By Kim Leary
The short version
- AI got sold as a writing tool, so that’s how most teams use it. Writing is the part it’s worst at making distinctive.
- The data exists, but it’s spread across tools that don’t talk to each other. AI can only read what you give it.
- None of it shows up on a report, which is why it keeps happening.
- Most of the advice on AI stops at the surface. So does most of the action.
- Join up the data first. Then point AI at the questions that actually matter to the business.
- It starts with agreeing the numbers, not with choosing a tool.
The board wants to know what you’re doing with AI. Your team has been told to go and try it. So they’re using it to write posts, draft emails and make graphics, faster than before.
None of that answers the question the board is really asking. Is it making the business better? Using AI to analyse marketing data is where that answer is.
Why it happens
AI got sold as a writing tool, so that’s how most teams use it. Writing is the part it’s worst at making distinctive.
Ask AI to write a post and you get the average post. Ask it to find patterns in three years of results and you get something no person on your team had time to find.
We’ve seen a marketing lead spend years trying to work out what made social posts perform. Short posts did well, then a long one did just as well, and the conclusions kept contradicting each other. AI went through the lot in minutes and found five or six clear patterns in what worked and what didn’t. That’s the kind of work that used to need hundreds of hours, so it simply never got done.
Why most businesses can’t do it yet
The data exists, but it’s spread across tools that don’t talk to each other. AI can only read what you give it.
Ads in one place, the website in another, enquiries in the CRM, revenue in finance. Nobody has joined it up, so there’s nothing whole for AI to look at.
It ends up summarising one spreadsheet at a time, which is useful but small.
What it’s costing you
None of it shows up on a report, which is why it keeps happening.
What doesn’t fix it
Most of the advice on AI stops at the surface. So does most of the action.
“Just trying AI is enough.” Trying it is how you start. It isn’t a plan, and it won’t give the board an answer.
“This tool will solve everything.” A tool without connected data and an agreed goal just gives you faster confusion.
“AI is for content.” It’s the least distinctive thing it does for you, and the one your buyers are already starting to spot.
“We can figure it out internally.” You can, eventually. Most teams don’t have the time, and nobody wants to be the one who asks the basic question.
What fixes it
Join up the data first. Then point AI at the questions that actually matter to the business.
Our connected data and reporting work joins the tools you already pay for, so spend, leads, pipeline and revenue sit in one place. That’s what gives AI something worth reading.
Then AI tools and training sets your team up to use it properly: which tools for which jobs, the guardrails, and half a day of training on your own work and your own data. From £2,800 + VAT.
If the bigger problem is that nobody has decided what AI is for in your business, start with growth consultancy. One day with your leadership team to agree the direction, then we work to it month by month. The day is £2,800 + VAT.
What joining it up looks like
It starts with agreeing the numbers, not with choosing a tool.
This is how we build it, in this order:
- Agree what counts as a lead. With sales and finance in the room, so the numbers settle arguments instead of starting them.
- Decide which system owns which number. Spend from the ad platforms and finance, website activity from GA4, leads from your CRM, revenue from finance.
- Follow a customer from first click to invoice. So you can see which marketing led to revenue, not just to a form fill.
- Build the Monday morning view. Where the pipeline is and whether you’re on track. Not forty charts.
Only then does AI get pointed at it. By that point it’s reading the whole picture, so what it finds is about pipeline and revenue, not just likes.
Keep reading
- What to tell your board about AI
- Stop moving the goalposts: agree what marketing success means
- Why AI-generated graphics all look the same
Questions about using AI to analyse marketing data
It can go through years of results in minutes and find the patterns people don’t have time to: which posts, pages and campaigns actually drive engagement and enquiries. It needs your data joined up to do it well.
You need it connected more than perfect. If your ad, website, CRM and finance data can’t be linked, AI can only look at one piece at a time.
No. It finds patterns. People still decide what to do about them, and that judgement is the part that’s hard to replace.
Start with the ones the board asks: which campaigns led to pipeline, where enquiries drop off, and what your best customers had in common before they bought. Agree what success means first, or the answers won’t settle anything.
Connected data and reporting is quoted, because it depends on how many tools you run and how well they already talk to each other. Tell us what you’re being asked for and we’ll tell you what it takes to answer it. AI tools and training starts from £2,800 + VAT.
Free resource
Build the business case. Work out what joining up your data is worth before you take it to the board.
The one thing to do next
Connected data and reporting
We join up the tools you already pay for, so spend, leads, pipeline and revenue sit in one place, and AI has something worth reading.
