Step 2 · AI Investment Review
Bring every AI proposal, plan and idea in front of you. In about 2 weeks we test each one against your own records and rank them, test the plan or investment thesis play by play, and hand you the full Truth Score™ (a score out of 100 for how far your data can be trusted), a remediation roadmap and a board readout. A plan and budget the CFO can defend line by line.
Starts with a 30-minute call. Commits you to nothing. Start with a Truth Score Read if there’s one decision in front of you; a Review is for when the question is every bet, not one.
2 of 6 bets have the data they need today.
What a Review covers
Vendor proposals, internal ideas, board requests, the plays in your plan. Each tested against the data it needs and given one call: fund now, fund after fixes, or don’t fund yet.
The business plan, growth strategy, value creation plan or investment thesis: the AI plays pulled out, each mapped to the data it relies on, and checked.
Every pillar shown, what was measured and what wasn’t, with the cap applied when a critical pillar is weak.
Which AI ideas you can use now, after fixes, or not at all: Deploy now, Needs work or Blocked, with the reason in your own records.
What to fix, in what order, with an owner, an effort estimate and a test for done. Written to hand straight to your team or any outside firm.
The verdict, the ranking and the roadmap presented to your leadership team or board, with the decisions they need to take.
Every bet, ranked
Every AI idea gets the same test: what does it need from your data, is that data there, and can it be trusted? Then one call per bet, so the budget follows the evidence.
Illustrative example, not a client.
Your plan, tested
Send the business plan, growth strategy, value creation plan or investment thesis. We read it, pull out the AI plays and the growth targets that lean on them, and test the data behind each play. We don’t forecast whether a target will be hit; we show whether the data each play depends on is there and can be trusted.
Commercial and operational priorities
Protect the base: Aflag renewal risk early across the top 400 accounts, ahead of the FY27 price review.
Margin: Boptimise list prices and discount bands by segment, using AI on deal history.
Cash: Cshorten collections with an AI agent working the overdue ledger.
Works from board packs, investment memos and 100-day plans, under NDA like everything else. Illustrative example, not a client.
The full Truth Score
Rule-based, not AI-generated. Every pillar lists what was measured, a weak critical pillar caps the score, and any pillar the data can’t support is marked limited signal or not assessed rather than guessed.
58/ 100
Band CSignificant work needed
Weighted average 58. The weakest critical pillar, Connectivity at 52, sets a cap of 64, so here the cap doesn’t bite.
MeasuredDuplicate rate, valid emails, phones and dates, outliers, format consistency
MeasuredFill rates on required and key fields, whole-record completeness
MeasuredOwners assigned, stale records, lifecycle tracking, audit trail
MeasuredStandard formats, fields that agree with each other, dates in order, values in range
MeasuredMatch rate between systems, contacts linked to companies, shared IDs
MeasuredWhere each record came from, how it was changed, version history
Illustrative example, not a client. How the Truth Score works
The remediation roadmap
Every Review ends with a prioritised roadmap: what to fix, in what order, what each fix unblocks, who owns it, how long it takes, and how you’ll know it’s done. It doubles as the brief for whoever does the work, so you can compare quotes like for like.
Illustrative example, not a client.
Board readout
The Review ends with a readout to your leadership team or board: the verdict first, then the evidence, then the decisions to take. Written so a non-technical board can act on it in the room.
About 2 weeks
Week 1
Week 2
The clock starts when complete exports and the plan arrive. Your team’s time: the kickoff, short conversations with the owner of each bet, and the readout.
Who it’s for
Leadership teams
Every department has an AI idea, and many come with a vendor behind them. A Review tells you which bets your data can carry, in what order to fund them, and what it takes to unblock the rest, so the budget you defend to the board is built on your own records.
Investors
The value creation plan leans on AI plays that nobody has tested against the company’s systems. A Review tests the thesis play by play while the plan can still change, and gives the company’s leadership a roadmap they own.
Questions
How a Review differs from a Read, what it asks of your team, and who does the fixing.
Book the callA Truth Score Read answers one decision in 5 working days: is the data behind this proposal or plan fit for it? An AI Investment Review takes about 2 weeks and covers every AI bet you put in front of us: it ranks them, tests the plan or thesis play by play, gives the full six-pillar Truth Score, writes a remediation roadmap and ends with a board readout.
Yes. Send the business plan, growth strategy, value creation plan or thesis. We read it, pull out the AI plays and the growth targets that lean on them, map each play to the data it relies on, and mark each one Deploy now, Needs work or Blocked, with what it would take to close the gap.
Accuracy 22%, Completeness 18%, Governance 18%, Consistency 16%, Connectivity 14%, Lineage 12%. Accuracy carries most weight because wrong data produces wrong answers fastest. The methodology page lists what we measure for each one.
Which AI ideas you can use now, after fixes, or not at all. Each AI play or use case is marked Deploy now, Needs work or Blocked, with the reason in your own records. It speaks to whether the data is ready for a use case, never to whether the use case will succeed. (Nothing to do with debt.)
No. We never fix what we find and never run the remediation. But you don’t get a list of problems and a shrug: every Truth Score Read ends with a prioritised fix list, and every AI Investment Review with a full remediation roadmap, with owners, effort and a test for done, written to hand to your own team or any outside firm.
Your own team, or any outside firm you choose. The roadmap doubles as their brief, so you can compare quotes like for like. Because we earn nothing from whoever you hire, nothing on the list is there to sell you something.
For a Truth Score Read: a 30-minute call, one person running standard exports, and a walk-through call when the memo is ready. An AI Investment Review adds a kickoff, a few short conversations with the people who own each AI bet, and the board readout.
A fixed fee, agreed before we start, that never moves with the verdict. It is a small fraction of the proposal it tests; if it wouldn’t be, we’ll tell you on the call. You get the number on the first call, once we know which systems and decisions are in scope.
A Review can begin with a Truth Score Read on the biggest bet. Either way, the fee is fixed before we start.
Starts with a 30-minute call. Commits you to nothing.