3 Things to Check Before Paying for More Website Traffic – An Interview with Ben Crowther
Ben Crowther brings around 18 years of marketing experience to his work with Crow Flies Growth Systems and Aiden Solutions, connecting websites, business information and AI to improve how companies handle customer enquiries. His starting point is often what happens after someone clicks “submit”: whether the right person receives the enquiry, has the information they need and follows up.
In this exclusive interview, Ben explains where digital systems lose potential customers, why useful AI depends on reliable company knowledge, and how smaller businesses can identify the first process worth automating.
Ben Crowther, Founder, Crow Flies Growth Systems & Co-founder of Aiden Solutions
Ben, through Crow Flies Growth Systems and Aiden Solutions, you’re changing how businesses use websites and AI. What inspired you to take this approach?
Frustration, mostly. A business would have a website, marketing, a customer database and reporting, all apparently helping it grow. Meanwhile, the owner was copying information between them and chasing up whatever had fallen through the gaps. An impressive amount of technology to leave someone doing the same job twice.
That was the part I wanted to fix. A customer could find the business, ask a perfectly sensible question and still disappear because the information was poor or the follow-up never happened. Buying another subscription didn’t seem like much of an answer.
With Crow Flies and Aiden, I’m interested in connecting that whole process. Help the right people find the business, give them useful answers, make it easy to enquire and make sure someone can act on what comes in. Then show the business what actually happened.
That’s where AI becomes useful to me. Can it remove repetitive work or help somebody get a better answer? If it adds another thing the owner has to manage, we need to have another look at it.
What separates a connected growth system from a traditional business website, and why does that distinction matter?
I’d look at what happens after somebody fills in the form. That tells you far more than the homepage does.
A basic website sends an email. A connected system can carry the product and application details with the enquiry, put it in the customer record, send it to the right person and start the appropriate follow-up. You can then see whether it became a conversation, a quote or a sale.
The useful bit is that the customer doesn’t have to explain everything again, and the team doesn’t have to reconstruct the enquiry from three different places.
Business knowledge, content, voice tools, automation and reporting can all support that process. Which ones you use depends on the problem. I’m quite comfortable leaving a feature out if it has no useful job to do.
It also changes what happens after launch. You can see where people get stuck and improve that part of the system. Otherwise, nobody can tell you where the enquiries went, which makes the next round of improvements rather speculative.
What led you to develop Aiden Solutions around a company’s own knowledge rather than relying on general-purpose AI tools?
Because knowing a lot about the world doesn’t mean an AI tool knows which version of your product manual is current.
If a dealer asks whether a component fits, a fluent, confident guess is still a problem. Confidence is a fairly unhelpful quality when the answer is wrong.
Aiden’s Business Brain is built around the idea of organising approved business knowledge so it can support customer conversations, staff, content and voice tools. That means the actual product information, manuals, policies and procedures the business wants people using.
There’s some unglamorous work involved: deciding which information is right, who can see it and who keeps it current. Give the system contradictory documents and you haven’t solved the contradiction.
Using the business’s own information can reduce the risk of invented answers, but it still needs testing and oversight. I want it to recognise when it doesn’t have enough information and bring in a person. That matters particularly with technical advice, safety or an expensive purchasing decision. Sometimes the most useful answer is, “We need to check that.”
You’ve suggested that the traditional five-year website cycle is disappearing. What is driving that shift?
I’m not convinced a website needs replacing simply because it’s had another birthday. I am convinced it needs attention before then.
The business keeps changing after launch. Products change, customers ask different questions, the sales process develops and search moves on. If the site stays exactly as it was, it gradually starts describing a business that no longer quite exists.
My preference is to keep improving it while it’s in use. Look at what’s happening, fix the next worthwhile problem and check whether the change helped. That might be clearer product information, a shorter enquiry form or a better connection to the customer database.
It’s easier to have a sensible budget conversation when you can explain the problem you’re fixing. “It looks a bit old” is a fairly thin commercial argument on its own.
Sometimes a rebuild is the right call, especially if the platform can’t support what the business needs. But I’d like that decision to come from evidence, rather than the calendar.
How is the rise of AI-powered search changing the way businesses in industries like automotive and manufacturing need to approach online visibility?
