Someone on your team fed an AI tool a description of what they wanted, and after a bit of back and forward, an hour later they had a dashboard. Charts, KPIs, the lot. It looks great. It probably looks better than the last thing you paid a consultancy for 🫣
So, here’s the question doing the rounds right now: if a dashboard takes an hour, why does proper analytics take twelve weeks (if you’re lucky), and cost what it costs?
It’s a fair question, and the honest answer isn’t ‘your mock-up is rubbish’ It probably isn’t. The answer is that what your team made is only a fraction of what your business actually needs, but happens to look similar on a screen.
Let’s dig into why.
The picture is the last 10% of the journey
Every dashboard worth trusting sits at the end of a complex data journey. Most of that journey is invisible. The chart on top is the only bit anyone ever sees, and it’s the bit AI just made easy to create a pretty picture of.
But making the picture at the front accessible doesn’t make the journey behind it any shorter. It just moved the only visible part to the front and hid everything that actually matters.
Let me wizz you through the journey quickly and I’ll show you where the hour went, and then where the twelve weeks go.
Stage one: the question
Before any data moves, someone has to decide what the business actually needs to know. Not “show me everything” – the specific questions a leadership team would change a decision over, or take a strategic action based on. Where’s margin leaking on the Anytown contract? Does our absence figure match what payroll actually paid out? Are agency invoices lining up with the shifts we scheduled, is our cover team the right size with the right mix of experience?
Your team’s mock-up can only ever show you what answering these questions could look like, not the actual answers that could change the business.
But, you’ve guessed it, answering “what do we want to see” is the easy part. It always was.
Stage two: capture
Now you need to identify the data that answers those questions. And in FM, that data doesn’t live in one place. It’s scattered across your T&A system, Finance, H&S, Operations, HR, Audits and spreadsheets, loooooots of spreadsheets. Different platforms, different formats, different owners, different language, built at different times by people who’ve probably since left 🤷♂️
A mock-up skips this stage entirely. The numbers on it were typed in, synthesized, estimated, or pasted from a spreadsheet that was accurate on the morning someone exported it.
Production analytics can’t skip it. The numbers have to come from the actual systems, automatically, on a schedule, cleaned, modelled and ready to rock n roll.
Stage three: connect
This is where it gets hard (by hard I mean way too hard for AI still).
Your systems have never agreed with each other. They were never designed to. T&A tells one story, payroll tells another, the agency invoices tell a third. Connecting them so they speak the same language – so a “shift” means the same thing in three systems that have always defined it differently – is real engineering. It’s slow because the problem is genuinely messy and needs thought + no two organisations are the same, so this is as much about the way your business operates, as it is the systems being used.
And, for my money, it’s where the value actually lives. This is the stage where important questions start to get answered. Where payroll variances you’ve been quietly carrying for years finally show up. Most FM providers walk out of this part of the work knowing more about their own operation than they learned in the previous five years. The mock-up, by contrast, just has either nonsense or something that was right at the time it was designed in it.
Stage four: reconcile
Connecting the systems isn’t enough (ask Sophia Lee how long her team spend on this!!) You have to make them agree… and prove it.
Does the absence figure match what payroll paid? Do the agency invoices match the scheduled shifts? When two systems disagree – and they will – which one’s right, and how do you know? Production analytics resolves those disagreements before a number ever reaches a screen, and it keeps a trail behind every figure: source, transformation, timestamp. Ready for the day a client or an auditor asks where a number came from.
A mock-up has nothing behind any figure. That’s fine when it’s a sketch on a leadership team’s table, or an idea in an SLT meeting. It’s a problem the moment you try to price a bid or staff a contract off it – obviously.
Stage five: the dashboard / the picture
Now, finally, you build the dashboard. The bit someone whizzed up in an hour with a Claude pro subscription.
Except now it’s built on a connected, reconciled layer of real data, in a platform that’s going to run it every day from here on. Same picture they imagined, completely different product reality underneath.
Stage six: keep it true
And this is the stage everyone forgets – a dashboard isn’t a file you make once. It’s a service that runs, forever.
Feeds break. Export formats change without warning. People leave and take the knowledge with them. The real question was never “can someone draw this” – your team already proved they can. It’s “who keeps the number true on the second Monday of the month, when the T&A export changes format and payroll stops matching the rota?”
That’s not prompting, that’s a department, a data department, Your Data Department.
So, what are our clients actually paying for?
Not (just) the picture. The picture is the smallest item on the list, and it should be quick, you could argue that the technology has just helped the front end work take as much effort as it always should have done.
You’re paying for the journey: the questions worth answering, the people to help you identify those questions, the audit of whether your data can answer them today, the connected layer where the real insights actually live, and the people who keep the numbers true every week after go-live. Most people can draw the chart now. The hard parts are a) making the number in it true, and b) keeping them true.
That’s the difference between a mock-up of a dashboard and a dashboard. One shows you the day it was drawn. The other helps you run your business, on-going.
What to do with the mock-up
At Datore we always say – Don’t bin it, bring it.
If you, or your team has sketched what good visibility would look like, you’ve done something most FM providers never get round to – you’ve shown, in pixels, exactly what you want. That’s the perfect starting point for a proper conversation about what it would take to make it real: which system feeds each KPI, whether that data’s usable today, how we could make it usable through great engineering, and what it would cost to keep it honest every week.
You know what? You bring the mock-up & we'll bring the questions - let's chat.
Contact Us
