Most data projects fail on trust and clarity, not technology. The four questions to ask any data partner before you sign, and why proof beats a slick promise.

You are buying trust, not software

Data projects rarely fail because the technology was wrong. Microsoft Fabric works. Power BI works. Azure works. They fail because, somewhere along the way, the people paying for it stopped believing in it, and a platform nobody believes in is worse than none at all, because now you are paying to maintain something everyone quietly works around.

We cover the practical side of this in our data strategy and leadership work, and Why The AAP Is Not A Traditional Retainer looks at a closely related question.

Nobody puts that in a proposal. When a mid-market business commissions a six-figure data engagement, it is not really buying software it could license itself. It is buying confidence: that the thing will work, that the people will still be there when it does not, and that someone will tell it the truth when a number looks wrong. Trust is the actual product. The dashboards are just where the trust becomes visible.

Why the confident pitch should worry you

The instinct, when serious money is on the table, is to want certainty up front. A fixed price, a guaranteed outcome, a deck delivered with total conviction. The trouble is that conviction is the cheapest thing in the room to fake and the hardest thing to hold anyone to. Anyone can promise a transformation. Far fewer will prove a small piece of it first, at their own risk, before you commit the budget.

So we work the other way round. Scope first, price after. Prove the value on something small and real before anyone signs up to the big number. It is slower to start, and it loses its share of beauty contests to a slicker deck, and we are at peace with that, because the clients it wins tend to stay for years. Proof is simply a more honest form of selling, and it quietly filters for the people who want a partner rather than a magician.

We did exactly that not long ago for a mid-market, sales-led business: a paid proof of value first, to demonstrate we could deliver what we had promised before either side committed to the full programme. It is a slower way to begin and a far surer one, and it meant the bigger decision was made on evidence rather than on a pitch.

Four questions worth asking anyone, us included

If you are weighing up a data partner, four questions reveal more than any case study. Ask them, and watch how comfortable the answers are.

Who will actually do the work, the people in this room or someone you will never meet. Can we start with a small, paid proof rather than a leap of faith. What happens, specifically, when you get something wrong, because you will. And how long do your people stay, because the person who learns your business is the one you least want to lose. A partner who answers all four without flinching is rare, and worth a great deal.

This is why the Analytics Acceleration Programme is built in stages rather than sold as one heroic leap, and why our analytics strategy work often opens by talking a client out of the largest version of what they asked for. The right-sized thing that gets trusted beats the ambitious thing that gets abandoned, every single time.

The platform you buy is only ever as good as your trust in the people who built it. Spend your diligence there, not on the feature list. Software you can always swap. A partner who tells you the truth and is still standing beside you in year three is the genuinely hard thing to find.

If any of this sounds familiar, talk to us about your data.

Related reading

• A Fixed Price For Unscoped Work Is A Red Flag
• Stop Counting Dashboards. Count Decisions.
• What The Pub Trade Teaches About Your Numbers