It’s making vague marketing copy even less useful, which is probably overdue. Someone looking for a component wants to know whether it fits, what it can handle and where it can be used. They’re unlikely to be reassured by another paragraph about your passion for excellence.
For automotive and manufacturing businesses, I’d start with accurate product pages, clear compatibility and application information, and useful technical explanations. Put important details where people can read them, with the current documents available when they need more depth.
The SEO fundamentals still matter. Google’s guidance says its AI search features use the usual Search requirements, with no special schema needed. Clear, useful content and good technical foundations remain important. Nobody can guarantee inclusion.
The opportunity is to make the expertise already inside the business accessible online. That takes more thought than producing another batch of generic articles, but it gives customers something they can actually use. Then look at the enquiries and commercial value as well as the visibility. A mention in an AI answer isn’t a sale.
When a business is attracting website visitors but struggling to generate enquiries, where do you usually look first?
Before getting excited about a redesign, I’d send a test enquiry. Does the form work? Does it reach the right person? Can you use the phone link on mobile? Is the reporting counting real enquiries? Fairly basic questions, but worth answering before paying for more visitors.
Then I’d look at who is arriving and what they came for. If the traffic is wrong for the business, a bigger traffic number won’t improve matters. It’ll just make the report look busier.
Once the right people arrive, can they understand the offer and find what they need? Is there enough detail to make a decision? Is the next step obvious, or are we asking them to complete a small administrative exercise before they can speak to anyone?
I’d also follow the enquiry beyond the website. Who responds, how quickly, and what happens next?
There’s usually a specific place to investigate. I want to find it before recommending more advertising or a new site. Otherwise, we’re charging the business to deliver more people to the same problem.
For a smaller business exploring AI and automation, how can it identify which processes are worth improving first?
Start with irritation. Ask the team which repetitive job they’d happily stop doing tomorrow. You’ll probably get a more useful answer than asking how the business can “embrace AI”.
Then work out how often the job happens, how long it takes and what goes wrong. Copying information between systems, finding approved answers and preparing routine reports are sensible places to look.
Choose something with clear inputs, a result you can check and manageable consequences if it fails. Write down the process first. Some jobs only need a straightforward automation; adding AI doesn’t automatically improve them.
Keep the first version small and give someone responsibility for checking it. If it produces customer-facing content or advice, put an appropriate review or escalation step in place.
Compare the time, errors and response speed before and after. Include the time spent maintaining it. Saving an hour while creating two hours of checking is a fairly poor exchange.
Get one thing working reliably, then expand. Trying to automate the entire business at once tends to produce a very sophisticated mess.
With around 18 years in marketing behind you, how has your experience shaped the way you approach business and leadership?
Around eighteen years in marketing gives you a healthy suspicion of anything being described as the thing that changes everything.
Every few months, someone announces that the old approach is dead. There’s usually something new to buy shortly afterward. Some changes genuinely matter, but I want to understand what has changed for the customer before reorganising a business around it.
People still need to understand what you do, trust you and have a reason to take the next step. I keep coming back to that when looking at technology or strategy.
With leadership, I value being clear about what we’re doing and why. People should understand what’s expected of them and have room to point out when something doesn’t make sense.
I also think you have to be willing to change your mind. If the evidence says an idea isn’t working, the fact that it was my idea doesn’t make it worth protecting. I’d rather fix the problem than spend a meeting explaining why it isn’t one.
Looking ahead, what is the most important change business owners should make today to prepare for the next generation of digital experiences?
Get your business information into a state where people can actually rely on it. “It’s somewhere on the shared drive” leaves quite a lot of work for the next person.
Start small. Pick a product range or the questions customers ask most often. Find the current answers, sort out the contradictions and make someone responsible for keeping them right. Decide what can be public and what needs to stay internal.
That gives you something useful to build on. The website, a staff assistant and a voice system can all draw from dependable information. Without that groundwork, you risk giving the same inconsistent answers through more channels.
Then try it on a real task. Can the customer find the right answer? Can a staff member use it without asking someone to dig through their emails? What happens when the answer changes or isn’t there?
You don’t need to predict exactly what the next digital experience will look like to do that work. I’d start there.
Read more from Ben Crowther
