percentage points. People ran fewer search sessions each day, though each session ran a little longer. About one in seven participants said they found it harder to get to the specific website they wanted.

Google tells a different story from its side. It reports record query volumes, 2.5 billion monthly users of AI Overviews and a billion for AI Mode in its first year, and says people who use its AI features search more. That may well be true. But Google hasn’t released the baseline or user-level data that would let anyone check it, so it’s a claim rather than a finding.

Your customers still use Google. More of them are getting the answer without ever visiting your website.

What can’t the research tell us yet?

Quite a lot, and it’s worth being straight about that before anyone rebuilds a marketing plan around one statistic.

The studies measure different things. Audience overlap, query volume, click-through rates and referral traffic each capture a different slice of behaviour, so they can all be accurate and still seem to disagree. A stable audience doesn’t mean stable engagement, and more queries doesn’t mean more clicks.

The overlap figure is weaker evidence than it looks. That 95 per cent of ChatGPT users also use Google tells you the audiences overlap. It doesn’t prove AI is only a supplement. Habit keeps plenty of people on Google even as they move particular kinds of questions elsewhere.

Cause and effect is hard to pin down. Search behaviour shifts for lots of reasons, and separating AI’s effect from everything else is difficult even in well-designed studies.

Nobody is tracking individuals across platforms over time. We can’t yet see how one person moves between Google, ChatGPT and a website to finish a real task, which is the thing a business most wants to know.

Much of the data is American, and some of it is from the platforms themselves. Australian behaviour is likely to be similar, but it hasn’t been measured the same way here.

And the ground is moving. Google is testing ads inside AI Mode, which could change both how satisfied people are and how often they click. The next round of earnings reports and peer-reviewed studies should make the picture clearer.

What should a business do while the picture settles?

Plan for both, because your customers are using both. In practice that means four things.

1. Keep the Google fundamentals strong. Google’s audience hasn’t shrunk, and the AI systems lean on the same signals: clear pages, accurate structured data and consistent business details.
2. Make your pages worth quoting. If informational questions are moving to AI first, the businesses that get named are the ones whose content answers those questions plainly. That’s the core of the GEO side of the work.
3. Stop judging content on clicks alone. A page that gets you named in an AI answer can do its job without a visit. Watch enquiries, phone calls and branded searches as well as sessions. We covered what you can and can’t measure in AI search in more detail.
4. Check where you stand. Ask ChatGPT, Perplexity and Google’s AI Mode the questions your customers ask, and see who gets named. It takes about half an hour, and our step-by-step guide walks you through it.

None of that depends on predicting which platform wins. It works whichever way the numbers go.

Frequently asked questions

Are people leaving Google for ChatGPT?

Mostly not. Similarweb data shows 95 per cent of ChatGPT users also use Google. What’s changing is how people use each one: some informational questions are moving to AI tools first, and people who get an AI answer click through to websites less often.

Is AI reducing Google searches?

For some people and some topics, yes. A Bocconi University study found US households with access to ChatGPT Search ran 9.4 per cent fewer traditional searches, rising to 17 per cent after 20 weeks, with the biggest drops for academic, reference and news questions. Google’s overall audience stayed about the same size.

Does Google’s AI Mode reduce clicks to websites?

The evidence says it does. In one experiment with 1,100 US users, AI Mode cut click-throughs to external websites by 18.8 percentage points. Google says its AI features increase search use overall, but it hasn’t released data that would let independent researchers verify that.

What should my business do about AI search?

Plan for both Google and AI tools, because your customers use both. Keep your Google fundamentals strong, make your content clear enough to be quoted in AI answers, measure enquiries and calls as well as clicks, and check regularly whether AI tools name your business for the questions your customers ask.

Want to know where you stand in both?

Our AI readiness review checks where your business appears across Google, ChatGPT, Perplexity and Google’s AI Mode, compares you with the businesses being named, and sets out what it would take to close the gap.

Or get in touch with the team. For the wider picture, our AI search optimisation guide is a good place to start.

Figures in this article are drawn from Similarweb’s 2026 audience data, a Bocconi University study of ChatGPT Search adoption, research by Gholami et al. (2026), an AI Mode field experiment by Wang et al. (August 2026) and Google’s May 2026 statements, as reported by Search Engine Journal.

By Brendan Brooks, Founder and Managing Director of HyperWeb. Brendan has spent 25 years watching search evolve and helping Australian businesses stay visible through every shift.

The post Are People Leaving Google for AI? What the Research Actually Says appeared first on HyperWeb.

If you’ve read could AI replace the software you’re paying for and landed on the third answer, replacing it with something bespoke, this is the next step. Here’s what goes into a scope, and what to have ready before you start one.

Why does a custom application need a scope at all?

Because “we need a system that handles our jobs” means something different to everyone in the room. The owner pictures the reporting. The office manager pictures the booking screen. The developer pictures a database. All three are right, and none of them has described the same thing.

Off-the-shelf software gets around this by deciding for you. It does what it does, and you adapt. A custom application does what you tell it to, which means someone has to tell it everything. The scope is where that happens, on paper, while changes are cheap. After launch, the same change can mean rebuilding whatever was put on top of it.

A change on a whiteboard costs a conversation. The same change halfway through a build costs rework.

What goes into a proper scope?

The detail varies with the project, but a scope worth building from covers seven things.

1. The process as it runs today. Every step, who does it, what tool they use and where the information goes next. Including the spreadsheet nobody mentions and the step that only happens when a particular person is in.
2. What the application replaces. The subscriptions, spreadsheets and manual steps it takes over, and what it deliberately leaves alone.
3. Who uses it and what each of them can do. Staff, managers, customers, suppliers. Each group has its own screens and its own permissions, and each one adds to the build.
4. The systems it connects to. CRM, accounting, payments, email, anything else holding information the application needs. For each one: what moves, in which direction, and what happens when that system is down. This is usually the most involved part of any build, which is why we’ve written separately about what CRM integration actually involves.
5. The information coming across. Existing customers, jobs, products or records that need to move into the new application, and how clean they are. Migration is easy to underestimate.
6. What happens when it breaks. Whether it takes payments, holds personal or sensitive information, or has to meet an industry requirement. That sets the level of testing, security and support it needs.
7. What finished looks like. A written list of what the application must do for the first version to be accepted, and a separate list of what can wait for version two.

That last one does more to protect a budget than anything else on the list. When a project keeps growing, it’s usually because “done” was never written down.

What do scoping sessions uncover?

What comes out is usually the gap between how a process is described and how it actually runs. When you walk through a real job from enquiry to invoice with the people who do it, the exceptions come out: the customer who’s billed differently, the approval that happens by text message, the report someone rebuilds by hand every month because the numbers never quite match.

Those exceptions matter. Leave them out and the new application handles the tidy version of your business and forces the messy parts back onto spreadsheets, which is often the problem you were trying to solve.

Once you’ve walked the process, the exceptions become part of the scope rather than surprises during the build.

How do you know a scope is good enough to build from?

You should be able to hand it to a developer who wasn’t in the room and have them understand what’s being built. Every screen and every integration should trace back to a step in your process. There should be a clear line between version one and later. And you should recognise your own business in it, including the awkward parts.

A scope that fails those tests produces quotes you can’t compare, because every developer fills the gaps with their own assumptions.

Ours runs over two to four weeks. We need three people from your side: whoever can approve the scope and the budget, the people who actually do the work every day, and whoever manages the software you’re paying for now. The people doing the work matter most, because that’s where the exceptions live.

You come away with a written scope, a fixed-price quote to build it, and a phased plan that separates version one from what can wait. Scoping is paid, and the fee comes off the build if you go ahead. The scope is yours either way, including if you take it to another developer.

What should you have ready before scoping starts?

You’ll get more out of the sessions, and spend less time in them, if you bring:

• the software and subscriptions involved, with what each one costs
• the spreadsheets and documents people actually use, not the official versions
• the people who do the work every day, as well as the ones who sign off on it
• a rough sense of what the process costs you now in hours, errors or lost jobs
• any hard dates, like a contract renewal or a busy season you need to be ready for.

