Time to
First Intelligence.
Every route to AI in a portfolio company has two clocks running. Most business cases only measure the first one.
Get theĀ five questions worth asking before you fund the next initiative
We will email you the PDF. No spam, unsubscribe at any time.
Only one of them touches EBITDA
Time to set up
Procurement, platform, migration, integration. Visible, easy to plan, quoted in every proposal. Nothing is learned while it runs.
Time to first intelligence
The day someone acts on an answer the system produced. Rarely quoted, almost never in the business case, and the only clock that moves the number.
of AI projects are expected to be abandoned through 2026 where the data is not ready for them. Not because the models failed, but because nobody checked the foundation first.
Gartner, February 2025. Survey of 248 data management leaders.
Gartner also found 63% of organisations either lack the right data management practices or do not know whether they have them. Read those two numbers together and the conclusion is uncomfortable: most organisations do not have data that is ready, and most projects that depend on it being ready get abandoned. The standard response is a cleanup programme first, which is months of spending before anything is learned.
We do not require the data to be ready. We connect to systems as they are, thirty different instances, spreadsheets, emails, legacy databases, and unify the fragmented data into one source of truth ourselves. No cleanup, no migration, no IT project on your side. Then the gates run before anything is modelled and tell you within weeks what that data can carry. Readiness stops being the gate on your timeline, which is how time to first intelligence lands in weeks rather than quarters, and why that 60% does not have to include you.
Where the gains actually come from
The data
Not volume. Whether it can support the specific decision you want to make. Fragmented across systems is normal, not a blocker, and the gates say within weeks what it can carry.
The workflow
Whether the output lands inside how someone already works. Of 25 attributes McKinsey tested, workflow redesign had the biggest effect on EBIT impact.
The model
Fit for the task, and replaceable. Rarely why a project fails or succeeds. A component with a shelf life, not a commitment.
The note also carries five questions to ask before you fund the next one, and an anonymised worked example from a real industrial order book where a third of the apparent customer churn turned out to be a data artefact.
The companion framework
Once you know the two clocks matter, the next question is which use case goes first. Our AI Use-Case Prioritisation Framework scores candidate ideas against seven criteria and returns a ranked shortlist, with the Excel score sheet built in.
Get the frameworkWant the same read on one of your portfolio companies?
We scope a fast read of one company's own data: where revenue is at risk, where margin is leaking, and whether the numbers can support the decision.
Talk to us