By David Robinson
Every business has experienced hearing about a contract once it's been awarded to someone else, and the first reaction is usually to wonder why nobody asked you to quote. The uncomfortable truth is that, in most cases, you were never in the running because someone you haven't met put together the shortlist at a desk, using a search engine and an AI assistant.
We work predominantly with engineering, manufacturing, and industrial companies, and the pattern is fairly consistent. By the time an RFP lands in your inbox, the buyer has already done most of their thinking, and you only ever see the final stage of a process that started without you. This post walks through that process from the buyer's perspective so you can see where you're succeeding and failing.
How a shortlist actually gets built
Picture the person doing the buying. It might be a project engineer who's been told to find someone who can handle a particular piece of fabrication, a procurement lead who needs three compliant quotes by the end of the month, or an operations director looking to replace a supplier who let them down. Whoever it is, they're busy, the task is one of a dozen on their list, and their main concern is not getting it wrong.
So they do what all of us do. They search for the problem, ask a colleague or two, and increasingly ask ChatGPT, Copilot, or Google's AI to suggest companies that do what they need. That gives them a longlist, often eight or ten names, which they then whittle down to three or four by looking at each company's website. Only then does anyone send an RFP or pick up the phone - remember phones?
The important bit is that the first two stages happen without anyone contacting you. You don't know you're on a longlist, and you certainly don't know you were removed from it, which is why so many good companies write off lost work to price or relationships when the real reason sits much further upstream. If you sell into long buying cycles, our post on marketing for complex B2B sales cycles goes into the wider picture.
The first cut: being found at all
Buyers rarely search for your company name, and they rarely search for your product name either. They search for the problem in their own words, something like "corrosion protection for subsea pipework" or "small batch CNC machining stainless Scotland", and they take whatever comes back as a reasonable picture of who's out there.
For businesses, AI tools have made this first cut harder, because a traditional search page gives a buyer ten results and some adverts to scroll through, whereas an AI answer will typically name a handful of companies and stop there, and if you're not one of them you don't exist as far as that buyer is concerned. Those tools pick up companies whose websites clearly say what they do, for whom, and where, in language that matches how people describe the problem, and they tend to favour sites with real evidence over sites full of generic claims. We've written more about this in AI search and B2B: what engineering companies need to know.
This is where B2B SEO and AI search optimisation do their real work. This work aims to improve the chances of a qualified buyer finding your company when they're trying to solve a business challenge.
The second cut: the five-minute website test
Once a buyer has a longlist, they open each website in a new tab and spend a few minutes on it. Nobody reads everything; they look for answers to a few questions that decide whether you stay on the list.
Can I tell what you do straight away?
If your homepage talks about "innovative solutions" and "a commitment to excellence" without saying what you actually make or do, the buyer has to go digging, and most won't bother when there are seven other tabs open. A plain sentence that says what you do and who you do it for helps far more than polished copy, a point we've been banging on about for years in Plain English: do you speak it?
Have you done this before, for someone like me?
Buyers want proof that you've solved a similar problem in a similar setting. Project pages, sector pages and case studies that name the type of work, the conditions and the outcome do this far better than a list of services, and they're often the first thing a buyer clicks on after the homepage.
Do you meet the requirements?
Procurement leads, in particular, will look for the accreditations, standards, and supplier registrations their organisation needs before they'll even consider you. If your ISO certificates, sector approvals or supply chain registrations are buried in a PDF three clicks deep, or missing altogether, a cautious buyer will assume you don't have them and move on.
Are you a safe pair of hands?
This one is rarely spoken about, but it matters a great deal. Buyers are checking that you're a going concern, with real people, a real address and recent signs of life, because nobody wants to recommend a supplier who turns out to be two people and a website. Named staff, recent project news and a sensible contact page all help here.
The third cut: the reasons nobody tells you
Some companies get dropped for reasons that have nothing to do with their capability, and they'll never hear about it. The most common ones are a news section that stopped in 2019, which makes a buyer wonder whether you're still trading, a contact page with nothing but a form and no phone number or named person, and a website that looks the same as every other supplier in the sector, so there's nothing for the buyer to remember you by when they're comparing notes.
There's also a bigger, quieter factor: procurement is mostly about avoiding risk. A buyer who chooses a well-known supplier and has a bad experience can defend the decision, whereas a buyer who chooses an unknown one and gets burned has some explaining to do. Your website is often the only evidence they have to justify putting you forward, so if it looks less established than the company really is, you're asking them to take a risk on your behalf, and most won't.
Overseas buyers tend to apply all of this with even less patience, because they can't drive past your premises or ask someone local about your reputation, so the website is most of the evidence they have.
Sitting in the buyer's chair yourself
The good news is that you can test most of this in an afternoon without spending a penny. Here's what we'd suggest:
- Write down three or four ways a buyer might describe the problem you solve, in their language and not yours, then search for each one and note who comes up.
- Ask ChatGPT, Copilot and Google the same questions, phrased as a request for suppliers, and see whether you're named, who is, and what the AI says about them.
- Give your website to someone outside the business, ideally someone with a procurement background, and ask them to tell you within five minutes what you do, what you've done and whether they'd put you on a shortlist.
- Do the same exercise on two or three competitors who keep winning work you think you should have had. Our guide to interrogating your competitors' websites gives you a framework for this.
In our experience, the results are usually uncomfortable but point to a small number of fixable problems rather than a need to start from scratch. If you'd like a deeper look at turning your website into something that wins work, a good next read is how to build a lead generation website.
If you'd prefer a second opinion from an industrial marketing agency that spends its days on this, we're happy to run the exercise above with you and tell you what we find, including when the answer is that you don't need us. No hard sell, just come and have a chat.
Get In Touch
Frequently asked questions
How do industrial buyers find new suppliers?
Most industrial buyers start by searching for the problem they need solved, asking colleagues and, increasingly, asking AI tools such as ChatGPT or Copilot to suggest companies. That produces a longlist, which they cut down to three or four suppliers by looking at each company's website before sending any RFQs.
Why are we not being asked to quote for work we could do?
The most common reason is that you're being left off the longlist or cut at the website stage, both of which happen before anyone contacts you. Buyers who can't find you when they search for the problem, or who can't quickly see what you do, what you've delivered and which accreditations you hold, will usually move on to a supplier who makes that easier.
Does AI search change how suppliers get shortlisted?
Yes, because AI answers usually name only a handful of companies, where a traditional search results page shows ten or more. Companies whose websites state clearly what they do, for whom and where, backed by real project evidence, are more likely to be named, and companies that aren't named are effectively invisible to that buyer.
What should an industrial company's website show to get shortlisted?
It should say plainly what you do and who for, show evidence of similar work through project pages or case studies, make accreditations and supplier registrations easy to find, and give clear signs that you're an established, active business, such as named people, recent news and full contact details.