What can a scope not tell you?

It won’t make the decision to build for you. A scope can show that a bespoke application is possible and roughly what it involves. It can also show that the job is smaller than you thought and an automation between your existing systems would do, or that the software you have is closer to fitting than it seemed. That’s a good result too. We’d rather a client found that out in scoping than six months into a build.

Frequently asked questions

What does scoping a custom web application involve?

Working out, before any building starts, exactly what the application will do, who will use it, which systems it connects to, what information needs to move across and what counts as finished. A good scope maps the process as it runs today, including the exceptions, and separates what’s needed for the first version from what can wait.

Why do custom software projects go over budget?

Usually because the scope was thin. When the process, the integrations or the definition of finished aren’t written down, the gaps get discovered during the build, when changes cost far more than they would have on paper. A clear list of what the first version must do is the best protection a budget has.

How long does scoping take?

For most business applications, scoping takes two to four weeks. It depends on how many people and systems are involved, and how much of the current process lives in people’s heads rather than on paper.

What should I prepare before scoping a custom application?

A list of the software involved and what it costs, the spreadsheets and documents people actually use, access to the staff who do the work every day, a rough idea of what the current process costs you, and any hard deadlines. The more of the real process you bring, the more accurate the scope and the quote.

Thinking about a custom application?

If you’re not yet sure whether you need one, start with our AI readiness review. We look at the software you pay for and the processes it supports, and tell you where a bespoke application would pay for itself and where connecting what you already have would do the job. If a build makes sense, scoping is the next step.

Already know what you need? Get in touch with the team and we’ll talk through scoping it. You can also read more about our AI and application integration work.

By Brendan Brooks, Founder and Managing Director of HyperWeb. Brendan has spent 25 years building websites and business applications for Newcastle, Hunter and Australian businesses.

The post What Scoping a Custom Web Application Actually Involves appeared first on HyperWeb.

Most businesses reach this question through a renewal bill. The annual invoice arrives, it’s gone up again, and someone asks whether AI could do the job instead. Here’s how we work through it with clients, in enough detail that you can run it yourself.

We touched on the short version in the two questions business owners keep asking about AI, and one line from that post is worth repeating because it holds up every time: the businesses getting real value started with a problem, not with the technology. The ones who started with the technology mostly ended up with a subscription and a story.

Where should you start?

With the gap between what you pay for and what you use. We call it the subscription-gap test, and it takes about an hour with whoever actually uses the software every day.

1. List what you pay for. Every platform, every seat, every add-on, with the annual cost. Include the ones that are charged to a credit card rather than an invoice.
2. Mark what you actually use. For each one, which features does your team open in a normal week? Be strict. “We might use the reporting one day” goes in the unused column.
3. Count the workarounds. Every spreadsheet that exists because the software can’t do something, every export that gets reformatted by hand, every step where someone copies information from one screen into another.

The workarounds are the important column. They’re where the software stopped fitting your business and your staff filled the gap themselves.

Once you’ve found a gap, pick the one process that sits in the middle of it and run three questions over it.

How often does the process happen?

Frequency is what makes automation pay. A task that happens fifty times a day repays setup costs quickly. A task that happens four times a year almost never does, however tedious it is when it comes round.

Daily beats weekly, and weekly beats monthly. Anything rarer usually belongs on a checklist.

How much of it is judgment, and how much is moving information?

Break the process down step by step and sort each step into one of two piles.

The first pile is information moving: copying a customer’s details from an email into a CRM, pulling numbers from three systems into one report, checking an order against a stock list. This is AI territory now, and it’s usually the bulk of the hours.

The second pile is judgment: deciding whether a quote is worth chasing, reading a tricky customer complaint, weighing up which job gets the crew first. That stays with a person. What changes is the information in front of them when they make the call, because the first pile has already been done.

Work that moves information is where AI earns its keep. Work that weighs things up stays with a person, with better information in front of them.

What is the time actually worth?

Put a number on it. “We’d save heaps of time” can’t be weighed against a quote.

The arithmetic is simple. Hours the process takes each week, times the loaded cost of the person doing it (wage plus super, leave and overheads), times the weeks in a working year. As an illustration: six hours a week at a loaded cost of $55 an hour, over 48 working weeks, is $15,840 a year. That’s the ceiling on what fixing that process is worth, before you count errors, delays or the better work that person could be doing.

If automating it costs a fraction of that, it’s worth doing. If it costs more than two or three years of the time it saves, leave it for now. Either way you’ve made a decision with a number behind it, which a general ambition to “do more with AI” never gives you.

What are the three answers you can land on?

Run the questions and you’ll end up at one of three places.

Keep the software. It fits the process, your team uses most of it, and the price is fair. That’s a good result: you’ve renewed on purpose rather than by default.

Build AI around it. The software is fine at its core job, but the work between it and your other systems is eating hours. This is the most common answer we see. AI and automation handle the moving of information between the platforms you already have, and nothing gets replaced. It’s often the cheapest of the three and the quickest to show a result. If your CRM is one of those systems, what CRM integration actually involves covers the practical side, and our AI and application integration work is built around this middle path.

Replace it with something bespoke. The gap is wide, you’re paying tens of thousands a year, and your process has been bent out of shape to fit the tool. Then a purpose-built application can cost less over a few years than the subscription, and it fits the way you work. It’s a bigger decision with a longer payback, and it starts with a proper scope. We’ve set out what that involves in what scoping a custom web application involves.

What have we learned the hard way?

We’ve been adopting AI inside HyperWeb for about three years, and not all of it has stuck.

Early on we paid for a platform that was meant to take work off us. It didn’t. It still wanted the same repetitive data entry, and every record keyed in by hand was another chance to get something wrong. What it was asking for was exactly the work a database lookup should have been doing, with AI handling the entry and the errors that came with it. We’d bought the tool and left the actual problem sitting where it was.

The lesson has been the same each time. When we started with a process we already knew was costing us, the result lasted. When we started with a tool that looked promising, we usually paid for it for a few months and then cancelled it. The three questions exist because we needed them ourselves. The longer version of that sequence, purpose before the platform, is set out in our 4P framework for AI adoption.

Frequently asked questions

Can AI replace the software my business pays for?

Sometimes. The test is the gap rather than the technology: how much of the product you use, how many workarounds you run, and what the subscription costs against what a purpose-built alternative would. Start with one process, not a platform decision.

Which business processes should be automated with AI first?

High-frequency, low-judgment work: moving information between systems, assembling reports, triaging enquiries. Work that needs someone to weigh things up should stay with a person, with better information in front of them.

How do I work out if automating a process is worth it?

Put a number on it. Hours the process takes per week, times what that time costs, against the cost of automating. A named process with a named cost is a decision you can make. A vague ambition to “use AI” gives you nothing to decide.

Do I need to replace my software to get value from AI?

No. Often the win is AI working between the systems you already have. Replacement is the answer when the gap between what you pay for and what you use is wide and the workarounds are eating staff time.

Want us to run it for you?

This is the process we run in the second half of an AI readiness review. We look at your systems and your software costs with you, run the three questions over the processes that matter, and give you a short list of what’s worth doing first, what’s worth leaving alone, and roughly what each would cost.

Or if you’d rather talk it through first, get in touch with the team.

By Brendan Brooks, Founder and Managing Director of HyperWeb. Brendan has spent 25 years building websites and business applications, and the last three putting AI to work inside his own agency and his clients’ businesses.

The post Could AI Replace the Software You’re Paying For? Three Questions to Ask appeared first on HyperWeb.

What did our own testing show?

In our last spot-check across ChatGPT, Perplexity and Google’s AI Mode, using the kind of local service questions real customers ask. On Google’s AI Mode, the businesses being named tracked review counts closely. The ones that made the shortlist were consistently the ones with the deepest, most recent review profiles.

