Marketing technology strategy for UK SMEs.

Follow one enquiry all the way through.

You can have a good website, a CRM, an email platform and a reporting dashboard, and still have no clear answer to a simple question: what happened to the last enquiry?

So that's where I'd start. Pick a real enquiry and follow it all the way through. Someone finds a page, fills in a form, receives a reply, speaks to the team and either becomes a customer or doesn't. Which system holds the record at each point? Who sees it? What happens if a field is empty or the person replies directly to an email? You begin to see where the story gets lost.

For a small team, those details matter more than the number of tools. A missed handoff leaves someone waiting. Adding another platform won't sort it on its own.

Sit down with the people doing the work.

I'd put that journey on one page with the people who actually handle it. No enormous diagram needed. At each step, write down:

• The task: what the person or system is meant to do.
• The owner: who checks that it happened.
• The data: what is captured, where it goes and whether anyone trusts it.
• The handoff: what triggers the next step, including manual copying.
• The exception: what happens when the form fails, a lead is duplicated or somebody is away.

Now look at a handful of recent enquiries, not just the process you think you have. If the website says an enquiry went through but the CRM has no usable record, you've found something to investigate. If sales have the information but marketing cannot tell which page generated the conversation, that's a different problem.

I'd leave that conversation with a short list of gaps, likely causes and evidence we still need. Our practical CRM guide goes further into choosing and setting up the place where customer records live.

Which fix would make the biggest difference.

For each gap, I'd ask: how much does it affect an important journey? How confident are we in the data? How difficult would the change be to maintain? You can mark each answer high, medium or low. It's a way to make the trade-off visible, rather than everyone arguing for their favourite tool.

Say a website form sends enquiries to a shared inbox, and somebody copies them into the CRM later. The obvious idea is to automate the transfer. But first check the form fields, consent choices, duplicate rules and who should own the follow-up. If those are unclear, the automation will simply move the confusion faster.

I'd put a reliable handoff ahead of a clever dashboard. Get the record into the right place, give someone responsibility for it, and check that the follow-up happens. Off the back of that, reporting has something solid to describe.

A rare manual step with a clear check may be fine for now. There is no prize for automating every task.

Put the changes in an order the team can manage.

This is where the audit becomes a roadmap. I'd start by making the journey reliable: fix broken forms, unclear fields, duplicate records and missing owners. Agree what a complete enquiry record looks like, then send example submissions through and check what arrives at the other end. If that basic test fails, there's your first job.

Next, connect the handoffs that the audit showed were causing trouble. Maybe that's the website passing an enquiry to the CRM, or the CRM putting it in front of the right person. Write down what should happen when a connection fails, because somebody will need to spot and correct it.

Only once that holds up would I start looking at which pages and activities bring useful conversations, and where people drop out. Record each change and its date, then compare it with a sensible baseline. Otherwise you're looking at a report and guessing why it moved.

Pause between those steps. If the data is still messy, don't rush into an AI pilot that depends on it. If the handoff works but nobody follows up, the next change may be a team process, rather than another integration. For every change, name the owner and agree what a successful test looks like.

What our own stack audit found.

We mapped our own route from a search result to an article, a contact form, the CRM and the report. In our September 2026 audit, Search Console showed this guide appearing in search, but the live sitemap did not list it. Our GA4 history also mixed live-domain and preview events. Those counts told us what the analytics system recorded; they did not tell us how many people made a genuine enquiry.

The form and tracking check gave us a useful warning. The local Astro form sends its fields and consent choices to a HubSpot Forms endpoint, while an older tracking trigger could count a failed form attempt. We treated that event as a measurement issue, rather than dressing it up as a lead number.

Off the back of the audit, we put the work in order. The website and SEO owner checks the guide's sitemap and canonical coverage. The CRM owner tests field mapping, consent, routing and notifications with an authorised submission. The analytics owner defines a success-only event without personal form data and tests it under each consent choice. Each step has an owner and a way to check the result.

That is the value of an audit and roadmap for me. You can point to what you found, agree who handles it and tell whether the change worked. Our SEO workspace walkthrough shows how we keep those page decisions and their evidence together.

And what about AI.

I think AI can be really useful for a defined piece of work, such as drafting a first pass of copy or organising information for a person to review. But I'd want to know which job it is meant to help with before adding it to the stack.

Choose one task. Decide what information it may use, what a good output looks like and who checks it before anything goes to a customer. Try it on a small sample, note the errors and compare the time and quality with the current way of working. Obviously AI won't get everything right. If it helps, keep going. If it doesn't, you've learnt that without reorganising the whole team around a tool. The AI workflows guide looks at that process in more detail.

Ultimately, I want a team to be able to say what happened to an enquiry without opening five tabs and asking three people. Map the journey, fix the gaps that matter, and then add the technology that earns its place. If you want another pair of eyes on your stack, we can work through the audit and roadmap with you.