Step 3 · AI Value Tracking
Is the data under each AI investment still holding up? An independent re-read, every quarter.
Data drifts. A migration, a new billing system or a team that leaves can quietly pull an AI investment back from Deploy now to Blocked. Every quarter we re-read the same systems the same way: your Truth Score™ (a score out of 100 for how far your data can be trusted), the Leverability read (which AI ideas you can use now, after fixes, or not at all), and progress on the roadmap. We measure the data, not the return, and tell the company’s leadership early when something slips. Like every DataTruth service, it sells no AI, software or fixes.
Starts with a 30-minute call. Commits you to nothing. Tracking follows a Truth Score Read or an AI Investment Review, so there is a baseline to measure against.
Q3: Connectivity down 9. The new billing system created 63 customers with no CRM ID. Churn prediction held at Needs work.
What each quarter covers
Three questions, asked the same way every quarter.
Can the data still be trusted?
The Truth Score and the Leverability read, re-run on fresh exports against the same field list, so a change in the score is a change in your data, not in our method.
Is the roadmap on plan?
Each fix on the remediation roadmap checked against its own test for done: done, on track, or slipping, measured in your records rather than reported in a meeting.
What is about to go wrong?
Early warning when something moves the wrong way: a new system creating records with no match, ownership gaps after a reorganisation, an AI play sliding back from Deploy now.
A quarterly read
One document a quarter. Signed off, like everything else.
Not a dashboard to log in to. A short written read for the company’s leadership, then a call to walk through it.
Truth Score™
67 ↓ 4 on Q2
Band B, minor issues · still above the baseline of 57
Leverability
2 of 4 plays deploy now
Was 1 of 4 at baseline
Early warning
1
Billing migration, see below
Roadmap against plan
- P1One customer ID across CRM and billingDone
- P2Trace closed-won deals with no matching invoiceDone
- P3Payment terms into the ERPOn track
- P4Discount approvals into the CRMSlipping
Leverability, this quarter
- Revenue forecastingNeeds workDeploy now
- Pricing optimisationBlockedNeeds work
- Customer churn predictionNeeds workHeld: see early warning
Early warning
The new billing system went live in August and created 63 customers with no CRM ID. Churn prediction depends on that match, so it stays at Needs work this quarter although the original fix is done. Add the CRM ID to the billing onboarding step before the Q4 renewal cycle.
Checked by hand · Signed off by DataTruth
Illustrative example, not a client.
Who it’s for
Sold to the company. Not run on it.
AI Value Tracking is bought by each company’s own leadership, for their own use. When an investor introduces us, the company is still the client: it sees every read first, and decides what it shares. What the fund sees is what the board sees.
What it is
- A quarterly read for the company’s own leadership
- The same method and field list as the baseline
- Evidence the leadership team owns, and shares with its board and investors as it chooses
- Per company, at a fee fixed per quarter
What it isn’t
- Surveillance of a company on behalf of a fund
- A dashboard anyone has to log in to
- Connections left running inside your systems
- A way to sell you the fixes
How a quarter runs
Same exports, same rules, a fresh read.
- 01
Fresh exports
One person runs the same standard exports as last quarter, against the same field list. No connections stay open between quarters.
- 02
Like for like
The same matching and the same rules, so movement in the Truth Score or the Leverability read reflects your data, not a change in method.
- 03
Roadmap checked
Every open fix tested against its own definition of done, in the records themselves.
- 04
Read and call
A signed quarterly read and a call with the leadership team. Early warnings go out as soon as they show in the data.
Questions
Before you sign up for a year of quarters.
What a quarter covers, who sees it, and what it costs.
Book the callWhat is AI Value Tracking?
A quarterly re-read, company by company: is the data under each AI investment still holding up? The Truth Score and Leverability read again, progress on the roadmap against the plan, and early warning when something slips. It measures the data, not the return. It is sold to each company’s own leadership, not run as surveillance for anyone else.
What do you keep after the engagement?
Quality metrics and the memo: figures such as how many customers matched between two systems, fill rates for key fields, and the Truth Score. The memo cites a handful of sample records as evidence; everything else is deleted after the engagement.
Do you work with private equity and other investors?
Yes. Start with one company, then use the same method across the portfolio: same pillars, same rules, same caps, so a 64 is built the same way at each company. Where a pillar is not assessed, the memo says so. Each company’s leadership is part of the work, and AI Value Tracking is sold to each company’s own leadership.
Do we need to log in to anything?
No. DataTruth is a service, not software. You send exports; we send back a signed memo, a review or a quarterly read, and walk you through it.
What does it cost?
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.
You’re new. Why should we trust the verdict?
We are new, and we won’t pretend otherwise. That is why the method is published, including the cap that stops strong pillars hiding a weak one. Every finding points to records your own team can open and check, and the scope line says what we didn’t check. You can stop at the memo. Our first clients get a founding rate in exchange for a private reference call once they’re happy with the work, and only with their consent: we never name a client, fund or target publicly without written consent.
Find out it’s slipping before the board does.
Data drifts. A quarterly re-read, on the same method, tells the leadership team first.
Starts with a 30-minute call. Commits you to nothing.