That makes sense, given AI Mode has Google’s own review data sitting right next to it. ChatGPT and Perplexity read more widely, which I’ll come back to.

I’ll be up front about the limitations. This is a spot-check, not a controlled study. AI answers vary between runs, and a deep review profile tends to travel with other things that help too: years in business, an active Google Business Profile, more mentions around the web. But when the same pattern holds across the categories we tested, it’s worth acting on.

Why do AI engines lean so heavily on reviews?

Think about the problem the engine is solving. Someone asks for a good electrician in Newcastle. There are dozens who could do the job, and the engine has to compress that down to a handful of names it’s prepared to put forward. It needs evidence, and reviews are the biggest body of independent, dated, plain-language evidence about a local business that exists anywhere.

Each part of a review profile tells the engine something different:

• Volume says plenty of real people have used you.
• Recency says you’re still trading and still doing good work, rather than coasting on a reputation from 2021.
• The review text supplies language the engine can match to a question.

That last one is the part most businesses haven’t clocked. As we explained in how AI assistants choose which local business to recommend, these engines break a question into a set of related searches and look for passages that answer them. A review that says “they rewired our 1950s house in Mayfield” contains a service, a property type and a suburb, written the way a customer talks. When the next person asks about rewiring an older home in Newcastle, that review is doing work your service page might not be.

A review that names the service and the suburb gives an AI engine something to match. Five stars and “great service” gives it almost nothing.

What does this change about asking for reviews?

For years the goal was the star rating. Get the average above 4.5, keep it there, move on. That still counts, because a poor average will keep you off a shortlist quickly. But the bigger lever now is what customers actually write.

A five-star review that says “great service, highly recommend” reassures a person. It gives an AI engine very little to work with. A review that says what you did and where gives the engine exact language to match against the next customer’s question.

The trick is asking for that without coaching people into copy that reads as fake. Scripted reviews all sound the same, to people and to machines. Here’s the approach we suggest to clients:

1. Ask soon after the job. While the work is fresh and the customer is still pleased, not weeks later on an invoice footer.
2. Make it one tap. Send the direct link to your Google review form, not a link to your profile that makes them hunt for the button.
3. Give a prompt, not a script. Something like: “If you’ve got a minute, we’d really appreciate a Google review. Mentioning what we did and roughly where you are helps other people nearby find us.” That’s enough. They’ll write it in their own words, which is the whole point.
4. Ask everyone. Don’t only ask the customers you’re sure were happy. Google’s policies don’t allow selectively asking for positive reviews, and a profile that’s nothing but glowing looks less credible.
5. Reply, briefly and personally. A reply shows the business is active. If it naturally mentions the job, it adds context, as long as it doesn’t read like it was written for a search engine.

Do reviews on other sites matter for AI search?

Yes, and this is where ChatGPT and Perplexity part ways with Google’s AI. Google reviews carry the most weight on Google’s own surfaces. ChatGPT and Perplexity draw on a wider spread of sources: industry review sites, directories, local forums and social platforms.

In practice that means three things. Know which review platforms matter in your industry, whether that’s a trades marketplace, a health booking platform or a hospitality site. Keep a genuine presence on two or three of them, not just Google. And make sure your business details are identical on every one. The same name, address, phone number and service description everywhere is how an engine works out that reviews on four different sites are all about the same business. It’s the same consistency point we made in why ChatGPT recommends your competitors and not you, and it applies to reviews as much as listings.

Are reviews enough on their own?

No. Reviews are one signal among several, and a great review profile attached to a website the engines can’t read will still lose. The engine needs to fetch your site, understand what you do and confirm the details match. If your services are buried in images, your contact details are inconsistent or your pages never answer a question plainly, reviews alone won’t carry you onto the shortlist.

And don’t buy or incentivise reviews. Offering a discount or a gift for a review is against Google’s guidelines, and fake or misleading reviews can breach the Australian Consumer Law, which the ACCC does enforce. Google already removes reviews it judges to be fake and can restrict profiles that attract them. The real risk is being filtered out entirely, and in a shortlist of three that’s the same as not existing.

In a shortlist of three, being filtered out is the same as not existing.

The good news is that the legitimate version is simple and cheap. Do good work, ask every customer, and make it easy for them to say what you did and where.

Frequently asked questions

Do Google reviews affect whether AI recommends my business?

Yes. AI answers name a handful of businesses rather than showing a list, and review volume, recency and content are part of how they choose. In our own testing, Google’s AI results for local services closely followed review depth. Reviews are one signal among several, so a strong profile works best alongside a clear, readable website and consistent business details.

How many reviews does my business need?

There’s no magic number, because it’s relative. AI engines compare you with the other businesses they could name for the same question. Check the review counts of whoever is currently being recommended for your key queries; that’s your benchmark. Recency matters too, so a steady flow of new reviews beats a big batch from a few years ago.

Do reviews on sites other than Google matter for AI search?

Yes. ChatGPT and Perplexity draw on multiple sources, including industry review sites and directories. Google reviews carry the most weight for Google’s AI, but a footprint across several platforms with consistent business details helps everywhere.

What should I ask customers to say in a review?

Nothing scripted. Ask them to mention what you did and where. A review that says the service and the suburb in plain words gives AI engines exact language to match against what future customers ask. Never offer incentives for reviews, and ask every customer rather than only the ones you know were happy.

Want to know where you stand?

If you’d like to know whether AI engines are naming your business, and why or why not, our AI readiness review covers exactly this: where you appear across ChatGPT, Perplexity and Google’s AI, how your reviews and listings compare with the businesses being named, and what it would take to close the gap. It’s one half of what we described in the two questions business owners keep asking about AI.

Prefer to start on your own? Our guide to checking where you really stand in AI search takes about 30 minutes. Or get in touch with the team and we’ll take a look together.

By Brendan Brooks, Founder and Managing Director of HyperWeb. Brendan has spent 25 years watching search evolve and helping Australian businesses stay visible through every shift.

The post Do Google Reviews Still Matter Now AI Answers the Question? appeared first on HyperWeb.

Here’s what sits in each bucket, and how you’d check your own position on either.

Why do businesses on page 1 of Google get no mention in ChatGPT?

I showed two rooms the same thing in the last month, a Newcastle Business Club lunch and a digital bootcamp out at Tocal. Take a business sitting on page 1 of Google. Ask ChatGPT the question that business’s customers would actually ask. Watch it name three other businesses and never mention them.

The room goes quiet at that point. Most people in it have never checked.

The reason is structural. Google ranks pages and hands you a list to choose from. AI engines assemble an answer, which means they have to decide who to name and who to leave out. They can only name businesses they can find, understand and trust, and those three things come from how your site is built, whether your business details are consistent everywhere they appear, and whether you’re mentioned credibly beyond your own website. Ranking well helps. It isn’t the same job. We pulled apart the five common reasons a business gets skipped in why ChatGPT recommends your competitors and not you.

The useful part for a business owner is that this is measurable. Ask the engines the questions your customers ask, logged out so you’re seeing what a stranger sees, and note who gets named. That’s your baseline, and it takes about half an hour. We wrote the method up in how to check where you really stand if you’d rather run it yourself.

Google ranks pages and hands you a list. AI engines assemble an answer, which means they have to decide who to name and who to leave out.

Where could AI do real work inside your business?

The second bucket emerges differently, organically. More and more clients are realising that AI capabilities can solve a pain point in their systems and processes. It usually starts with something specific: the admin that eats a day a week, the quoting process that lives across three spreadsheets and one person’s memory, or the software subscription that costs thousands a year and does maybe a third of what the business actually needs.

Those are good starting points, because they’re specific and they already have a cost attached. Vague AI ambitions go nowhere. A named process that wastes a named amount of time is something you can put a number against and either fix or leave alone on purpose.

