# Rank clients by LTV

The skill sums every paid invoice by customer and sorts from your biggest down, with real currency amounts and an as-of date. To be upfront: this ranks revenue paid to date, not a predictive model of future value, because there is no churn or retention data behind it.

`npx skills add wellapp-ai/skills --skill rank-clients-by-ltv`

Who are our best customers?

Stripe leads at €412K paid to date, then Mercury at €288K, Qonto at €196

## How the skill answers

Every step of the run, in order, with the card each one returns. A step that hands its work to another skill says which one.

01. You ask

Who are our best customers?

02. Pin the workspace

### Workspaces
    Pick the workspaces to work in.

03. Confirm the connections this answer needs

### Connect a tool
    Pick a tool to connect it to Well.

04. Verify the data itself has landed

05. Resolve your own company

06. Resolve the time window

07. Query paid revenue by customer

08. Normalize currency

09. Sort and limit

10. If any required step errors or returns unusable data

11. You get

### Companies
    | Company Name | Total amount |
    | --- | --- |
    | Stripe | €412,000.00 |
    | Mercury | €288,000.00 |
    | Qonto | €196,000.00 |
    | HSBC | €154,000.00 |

Stripe leads at €412K paid to date, then Mercury at €288K, Qonto at €196K, and HSBC at €154K. This ranks realized revenue actually paid, not a predicted future value.

## Frequently asked questions

Is this really lifetime value?
It is realized revenue to date, which is the honest version of the question. A true LTV model needs churn and retention data that invoices alone cannot provide.

Does it count unpaid invoices?
No. Only invoices actually paid count toward the ranking, so the order reflects cash received rather than cash promised.

Can it rank by something else?
Yes. Ask for a specific window, a currency, or a segment and the ranking recomputes.