The website, the AI and the contenta company your size should have in 2026.

How do you know your website is outdated?

When a buyer can read the year it was built.

A company that has operated for decades usually has a website built for the company it was: a smaller fleet, fewer sectors, projects that are no longer the biggest thing it has done. The signs are plain. The photography is old or stock. The projects page stops years ago. The staff portal link leads nowhere. It does not read properly on a phone. Search engines and the AI answers people now read cannot find anything on it to quote. Only 52.9 per cent of Australian businesses with 200 or more staff have a digital business strategy at all, and only 32.6 per cent measure what digital contributes; in construction it is 3.9 per cent. The website is where that shows first.

Does an old website actually cost us work?

Yes, before you ever hear about it.

In the largest recent survey of business buyers, 68 per cent began with a preferred supplier already in mind, and that supplier won 55 per cent of the time. In another, 94 per cent of buyers had ordered their shortlist by preference before speaking to any seller. The shortlist is built from what the buyer, their team and their procurement people can see, and the website is what they see. A site that undersells a 500 person company loses the tenders it never hears about. The last one we took on had a website that did not reflect how big the company was and a portal link that led nowhere.

Is Australia behind on AI, and does it matter to a company like ours?

Behind, and most of the businesses counted as using it are barely using it.

In the Australian Bureau of Statistics’ latest count, 12 per cent of Australian businesses used AI in 2024-25, and that is a tick box that counts a business where one person asks a chatbot questions. Among businesses with 200 or more staff it was 34.7 per cent. In construction it was 6.4 per cent, in manufacturing 9.3 per cent, in transport 1 per cent. The businesses that actually run on it are a fraction of those: when the UK’s statistics office measured depth, only one adopter in ten used AI extensively. The Productivity Commission put it plainly in December 2025: Australian businesses appear to be embracing AI less quickly than their international peers. In the United States, 37 per cent of businesses with 250 or more staff had used AI in the previous fortnight when the Census Bureau last asked. The tools those companies use are available in Melbourne today, at the same price. Being late is the opportunity: in most Australian industries the first company at your scale to run its operations on it, not just ask it questions, will be the only one for a while.

Aren’t we already using AI?

Replacing a search engine with a chatbot is not using AI in the business.

Most companies that say they use AI mean that some of their people ask ChatGPT or Claude questions and draft the odd email with it. That is a better search engine. It changes nothing about how the business runs: the enquiries still land in an inbox, the quotes still wait for tonight, status still costs a phone call. Using AI in a business means systems that answer from your own records, do the routine work without being asked, and hand people time back. That last part is the whole point as we see it: AI’s job is to give humans more of their day. The best measured trial of an assistant over a company’s own email and documents, across 66 companies and 7,137 staff and read from usage logs rather than surveys, found two hours a week back per person, and the reason it never showed up in the payroll is that people spent it on other work. Our own studio is the proof at small scale: two founders run a full stack agency because the back office runs almost entirely on AI, Claude, Codex, Grok and X doing the routine work between us. We build the same thing for companies with hundreds of staff.

What can AI actually do in a business like ours?

Answer questions from your own records, draft the routine work, and give people hours back.

Overseas, companies your size are already putting it to work: a 385 person engineering firm runs an assistant over its regulatory manuals and reports research time cut by 90 per cent; a builder ran a chatbot over a single job’s 21,000 documents and it answered 246 questions in 37 working days, 87 per cent of them correctly; a 100 year old machinery maker cut project documentation from more than a week to four to eight hours. In the system we are building for a 500+ person electrical contractor, the assistant answers what is on today from the live project data: what is lagging, who is carrying too much, what is waiting on someone. We build its tools ourselves, and they do one thing: read the system’s data and answer from it. Nothing else. Heavy instructions around what it may and may not say, each person seeing only what their role allows, and once it is live, we watch it hard. What is possible in your business depends on where the hours go, which is why the fastest way to find out is thirty minutes with us and the org chart.

Why does brand matter when we already have the relationships?

Because the buyer who has never heard of you is not in the market yet.

The Ehrenberg-Bass Institute’s rule of thumb for business buying is that up to 95 per cent of buyers are not in the market for what you sell at any one time; they enter later, and if they have never heard of you they do not consider you. It is a heuristic, not a survey, and the Institute says so, but the surveyed version says the same thing: 68 per cent of buyers start with a preferred supplier and that supplier usually wins. The relationships that built the company were made by people who are retiring. The next decade of work goes to the company the next buyer already knows.