Three questions are worth asking about any process you’re considering. How often does it happen? How much of it is judgment and how much is moving information from one place to another? And what would getting that time back actually be worth? Work that’s high frequency and low judgment is where AI earns its keep first. Work that needs someone to weigh things up stays with a person, ideally with better information in front of them. We’ve written up the full process in could AI replace the software you’re paying for.

Can a bespoke application replace expensive off-the-shelf software?

Often, yes. It’s also the question the fewest people think to ask.

Off-the-shelf software is built for the average of thousands of businesses. You pay for the whole product, use the slice that fits, and bend your own process around the parts that don’t. Per seat, per month, indefinitely.

A bespoke application works the other way around. It’s built around how you actually operate, it does the things you need and nothing else, and you own it.

It isn’t automatically the right answer. If the off-the-shelf product fits and the price is fair, keep it. The test is the gap between what you’re paying for and what you actually use, plus how much of your process you’ve had to distort to suit the tool. When that gap is wide and the subscription is well into five figures a year, it’s worth costing the alternative properly. AI has moved that maths, because building software specific to one business is meaningfully cheaper than it was three years ago.

What does an AI readiness review cover?

Two halves, matching the two buckets.

Where you appear to your customers:

• How your business is named across ChatGPT, Perplexity and Google’s AI, on the questions your customers actually ask
• Who’s being named instead, and what they’re doing that you aren’t
• The technical reasons you’re being skipped: content structure, schema markup, crawler access, and whether your business details are consistent across the web
• What to fix first, and what can wait

Where AI could do the work:

• A look at your systems and processes, and where the repeat work actually sits
• What your current software costs measured against how much of it you use
• Where a bespoke application would pay for itself, and where it plainly wouldn’t
• A short list of what’s worth starting now against what’s better left for twelve months

You get a written summary either way. Some businesses take that list and hand it to their own people, and that’s a perfectly good outcome.

What have we learned running this on ourselves?

We’ve spent the last three years doing this to our own business first, riding the heavy wave of AI change and trying to stay just under the lip. Plenty of what we tried didn’t stick.

Two things have remained consistent. The first is that the tools change every few months, so the useful skill isn’t picking the right tool, it’s being able to swap tools without rebuilding how your business runs. The second is that the businesses getting real value started with a problem, not with the technology. The ones who started with the technology mostly ended up with a subscription and a story.

We’ve started doing the same for clients, and we’re set up to help more.

Frequently asked questions

What is an AI readiness review?

It’s an assessment of two things: how visible your business is in AI search across ChatGPT, Perplexity and Google’s AI, and where AI could do useful work inside your systems and processes. The first half looks at why you are or aren’t being named when customers ask AI engines for a business like yours. The second half looks at repeat work, current software costs, and whether a bespoke application would serve you better than an off-the-shelf subscription. You get a written summary with what’s worth fixing first.

Isn’t AI search visibility just SEO with a new name?

They overlap heavily, and good SEO work makes AI visibility easier. The difference is the finish line. Google shows a list of links and lets the customer choose. An AI engine assembles a single answer and names a handful of businesses, so being one of the options is a different result from ranking well. Content structure, schema markup, consistent business information and crawler access all matter more when there’s only room for three names.

How do I know if a custom application is worth building instead of paying a subscription?

Compare what you’re paying for against what you actually use, and count how much of your own process you’ve reshaped to suit the software. If you use most of the product and it fits how you work, stay where you are. If you’re paying well into five figures a year for a fraction of the features, working around the gaps manually, and adding staff time to make it fit, it’s worth costing a bespoke build properly. Building software around one business is cheaper than it was three years ago, which has changed where that line sits.

Is my business too small for any of this?

No. The AI search side matters more for small businesses, not less, because an AI answer names three businesses where a Google results page shows ten. On the process side, the starting point isn’t a large project. It’s usually one repetitive task that eats a predictable amount of time each week. Small businesses often see the value faster because there are fewer people and systems to change.

Want to know where your business stands?

HyperWeb has been building websites and digital strategy for Newcastle and Hunter businesses since 2000, and both halves of this are now part of that work. If you’d like to know where you stand on either front, being found or getting the work done, book an AI readiness review below or get in touch with the team.

If you’d rather read first, our working guide to GEO and AEO covers the search side in full, and our AI and application integration page covers the build side.

The post AI Readiness: The Two Questions Business Owners Keep Asking appeared first on HyperWeb.

Whichever way it lands, the website is the smallest part of the decision. What decides whether a winery does well online is the club, the stock, and whether those two talk properly to the shop, the cellar door and the back office. Sort those out and the platform question mostly answers itself.

We’ve built winery sites on both platforms for years, most recently Mitolo Wines in McLaren Vale. Here’s what we’ve learned doing it.

What did the WooCommerce builds actually involve?

Our experience went from simple stores in the early 2000s to custom WooCommerce work. For Tyrrell’s and De Iuliis Wines we built a wine club platform with allocations and pre-orders, then wired it into NetSuite for warehousing and shipping, and into Mailchimp or Klaviyo for member communications.

That’s the real job. A wine club isn’t a subscription box. Members hold allocations, allocations have rules, releases open in waves, and a pre-order has to reserve stock that doesn’t exist yet without letting the warehouse ship it early. Every one of those rules had to be written, tested and maintained by us.

Woo did the job, and did it well. But every clever thing we built was a thing we owned forever.

Why did we start building on Commerce7?

Mercer Wines was the first. A small winery moving off WooCommerce onto Commerce7, while keeping their WordPress front end. Since then we’ve built on Commerce7 for Mitolo Wines, Glenarty Road, Amato Vino, Koonara Wines and Thistle Meadery, across the Hunter, McLaren Vale, Margaret River and Barossa wine regions.

The reason is simple enough. Commerce7 was built for wineries. Clubs, allocations, tastings and cellar door point of sale live in the one system, and the club logic that we used to write by hand is just there. Less custom code to maintain means fewer things to break at 6am on a release day.

What does a Commerce7 build on WordPress look like?

Mitolo is the clearest example. We rebuilt the site in WordPress with Commerce7 running the shop, the wine club and gift cards behind it. They asked us to keep their existing look and feel, so we did.

That matters, because most replatforming stories go the other way. A winery changes systems, the website gets dragged along, and customers land on something they don’t recognise a week before a release. Same site their customers already knew, different machinery underneath.

So which one, WooCommerce or Commerce7?

The feature tables won’t settle it. What settles it is how you run.

Commerce7 wins when the club is the business. If members, allocations and cellar door sales are the operation, Commerce7 is usually the better fit, because the club and the point of sale run off the same customer record. Woo has no equivalent. The club logic is native, and there’s less bespoke code sitting between you and your own data.

WooCommerce holds its own when the shop is the business. If you’re mostly selling wine online, with a light club or none at all, Woo gives you total control of the build and no per-platform subscription sitting on top. It’s also the sensible answer when you’ve already got working integrations you don’t want to rebuild. If your warehouse and accounting are wired into a WooCommerce store that’s been running fine for five years, moving platforms means rebuilding all of that to get back to where you already are.

What’s harder about each?

Both have a cost, and it’s worth being clear about them before you sign anything.

Commerce7 is a hosted platform, so the usual platform economics apply. There’s a subscription. And it has its own way of doing clubs, so your processes fit into it rather than the other way round. The front-end integration is also real development work, not a plugin you switch on. That’s the part people underestimate. It’s why we build a custom WordPress front end on ours. You keep full control of page speed, SEO and the brand experience, and Commerce7 does what it was built for behind it.

With WooCommerce, the flexibility costs you later. Everything clever is custom, custom code needs maintaining, and each plugin you add is another thing that can break on update. The day-to-day cost lands on your team too: the admin wasn’t designed for a cellar door crew doing customer service, and managing club orders through it takes patience or custom admin work. It’s one of the reasons club-heavy wineries end up moving.

What hasn’t changed?

More than ten years of these builds and the pattern holds on both platforms. The wineries that do well online are the ones with clean stock data, a club their team actually understands, and systems that talk to each other properly.

