Know your customers like never before
Per-customer value, risk and churn — so you collect smarter and keep the right customers.
Live product UI — sample data.
A score per customer
Lifetime revenue, average order value, payment delay and profit contribution roll into a single customer score.
See churn before it happens
Churn probability and revenue-at-risk from purchase recency and frequency — so you act in time.
Collect smarter
Risk tiers and overdue insight focus your collections where they matter most.
What this actually means in practice
Two engines read the same party history. The customer score reads every non-cancelled sales invoice raised against a party, credit notes issued for returns, receipts including those posted on account rather than against a specific bill, and the due date on each invoice so settlement delay can be measured. The churn view reads a narrower slice: order count, first and last invoice dates, and net lifetime revenue after returns. Neither engine reads anything you have not actually billed.
Value is your revenue from the party measured against your largest customer. Loyalty blends order frequency with recency. Payment behaviour starts from average settlement delay against the due date and is reduced by the overdue share of that party's balance. The three combine at forty, thirty and thirty per cent into a 0-100 score, a tier — VIP, Regular, Occasional, or Dormant beyond a hundred and eighty days — and a risk level. Churn is arithmetic, not prophecy: how far past a customer's own average gap the next order has drifted.
Collections works the list from the top down — who owes, who pays late, whose balance is mostly overdue — while sales works the other end, ringing the regular buyer whose ordering interval has quietly doubled. Revenue at risk puts a rupee figure beside that drift. The boundary is firm: the engine ranks and explains, it does not send a reminder or block a credit sale of its own accord. A customer with a single order has no interval to measure and is reported as such, never guessed at.
| Component | Read from | Turns into |
|---|---|---|
| Value (40%) | Net lifetime revenue against your largest customer | The value half of the score |
| Loyalty (30%) | Order count blended with days since the last order | Frequency and recency signal |
| Payment (30%) | Average settlement delay against invoice due dates | Cut further by the overdue share |
| Tier | Recency plus value relative to others | VIP, Regular, Occasional or Dormant |
| Risk level | The payment component alone | Low, medium or high |
| Churn figure | Days since last order over their own average gap | 0-1, banded active to churned |
| Revenue at risk | Annual run-rate multiplied by that figure | Rupees attached to the drift |
What’s included
- Lifetime value & AOV
- Payment-behaviour analysis
- Churn probability + revenue-at-risk
- Risk tiers (VIP / regular / dormant)
- Per-customer score
- Runs offline, every plan
What it does not do
- · It measures buying, not intent. A customer who stopped ordering because of a factory shutdown looks identical to one who left for a rival — the engine reports the gap, you supply the reason.
- · Counter sales booked without a party never attach to a customer, so a walk-in-heavy shop will find much of its revenue sitting outside the analysis.
- · Single-order customers have no interval and no settlement history; they are reported as such rather than handed a manufactured churn figure.
- · Deterministic customer intelligence is included from the ₹249 plan upward; asking questions of it conversationally is the ₹2,499 Enterprise tier.
Frequently asked
Is this included in Basic?
Yes — customer intelligence is deterministic Tier-A and ships in every plan.
How is churn predicted?
From purchase recency and frequency vs the customer’s own expected interval — transparent and offline, no LLM.
Where does the churn figure come from?
From the customer's own rhythm, not a model trained on other businesses. LekhaPro takes the average gap between their past orders and compares it with the days since the last one. The further past their normal interval they drift, the higher the figure climbs, until roughly three cycles late is treated as gone.
Does it work for a business with no credit sales?
Partly. Value, loyalty and the churn figure all work on cash sales provided the bill carries a party name. The payment component needs due dates and settlement to mean anything, so in an all-cash counter book it stays neutral and the score leans on value and buying frequency instead.
How is ai · deterministic handled differently here?
It rests on a score per customer, see churn before it happens and collect smarter. The distinction that matters is that these are structural rather than cosmetic — the behaviour is built into how records are posted, not layered on as a report you have to remember to run.
Does AI · deterministic work offline?
On the Desktop edition, yes — completely. lifetime value & AOV and payment-behaviour analysis run against a local database on your own machine, so the screens behave the same with the network unplugged as with it connected. AI and billing are the only two things that reach out. On the Cloud edition ai · deterministic runs in the browser and needs a connection.
Does AI · deterministic post to the real books?
Yes, and that is the whole reason it lives in this product rather than beside it. What a score per customer records lands in the same posted double-entry ledger the Trial Balance, P&L and GST returns are built from — so there is no second set of ai · deterministic numbers to reconcile against the first.
How does AI · deterministic affect my GST?
It feeds it directly rather than sitting alongside it. CGST/SGST and IGST are split from HSN or SAC and place of supply as each document is raised, so whatever ai · deterministic produces is already correct when GSTR-1, 3B and 2B reconciliation are drawn off the same books. The return matches the accounts because it was never a separate exercise.
Does AI · deterministic cost extra?
No. AI · deterministic is part of the accounting core rather than a paid add-on, so see churn before it happens is there on the entry tier exactly as it is on the highest one. What the tiers change is reach and depth — devices, multi-branch and multi-warehouse reporting, and whether the conversational AI copilots are switched on.
Can I try AI · deterministic before committing?
Yes — 14 days, no card. Test ai · deterministic against your own masters and your own transactions rather than sample data: the questions worth answering here are about your business's edge cases, and a demo dataset is built not to have any.
Chase the right customers first
Per-invoice ageing meets customer risk and churn — so your follow-ups go where the money and the relationship are most at stake. Draft a WhatsApp or email reminder in one tap.
Ageing & risk computed offline · reminders open WhatsApp/email pre-filled (one tap). Sample data.