AI use cases
Every AI use case needs different data. Find out what yours needs.
Every AI proposal assumes your data is fine. Nobody has checked. The model is the easy part; whether your data can support the use case is the hard one. Each page below sets out what one AI use case needs, which systems it lives in, where it breaks between them, and the questions to ask before you sign.
Starts with a 30-minute call. Commits you to nothing.
By use case
Pick the AI you’re being sold.
Revenue
AI on the CRM, pricing and the customer base.
- Revenue forecasting Needs won deals that match invoices, and close dates nobody moved after the fact. Needs: CRM · Billing and finance · Spreadsheets and documents
- Lead scoring Needs clean win and loss history, and one record per person across marketing and CRM. Needs: CRM · Marketing and web
- Customer churn prediction Needs one customer ID across CRM, billing and support, and a renewal date on every contract. Needs: CRM · Billing and finance · Support · Product and warehouse · Spreadsheets and documents
- Pricing and rebate optimisation Needs net prices that match the ledger, and one product code from quote to invoice. Needs: CRM · Billing and finance · Operations · Spreadsheets and documents
- Personalisation Needs one view of each customer, and consent that agrees in every channel. Needs: CRM · Billing and finance · Marketing and web · Operations
Finance
AI on the ledger, cash, payments and documents.
- 13-week cash flow forecasting Needs invoices, payments and terms that reconcile, in the system rather than a spreadsheet. Needs: Billing and finance · HR and payroll
- Financial close and FP&A Needs sub-ledgers that agree with the ledger, and one chart of accounts across entities. Needs: Billing and finance
- Fraud detection Needs complete, consistent transaction history, and approvers who still work for you. Needs: Billing and finance · HR and payroll · Spreadsheets and documents
- Document intelligenceWith your team Assessed with your team, not from exports alone. Needs: CRM · Billing and finance · Spreadsheets and documents
Operations
AI that runs processes, plans demand or talks to customers.
- Process automation Needs a process recorded the same way every time, with keys that carry from step to step. Needs: CRM · Billing and finance · HR and payroll · Operations · Spreadsheets and documents
- Demand forecasting Needs order history without gaps or duplicates, and stock-outs recorded somewhere. Needs: Billing and finance · Marketing and web · Operations · Spreadsheets and documents
- Customer service automation Needs tickets tied to customers and contracts, and one current version of every policy. Needs: CRM · Billing and finance · Support · Spreadsheets and documents
- Natural-language analyticsWith your team Assessed with your team, not from exports alone. Needs: CRM · Billing and finance · HR and payroll · Product and warehouse · Spreadsheets and documents
The dependency map
Which systems each use case leans on.
Count of requirements that live in each family of systems. The more columns a use case touches, the more gaps between systems it has to survive, and the more its Truth Score™ (a score out of 100 for how far your data can be trusted) leans on Connectivity: whether the systems agree.
| Use case | CRM | Billing and finance | HR and payroll | Support | Marketing and web | Product and warehouse | Operations | Spreadsheets and documents |
|---|---|---|---|---|---|---|---|---|
| Revenue forecasting | 4 | 3 | none | none | none | none | none | 1 |
| Lead scoring | 4 | none | none | none | 3 | none | none | none |
| Customer churn prediction | 2 | 4 | none | 1 | none | 1 | none | 1 |
| Pricing and rebate optimisation | 2 | 4 | none | none | none | none | 1 | 2 |
| Personalisation | 2 | 1 | none | none | 5 | none | 1 | none |
| 13-week cash flow forecasting | none | 5 | 1 | none | none | none | none | none |
| Financial close and FP&A | none | 5 | none | none | none | none | none | none |
| Fraud detection | none | 5 | 2 | none | none | none | none | 1 |
| Document intelligence | 1 | 2 | none | none | none | none | none | 4 |
| Process automation | 2 | 3 | 1 | none | none | none | 4 | 1 |
| Demand forecasting | none | 4 | none | none | 3 | none | 4 | 1 |
| Customer service automation | 2 | 2 | none | 4 | none | none | none | 1 |
| Natural-language analytics | 1 | 3 | 1 | none | none | 4 | none | 1 |
Drawn from the requirements on each page. Your own estate decides which systems hold each requirement.
Also by
Start from a system or an industry instead.
- 21 system guides Data quality for AI, by system Salesforce, HubSpot, NetSuite, SAP, Xero, Workday and more: what each system is right about, and where it disagrees with the systems beside it.
- 7 industries AI readiness, by industry The AI plays proposed in your sector, the systems a mid-market company there runs, and the gaps that sink the plays.
What you get back
See what comes back: a real sample memo.
One fictional B2B software company, Example Co: 72,837 records in 7 exports from its CRM, finance and HR systems, run through the same engine and the same checks as a paid Truth Score Read. Each system passes its own checks. Matched across systems, only 13% of deals tie to an invoice.
Questions
AI use cases, answered plainly.
Anything else, ask on the 30-minute call. It commits you to nothing.
Book the callWhat is AI use case readiness?
Whether your data can support a specific AI use case: the records it needs exist, are complete enough, and agree across the systems they live in. It is a different question for every use case, which is why each one has its own page.
Why does each use case need different data?
A churn model needs one customer across CRM, billing and support. A cash forecast needs invoices, payments and terms that reconcile. Fraud detection needs clean suppliers and approvers. The same company can be ready for one and blocked on another.
How do you test a use case against our data?
With a Truth Score Read. You send exports from two or more of the systems behind the use case. Our engine matches the customers, people, deals and invoices across them, and every red flag is checked by hand against sample records before sign-off. Within 5 working days of complete exports you get a signed memo: your Truth Score, a Leverability read on the use case (which AI ideas you can use now, after fixes, or not at all: Deploy now, Needs work or Blocked) and a verdict: sign, fix first or walk away.
Which use cases can’t be tested from exports?
Document intelligence and natural-language analytics depend on documents, data models and definitions as much as records, so we assess them with your team in a working session, alongside exports.
Do you recommend AI vendors?
No. We sell no AI, no software and no fixes. We take no referral fees, commissions or margin from anyone. Our fee is fixed before we start, so a ‘sign’ pays us exactly what a ‘walk away’ does. We tell you whether your data can support the use case, and what to fix first.
Do you put our data into an AI model?
The Truth Score is rule-based: the same data, scored on the same date, gives the same score, and no AI model decides it. We do use a language model (Anthropic’s Claude, through its commercial API) to read the plan or proposal you send and to help draft the memo’s wording from our findings. A person checks every line before it reaches you. If you’d rather we didn’t use it on your engagement, say so and we do that part by hand.
One use case, tested against your data. Before you sign.
Tell us the AI you’re being sold. Send exports from two or more systems behind it. A signed verdict within 5 working days of complete exports. Starts with a 30-minute call. Commits you to nothing.
Starts with a 30-minute call. Commits you to nothing.