Get the club, the stock and the integrations right and the sales tend to take care of themselves. That was true on WooCommerce ten years ago and it’s true on Commerce7 now.

Frequently asked questions

Should my winery use WooCommerce or Commerce7?

If your business is club-first, with allocations, releases and a cellar door, Commerce7 is usually the better fit because that logic is built in rather than built by hand. If you’re shop-first, with heavy custom integration into systems you already run, WooCommerce can serve you better and gives you more control. Both need the same discipline about stock data and integration, and that discipline matters more than the platform choice.

Can we move to Commerce7 without changing how our website looks?

Yes. That’s the pattern we used for Mitolo Wines and Mercer Wines: Commerce7 runs the shop, the wine club and gift cards behind a custom WordPress front end. Customers see the same site they already know. The machinery underneath changes.

Does Commerce7 work with WordPress?

Yes, through a custom front-end integration. Commerce7 handles the commerce, the club and the customer records, and WordPress handles the site your customers see, with the two connected so that products, club sign-ups and checkout behave as one experience. It’s development work rather than a plugin install. We’ve written about what that kind of work involves in what CRM integration actually involves.

What systems does a winery website need to talk to?

Usually four. The club or CRM that holds members and allocations, stock and warehousing, the email platform for member communications, and the cellar door point of sale. On our builds that has meant NetSuite for warehousing and shipping, Mailchimp and Klaviyo for email, and Commerce7’s own club and POS. The question to ask about any of them is which system wins when two of them disagree.

Weighing it up?

If you run a winery and you’re deciding between WooCommerce and Commerce7, we’re happy to talk it through and give you an honest read on which fits how you actually operate. No pitch, just a conversation about your club, your stock and what’s already working. Get in touch with the team, or have a look at how we approach winery web design.

If you want more of the background first, our post on what CRM integration actually involves covers the wider integration picture. How much a website costs in 2026 explains why builds like these are priced the way they are. And there’s a broader platform view in our piece on moving to eCommerce in 2026. If you’re looking further ahead, agentic commerce covers what changes when AI agents start doing the buying.

Brendan Brooks is Founder and Managing Director of HyperWeb, a Newcastle web development and digital marketing agency building for Australian businesses since 2000.

The post WooCommerce or Commerce7? What We’ve Learned Building Winery Websites appeared first on HyperWeb.

Ask ChatGPT for a plumber or a physio yourself and you’ll see why that last step matters. You don’t get ten blue links. You get three names, maybe five, a line on each, and a reason to pick one. Being findable used to mean ranking. Now it also means being picked, by a system you can’t log into and can only partly measure.

Here’s what’s actually happening under the bonnet, what Google will and won’t tell you about it, and the work that still moves the needle. If you want the companion piece on why you might be missing from those answers entirely, we covered that in why ChatGPT recommends your competitors and not you.

How does AI search actually pick which businesses to mention?

Two techniques do most of the work: retrieval augmented generation and query fan-out. Neither is as complicated as it sounds.

Retrieval augmented generation (RAG) means the model doesn’t answer from memory alone. It fetches live pages from the search index first, reads them, then writes an answer grounded in what it found. This is why your website content matters as much as your general reputation. If there’s nothing fetchable that answers the question well, you can’t be part of the answer.

Query fan-out is the other half. Instead of running the one question you typed, the system breaks it into a set of related sub-queries and runs them all at once. Ask for “a good physio in Newcastle for a running injury” and behind the scenes it’s likely searching for physios in Newcastle, sports physiotherapy, running injury treatment, opening hours and reviews, plus a few angles you’d never have thought to type. It then assembles one answer from the best of all of that.
So you’re not competing for the one query any more, you’re competing across a dozen questions you never saw. There’s a number worth sitting with here. When Whitespark tested 540 local queries across six industries this year, Google’s AI answers showed up on only 15 per cent of the direct “physio near me” style searches. On hybrid questions, the ones that carry a buying decision inside them, “should I see a physio for this knee or wait it out”, they showed up 97 per cent of the time. Your customers are meeting the AI answer while they’re still deciding whether to buy, not just who from.

The practical upshot is that narrow keyword optimisation matters less than it used to. Covering your topic properly, in the language a real customer would use, matters a lot more. That’s the core of the GEO side of the work, and it’s a slower build than chasing a single phrase.

Why do listings and reviews still do the heavy lifting?

Because the retrieval step has to find something it trusts.

An AI assistant naming a business is making a small reputational bet on your behalf. It favours businesses it can describe confidently: the ones where the name, address, phone number, hours, categories and service description line up everywhere it looks. Where those details are inconsistent, the system plays safe and names someone clearer. We go into that in more depth in entity SEO.

Reviews are the other half of it, and not only for the star rating. Review text is retrievable content written in customer language, describing what you actually do and how well you do it. Volume, recency and specificity all count. A steady trickle of recent, detailed reviews is worth more here than a pile of three-word ratings from 2021.

It’s unglamorous work. It’s also the highest-return thing most local businesses can do, and it costs nothing but attention.

Can you actually measure any of this?

Partly. This is the uncomfortable bit, and anyone selling you a clean dashboard is overselling.

Google’s reporting on AI features will show you impressions and clicks for links surfaced there. That’s useful, and it’s more than we had a year ago. But it’s aggregate. You don’t see which question triggered you, you don’t see who else got named alongside you, and you certainly don’t see the answers where you weren’t mentioned at all. It also only covers Google. Nothing in it tells you what ChatGPT, Perplexity or Copilot said about your industry this morning.

So most businesses fall back on two workarounds.

Fixed query monitoring. Pick twenty or thirty questions your customers actually ask, run them across the major assistants on a set schedule, and record whether you get named and who else does. It’s crude, but it’s honest. The catch is that answers vary between sessions, accounts and locations, so you’re sampling rather than measuring. Run them logged out, and run them consistently, or the comparison is worthless.

Referral traffic analysis. Filter your analytics for referrals from the AI platforms. Real signal, but it undercounts badly, because plenty of people read the answer, form a view, and arrive later through a branded search or a direct phone call. In our experience being named in an AI answer shows up as a phone call more often than a session.
Neither method is a reporting suite. Treat both as directional, watch the trend rather than the number, and don’t make big decisions off a single week.

What does Google actually recommend you do?

Google’s own guidance for being eligible in AI features is refreshingly unromantic. Three things:

• Put your key business information in text on the page, not locked inside an image, a PDF or a video with no transcript.
• Keep your structured data accurate and matched to what’s actually visible on the page.
• Keep crawling permissions open, so the systems that want to read you can.

Two things Google says it does not do, since both are being sold to local businesses right now: it ignores llms.txt and similar files, and it does not need your content chopped into special AI-sized chunks. Plain, well-structured pages are the whole brief. One caveat from the same guidance that lands squarely on the build side of our work: sites that render their content through JavaScript are, in Google’s words, generally more complex to work with. If your key information only exists after a script runs, you’re making the reader work harder than it needs to. We’ve inherited sites like that, and the fix is usually a rebuild decision, not a content one.

Nothing exotic. It’s the same fundamentals that have made sites findable for two decades, pointed at a new kind of reader. We wrote up the local application of that in local SEO and structured data.

Why has accessibility become an AI readiness issue?

This is the bit most businesses haven’t clocked yet, and it’s the one I’d put money on mattering more over the next couple of years.

AI browser agents that carry out tasks, booking an appointment or filling in an enquiry form, often read a page through its accessibility tree. That’s the structured representation browsers build for screen readers. It’s the same map, used by a different reader.

Which means the accessibility work you might have parked as a compliance chore now determines whether an agent can complete a booking on your site. Unlabelled buttons, form fields without labels, fake buttons built out of divs, custom widgets with no ARIA: a screen reader user struggles with all of it, and so does the agent. If the agent can’t work out which control submits the form, it moves on to a competitor whose site it can read.

Building an accessible site and building an agent-ready site turn out to be the same job. That’s a rare thing in this business, one piece of work that pays off twice.

Where should you start?

In this order, because the early items are cheap and the later ones compound:

1. Get your business information identical and in plain text everywhere: website, Google Business Profile, directories.
2. Check your structured data is accurate and reflects what’s on the page.
3. Confirm nothing in your robots file, firewall or security plugin is blocking the crawlers you want.
4. Fix accessibility on your conversion paths first: enquiry forms, booking flows, contact details.
5. Keep reviews recent and encourage detail, not just stars.
6. Set a baseline with twenty or thirty fixed queries, then re-check quarterly.

None of that is a trick, and none of it is beyond a small business. The businesses doing this work now, while most are still deciding whether AI search is real, are the ones that’ll be named for the next few years.

We see the gap most weeks. When we ran the check for one long-standing Newcastle manufacturer, holding page 1 Google positions for their core product terms for years, Perplexity answered their customers’ buying question by naming eight of their competitors. They weren’t on the list. Their listings were inconsistent, their site described products rather than answering questions, and nobody had ever asked the engines about them. Every one of those is on the list above.

Figures in this article are drawn from Google’s published guidance on AI features, Whitespark’s 2026 local query analysis and Uberall’s quick-service benchmark, as reported by Search Engine Journal.

Frequently asked questions

How do AI assistants decide which local businesses to recommend?

They fetch live pages from a search index and write an answer grounded in what they find, a technique called retrieval augmented generation. Most also use query fan-out, breaking your single question into several related searches and assembling one answer from all of them. Businesses that get named are the ones with consistent listing information, content that answers real questions in plain text, accurate structured data and open crawling permissions. AI answers appear far more often on questions that contain a buying decision than on simple “near me” searches, so answering those questions on your site matters most.

Can I track whether my business appears in AI search results?

Only partly. Google’s reporting on AI features shows aggregate impressions and clicks for your links, but not which query triggered them, who else was named, or what happened outside Google. The practical alternatives are running a fixed set of customer questions across the major assistants on a schedule and recording the results, and filtering your analytics for AI platform referrals. Both are directional rather than exact.

Does website accessibility affect AI visibility?

Increasingly, yes. AI agents that complete tasks on websites, such as making a booking or submitting an enquiry, often read pages through the accessibility tree that browsers build for screen readers. Unlabelled buttons, missing form labels and custom controls without ARIA make a site hard for both screen reader users and AI agents. Fixing accessibility on your conversion paths improves both at once.

Is optimising for AI recommendations different from local SEO?

They overlap heavily. Consistent business information, genuine reviews, accurate structured data and crawlable content serve both. The differences are in emphasis: AI answers reward content that covers a topic thoroughly in natural language rather than content built around a single keyword, and they put more weight on being describable as a clear, consistent business entity.

Want to know where you actually stand?

HyperWeb has been building websites and digital strategies for Newcastle and Hunter businesses since 2000, and getting our clients named in AI answers is now part of that work. If you’d like us to check where your business currently sits in AI search and what it would take to shift it, get in touch with the team. For the wider picture, our guide to getting found in AI search is a good place to start.

The post How AI Assistants Decide Which Local Business to Recommend appeared first on HyperWeb.

More of your customers are starting with AI than you think. Instead of Googling “web designer near me” and scanning a page of results, they ask ChatGPT, Perplexity or Google’s AI Mode a full question and take the shortlist it gives them. If your business isn’t on that shortlist, you’re losing the enquiry before you ever knew it existed. We’ve watched it happen to businesses that have ranked on Google’s first page for years: they ask an AI engine the question their customers ask, and it recommends a handful of competitors without ever mentioning them. This is the same shift we wrote about in our working guide to GEO and AEO, looked at from the angle that matters most to a business owner: why am I not there, and what do I do about it?

How does AI search decide who to mention?

It helps to understand what’s happening under the hood, because it explains every fix that follows.

An AI engine answers a question in two steps. First it draws on what it absorbed during training, a broad sense of which businesses are known and credible in a field. Then, for anything current, it fetches live pages from the web, reads them, and lifts the parts that answer the question. To be named, you need to either be a brand the model already recognises, or have a page it can fetch and understand well enough to quote. Ideally both.

That’s the whole game. Everything below is about making your business one of the brands the model knows, and your pages the ones it can lift a clean answer from.

What are the real reasons your business is invisible?

In our experience there are five, and most businesses have more than one.

1. Your content doesn’t answer questions directly. AI engines lift short, self-contained answers. A page that buries its point under five paragraphs of preamble, or never states the answer plainly, gives the engine nothing to quote. Pages that lead with a clear answer and back it up get cited. If your content reads like a brochure rather than an answer, it’ll be skipped.

2. Your business isn’t a clear entity. AI engines build a picture of who you are from your name, address, phone number and description as they appear across your website, Google Business Profile, LinkedIn, directories and data aggregators. When those details are inconsistent, or barely present, the engine can’t form a confident picture of your business, so it plays safe and names someone clearer. We go deeper on this in entity SEO, but the short version is: be the same business everywhere, in exact detail.

3. You have little or no schema markup. Schema is structured data that spells out, in a format machines read directly, what your page is and which sentence answers which question. Most businesses treat it as an optional plugin setting. AI engines and search engines use it to confirm what they’re looking at. Without it, you’re asking the machine to guess. Our piece on schema markup covers what actually matters here.

4. You’re accidentally blocking the AI crawlers. AI engines send their own bots to read your site: GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot and others. If your site blocks them, whether by choice or by a setting nobody remembers changing, you remove yourself from AI answers entirely. It’s worth checking, and we explain the decision in GPTBot and your website.

5. You haven’t earned enough authority. AI engines lean toward businesses that appear credibly and repeatedly across the web: mentions, links, reviews, coverage. If your business barely shows up beyond its own website, the model has little reason to trust you over a competitor who’s cited everywhere. This is the slowest to fix, but it compounds.

How do you check where you actually stand?

Before fixing anything, find out how invisible you really are. Ask ChatGPT, Perplexity and Google’s AI Mode the questions your customers would ask: “best [your service] in [your town]”, “who should I hire to [do the thing you do]”. See whether you’re named, and who’s named instead. We wrote a full method for this in how to check where you really stand. Do it logged out, so you’re seeing what a stranger sees, not a result skewed by your own history.

We run this check on our own business, and it keeps us honest. One recent morning we asked three engines the same simple question, who are the best web developers in Newcastle. ChatGPT put us first. Perplexity and Google’s AI Mode didn’t mention us at all, even though we’ve ranked on Google’s first page for our own services for years. Same question, same day, three completely different answers. That’s the nature of these engines: being named by one is no guarantee with the next, and the answers shift over time, which is why one check is a snapshot, not a scoreboard.

Whatever comes back is your baseline. If competitors are named and you’re not, the five reasons above are your checklist.

How do you fix it?

The work sorts into quick wins and slow burns. Do the quick ones first.

Fix your business information everywhere. Make your name, address and phone number identical across your website, Google Business Profile and every directory you can find. It’s unglamorous, and it’s the single highest-return thing most local businesses can do, because it feeds the entity picture directly.

Rewrite your key pages to answer questions. On the pages that matter, lead with a direct answer, then support it. Add an FAQ section that answers the real questions customers ask. This helps human readers and gives AI engines clean passages to lift.

Add proper schema markup. Organization, LocalBusiness, Service and FAQPage schema on the relevant pages. It’s technical work, but it’s a one-time setup that keeps paying off.

Confirm the AI bots can reach you. Check your robots file and any security or firewall settings aren’t quietly blocking the AI crawlers you want reading your site.

Build authority over time. Publish genuinely useful content on your topic, earn mentions and reviews, and get listed where your industry gets listed. This is the long game, and it’s the difference between being occasionally fetched and being a brand the model knows.

Why is this a technical job, not a magic trick?

Plenty of people are already selling “get your business into ChatGPT” as if it were a switch. It’s not. Appearing in AI search is the product of clean structure, consistent data, real schema, open access for the right crawlers, and earned authority. That’s disciplined web and content work, which is exactly what a good web team does. If a service promises instant AI visibility with no changes to your site, it’s selling you something that doesn’t exist.

None of this is beyond a small business. It’s the same fundamentals that have always made a site findable, pointed at a new reader. The businesses that do it now, while most are still working it out, are the ones AI engines will name for the next few years.

Frequently asked questions

Why does ChatGPT recommend my competitors but not me?

Usually because your competitors are easier for the AI to find, understand and trust. Their business information is consistent across the web, their pages answer questions in a liftable way, they have schema markup, they allow the AI crawlers, and they’re mentioned in more places. Any one of those gaps can be enough for the engine to name them and skip you. The fixes are the same disciplined web and content work that makes any site findable.

How do I get my business to show up in AI search?

Start by making your business information identical everywhere (website, Google Business Profile, directories), rewrite your key pages to answer real customer questions directly, add proper schema markup, confirm you’re not blocking the AI crawlers, and build authority over time through useful content and mentions. Check your starting point first by asking AI engines the questions your customers would ask, logged out.

Is appearing in AI search different from ranking on Google?

They overlap heavily. AI engines draw on much of the same web Google indexes and reward the same signals: clear content, consistent business data, strong schema and real authority. The difference is the finish line. Google shows a list of links; AI search lifts an answer and names a source. Work that improves one generally improves the other.

Can I pay to be included in ChatGPT or AI answers?

No. There’s no paid placement in the organic answers ChatGPT, Perplexity or Google’s AI Mode generate. Being named is earned through the visibility, structure and authority of your business online. Anyone promising guaranteed paid inclusion in AI answers is misrepresenting how these engines work.

Want a straight answer on where you stand?

HyperWeb has built websites and digital strategies for Newcastle and Hunter businesses since 2000, and getting our clients found in AI search is now part of that work. If you’d like us to check where your business currently stands in AI search and what it’d take to fix it, get in touch with the team. For the full picture, our guide to GEO and AEO is the place to start.

The post Why ChatGPT Recommends Your Competitors and Not You appeared first on HyperWeb.

• A small brochure site that describes your business. Custom-built on WordPress at a lean scope, your content presented cleanly, with solid foundations. $3,000 to $6,000.
• A site designed around how you sell. Custom design shaped by your customers and your sales process, custom build, eCommerce online shopping, SEO foundations done properly. $8,000 to $25,000.
• An operational platform, not just a website. Custom web application development that runs a business process end to end: client portals, booking or training systems, with CRM, accounting and third-party application integration. From $20,000.

The most common conversation I have with new clients starts with this question, usually after they’ve collected quotes an order of magnitude apart and concluded someone must be having a lend. Almost always, nobody is; they’ve been quoted three different products. This post is about telling them apart.

Why do website quotes vary so much?

Send the same enquiry to five developers and the quotes can differ tenfold. Three things explain nearly all of it.

First, scope and approach. One builder is pricing your content dropped into an off-the-shelf design; another is pricing design work that starts with your customers and works backwards. We sit firmly in the second camp, and we do not build from templates, but both are legitimate ways to sell a website. They are just different jobs.

Second, who’s doing the work and where. Offshore teams and freelancers carry different cost structures to a local studio that answers the phone when something breaks. That’s a trade-off to make knowingly.

Third, and this is the big one: whether anyone asked what the website needs to do. A quote that assumes “describe the business, collect enquiries” will always be a fraction of one priced for payments, logged-in users or system integration. A cheap quote usually isn’t dishonest. It’s quoting a smaller job than the one you actually have. The danger isn’t the price, it’s the mismatch.

What actually moves the price?

Four levers do most of the work.

Content. Who’s writing the words and sourcing the images? “Client supplies all content” is the quiet assumption behind many low quotes, and it’s where projects stall for months. A quote that includes copywriting and content migration costs more and finishes sooner.

Design. An off-the-shelf design gives you a layout that thousands of other businesses also use. Custom design starts with questions: who buys from you, what do they need to see to trust you, what does the path to enquiry look like? For a cafe, a small custom brochure site is often plenty. For a firm selling high-value engagements, design built around how buyers decide tends to pay for itself.

Functionality. This is the steepest curve. A contact form is trivial. Online payments raise the stakes on security. Logged-in user accounts change the architecture entirely. And connecting the site to your CRM or accounting system is a business-systems job, not a plugin; we’ve written about what CRM integration actually involves.

The invisible layer. SEO foundations, page speed, security hardening, accessibility. Two sites can look identical while one is built to be found and to hold up, and the other quietly leaks rankings and breaks under load. You can’t see this layer in a screenshot, which is why it separates quotes that look inexplicably far apart.

What’s the tier most people don’t know exists?

Most buyers compare the first two tiers and stop. The third changes the maths completely: an operational platform built through custom web application development, a website that runs a program instead of describing one.

We built one recently for the Hunter Melanoma Foundation. Their SWAP sun-safety training platform delivers the courses, takes the payments, issues personalised certificates automatically, gives employers live reporting, and feeds enrolments straight into the Foundation’s CRM and accounting systems. A small charity now operates with the reach of a much larger training organisation, and each new enrolment adds zero admin. The full story is in our post on building a website that runs a program. There’s a platform-level version of the same question in WooCommerce or Commerce7 for winery websites.

The point isn’t “spend more”. It’s that the return calculation changes completely. A brochure site earns its keep by winning enquiries. A platform earns its keep by replacing wages-worth of manual administration, every week, forever. Once a website removes a process instead of describing one, you stop comparing its cost to other websites and start comparing it to the cost of doing that work by hand.

What does a cheap website really cost?

Here’s the pattern I’ve watched play out again and again: a business buys the cheapest quote, spends eighteen months fighting it (content that’s hard to update, pages that don’t rank, a form that silently stops working), then pays again for the site they should’ve bought first. The cheap site wasn’t a bargain, it was a deposit on the real one.

But the other side is real too. If you’re early stage, still validating the idea, or the budget simply isn’t there yet, a DIY or off-the-shelf site is the right call. A modest site you can afford beats an ideal site you can’t. We said the same when we asked whether AI can build your business website: sometimes yes, if you know which situation you’re in. Just buy it as what it is: a starting point, not a destination.

What does a website cost to run?

Every website has ongoing costs, and any quote that doesn’t mention them is incomplete. Expect to budget around $140 per month for a business-grade setup.

What that should cover: quality Australian hosting, software and plugin updates, security monitoring and patching, backups you can actually restore from, and access to a human when something needs fixing. Ask what’s included, and what happens when something breaks outside the plan. This is simply the operating cost of a business asset, like insurance or rent, and it’s modest when set up well.

How do I compare website quotes?

Ask every developer the same questions (including us) and the vague quotes reveal themselves quickly:

• What exactly is included in content: who writes the copy, who supplies images, who migrates the old site?
• Is the design custom or off-the-shelf, and can you show me which?
• Who owns the site, the domain and the hosting account when we’re done?
• What happens after launch: what does support cost and what does it cover?
• What’s the plan for being found, on Google and in AI search? Push for specifics.
• What are you assuming the website needs to do? If nobody asked about your business process, they’re quoting a brochure.
• Can I speak to a client from a project like mine?

Any decent developer will welcome these.

Frequently asked questions

How much does a small business website cost in Australia?

In 2026, a professionally built small brochure website typically costs $3,000 to $6,000. A custom-designed website or online store built around your business and customers usually runs $8,000 to $25,000. An operational platform, built through custom web application development with portals, booking or training systems and CRM or accounting integration, generally starts from $20,000. The right tier depends on what the website needs to do, not on how much a website “should” cost.

Why do website quotes vary so much?

Because different builders are quoting different products. One is pricing your content dropped into an off-the-shelf design, another is pricing custom design and SEO foundations, and a third is pricing functionality like payments or CRM integration. Offshore and local teams also carry different cost structures. A cheap quote usually isn’t dishonest; it’s usually quoting a smaller job than the one you actually have. Compare what’s included, not just the number.

Is a cheap website worth it?

Sometimes, yes. If you’re early stage, validating an idea, or the budget isn’t there yet, a DIY or off-the-shelf site is a sensible starting point. The false economy is buying a cheap site to do an expensive site’s job: it costs you in rankings, updates and enquiries for eighteen months, and then you pay again for the site you needed first. Match the build to the job.

How much does it cost to maintain a website?

Budget around $140 per month for a business-grade setup. That should cover quality hosting, software and plugin updates, security monitoring and patching, reliable backups, and access to support when something needs attention. Treat it as the operating cost of a business asset: an unmaintained website drifts out of date, slows down and becomes a security risk, and recovering from that costs far more than maintaining it would have.

Want a straight answer on yours?

Every project sits somewhere specific on these ranges, and finding that spot takes a conversation, not a calculator. If you’d like a straight answer on your project (a number and the reasons behind it, no obligation), get in touch for a free initial consultation, or see how we approach a build on our website development services page.

By Brendan Brooks, Founder and Managing Director of HyperWeb. Brendan has spent 25+ years building for the web, and answering this question over coffee.

The post How Much Does a Website Cost in 2026? The Honest Answer appeared first on HyperWeb.

That last part is the point of this post.

Most businesses think of their website as a description of what they do. Services, team, testimonials, contact form. That model is so common we barely notice it, and for plenty of organisations it’s enough. But it quietly sets a ceiling: the website tells people about your work, and then humans do all of the actual work behind the scenes.

The platform we recently built for the Hunter Melanoma Foundation breaks that model, and it’s a useful case study in what becomes possible when you stop asking “does our website describe us well?” and start asking “what could our website actually do for us?”

A mission bigger than the headcount

The Hunter Melanoma Foundation is a registered charity working to reduce the impact of melanoma in a country with some of the highest skin cancer rates in the world. Its SWAP program, the Sun Safe Workplace Awareness Program, targets the people who carry a disproportionate share of that risk: outdoor workers. Construction crews, landscapers, road workers, trades. People whose workplace is the sun.

SWAP had been delivered in person, and it worked. But in-person delivery has a hard limit: the size of the team delivering it. The Foundation is a small charity, and the number of workplaces that need this training is enormous. Every hour spent on scheduling, paperwork and certificates was an hour not spent reaching another workplace.

The mission was bigger than the headcount. That gap is exactly what a well-built web application is for.

Not videos on a webpage

It would have been easy to treat this as a content problem: record the training, put it behind a login, done. That version gets built a lot, and it fails a lot, because delivering a real training program requires much more than hosting videos.

The Foundation needed structured courses with lessons, assessments, learner accounts and progress tracking. It needed provable completion: personalised certificates for workers and reliable records for employers, because this is workplace safety training, not casual viewing. It needed payment flexibility for everyone from a sole trader booking one course to a large employer enrolling a whole workforce. And it needed all of that to run without manual administration, because a small charity team can’t spend its week re-keying enrolment data.

One more constraint shaped everything: the audience. Outdoor workers don’t complete training at a desk. The platform had to work on a phone, in a work vehicle or a site shed, on patchy regional internet.

What we built

The SWAP training platform now lives at training.hmf.org.au, a dedicated learning management system built on the LearnDash and BuddyBoss stack within WordPress. It carries the Foundation’s brand while standing visually distinct as its training arm. In practical terms, here’s what it does:

• Delivers the courses. Structured lessons and assessments, learner dashboards, and password-protected resources like toolbox talks and workplace checklists. New courses can be added as the program grows.
• Takes the money. Stripe payments supporting one-off purchases, subscriptions, discount coupons, free courses and custom packages for larger organisations, all managed by the Foundation’s team without a developer in sight.
• Proves the training happened. Completing the final assessment generates a PDF certificate personalised with the learner’s name, downloadable on the spot and emailed automatically. Employers can run reports by course, learner or group and see exactly where their team is up to.
• Talks to the Foundation’s systems. Integrations with Zoho CRM and Xero mean enrolments, payments and contact records flow through automatically, with email notifications keeping administrators and learners informed.

Underneath all of that sits the unglamorous work that makes a platform trustworthy: security hardening around user accounts and payments, speed optimisation, responsive development, and on-page SEO from day one. We finished with hands-on training for the Foundation’s team at their Waratah premises, because a platform you can’t run yourself is a liability, not an asset.

The number that matters is zero

Here’s the outcome I keep coming back to. When a new business enrols its team in SWAP today, the administrative cost to the Foundation is zero. No forms to process, no certificates to create, no records to update, no invoices to chase into the accounting system. The platform does all of it.

That’s what lets a small charity operate with the reach of a much larger training organisation. Scale didn’t come from hiring; it came from removing the administration between the Foundation and every workplace it serves.

“Working with Brendan and the Hyperweb team has been an excellent experience. They’ve consistently demonstrated creativity in responding to our requests and have been timely and reliable in all communication. The result is a high-quality product in the SWAP training platform, which reflects their skill, professionalism, and commitment to delivering great work.”

Sue-Ellen Evans, Project Manager, Hunter Melanoma Foundation

You can read the full project breakdown in our SWAP training platform case study.

What this means for your organisation

You may not run a sun-safety charity, but the underlying question travels well: what work is your team doing by hand that your website could be doing for you?

If you deliver training, the answer might look like SWAP. But the same pattern applies to membership content, certification and compliance, client onboarding, bookings and recurring billing, gated resources, or reporting that customers currently have to email you for. In each case the shift is the same: from a website that describes the service to a web application that performs it.

A few tests we would apply to your situation, based on this build:

• Is the process repeatable? If you run the same sequence for every customer (enrol, deliver, certify, record), that sequence can almost certainly be systematised.
• Is admin the bottleneck? If growth means proportionally more paperwork, automation changes your economics, not just your convenience.
• Does proof matter? Where completion, compliance or membership status must be provable, automating the record-keeping is worth more than automating the delivery.
• Can your team run it without us? A purpose-built platform should end with handover, not dependence. If a proposal keeps the developer permanently in the loop for daily operations, question it.

To be clear, this isn’t an argument that every business needs a web application, and it isn’t an AI story either; SWAP is a learning platform, built on proven tools, doing structured work reliably. Some organisations genuinely need a well-crafted brochure site that tells the right story, and nothing more. The mistake is never asking the question, or assuming the answer requires enterprise budgets. It doesn’t.

Frequently asked questions

What is the difference between a website and a web application?

A website presents information about your organisation. A web application performs work: it handles accounts, payments, content delivery, records and reporting. The SWAP training platform is a web application, delivering an entire training program without manual administration.

How do you deliver online training with certificates from a WordPress website?

A learning management system built on WordPress with LearnDash and BuddyBoss can deliver structured courses, assessments, learner accounts and progress tracking, generate personalised PDF certificates automatically, and take payment through Stripe, including subscriptions and organisation packages.

How much does a custom learning platform cost to run day to day?

Built correctly, very little. Automating enrolments, certificates, notifications and CRM and accounting integration means each new learner adds no administrative workload. The Hunter Melanoma Foundation team runs the SWAP platform themselves, without developer involvement.

Start with the program, not the platform

The SWAP build worked because we started with the program: who the learners are, where they train, what employers need to prove, what the Foundation’s team could sustain. The technology choices followed from those answers, not the other way around.

If your organisation delivers training, membership content or certification, and the administration is holding back the mission, we should talk. Book a free initial consultation with the HyperWeb team, or browse more of our case studies to see what purpose-built looks like in practice.

By Brendan Brooks, Founder and Managing Director of HyperWeb. Brendan has spent 25 years helping organisations turn their websites into working parts of the business rather than separate islands.

The post Your Website Can Run a Program, Not Just Describe It: Building SWAP for the Hunter Melanoma Foundation appeared first on HyperWeb.