un-burn reads the financial and operating data your business already produces and shows what changed, what looks unusual, what deserves your attention, and what the files could not show. Where something material appears, we go through it with you.
No bank login. Your files are read in your browser; only bank and card transactions leave your machine, and only if you buy the written report. Free to see what's flagged. Paid from ₹399/month.
The software reads a bank or card export, and a sales export where one exists. It states what changed, what moved enough to be worth a look, which recurring costs are worth reviewing, and what the files could not show.
A finding is not a conclusion. Spend that rose is not proof of a price increase, a recurring charge is not proof of waste, and a payout that does not match your sales is not proof of missing money. Where a finding is material, we go through it with you, and with whatever further data is needed to establish what is actually happening.
If the investigation establishes a problem we can credibly help with, that work is scoped separately. If it does not, we say so.
The monthly check-in continues underneath. Each one is compared against the last, so whatever follows the work shows up in the next report.
One month, worked through the way the software works through it. Each step is more precise than the one above, and the last two are the ones most products leave out.
Sales fell ₹18,400
₹259,000 last month, ₹240,600 this one. Down 7.1%.
Order count barely moved: 1,040 to 1,018, down 2.1%. Average order value fell 5.1%, from about ₹249 to ₹236.
Smaller orders account for about ₹13,100 of the drop; fewer orders for about ₹5,300.
The typical order hardly changed. The median went from ₹196 to ₹193, a 1.5% move against an average that fell 5.1%.
So “customers spent less” is the wrong reading. Something has happened at one end of the order book, not across all of it.
| Order size | Last month | This month | Revenue |
|---|---|---|---|
| below ₹196 | 520 | 519 | -₹4,200 |
| ₹196 to ₹283 | 260 | 262 | +₹300 |
| ₹283 to ₹455 | 156 | 161 | +₹800 |
| ₹455 and above | 104 | 76 | -₹15,300 |
Orders of ₹455 and above carry 83% of the movement while making up 29% of last month's revenue. There were 104 of them, then 76.
That is where the decline sits. It is not why it happened, and these files cannot say. Discounting fell over the same two months, so it cannot account for the fall. Nothing in an order export records why an order that used to be placed was not placed.
Observed ₹18,400. Where it sits: orders of ₹455 and above. Cause established: none.
The same two months at item level, with item name, quantity and unit price on each order, would test whether the change came from what customers ordered or from what those items cost.
An example in the format un-burn uses. The figures are invented and belong to no real business; the wording, the structure and the arithmetic are what the product produces from your own files.
| Date | Statement line | Amount | Finding |
|---|---|---|---|
| 03 Jun | JUDGE.ME LTD | 15.00 | duplicate function: Yotpo does the same job |
| 04 Jun | YOTPO INC | 79.00 | — |
| 05 Jun | KLAVIYO SOFTWARE | 184.00 | price creep +23% since March |
| 06 Jun | SHOPIFY* 118427490 | 79.00 | — |
| 11 Jun | SHIPSTATION | 29.99 | — |
| 14 Jun | GORGIAS INC | 60.00 | — |
| 18 Jun | LATER.COM | 25.00 | recurring since Jan, confirm anyone still uses it |
Your sales for the period if you upload a sales export, the costs and recurring spend worth reviewing, anything that moved enough to be worth a look, and what your files could not show. Where the two files disagree on what you were paid, it says so. From the second check-in on, it opens with what changed since the last one.
Works best where costs arrive named: software, platforms, gateways, card and transfer fees. If most of your money leaves as salaries and rent, there is less to check.
Your first upload is the baseline. Every check-in after it is compared against the last one, so the report opens with what went up, what started charging, what stopped, and which findings from last time are gone.
Recurring spend across every vendor is up ₹4,120 a month since your last check, at ₹68,400 a month.
Example of a monthly comparison, with sample figures in the format your second check-in uses. Meta Ads reads as "more spent" rather than a price increase because ad platforms bill on what you use, and a statement cannot tell a higher rate from a bigger budget.
What is compared across check-ins is your recurring vendor payments and the findings from your last report. Sales, when you upload a sales export, are compared between months inside that file rather than against your last check-in.
Founders, managing directors, and whoever actually reconciles the accounts. What matters is complexity rather than industry: money leaving through enough systems, vendors and platforms that a change can run for months before anyone puts the pieces side by side.
Your files are read in your browser. Bank and card transactions reach our server only when you buy the written report. A sales export is different: it is analysed in your browser and never uploaded, because it holds your customers' details and not only your own. We keep your email and a summary of each report, so the next one can show what changed. Findings are rule-based arithmetic on your own files, not financial advice.
₹399/month
or ₹3,199/year (a third less).
Self-serve, and running today. Upload a statement, see your flagged total free, and pay only for the written check-in. Run it as often as you have a fresh file; each one opens with what changed since the last. Most months should come back clean, and that is the check-in working.
Scoped
per engagement.
For a business where a finding is material enough to be worth establishing properly. We work through it directly with management, on the findings and on whatever data sits behind them, and anything that follows is scoped on its own. Start with an email and the diagnostic on a real file.
un-burn starts with what your own data can actually support, so a conversation about your costs begins with your statement lines rather than a general opinion about your industry. A signal is not treated as a conclusion, and where the data cannot answer a question, we say so. That is the whole discipline, and it is the reason the investigation step exists at all.
You can run the diagnostic on your own file in the next few minutes, and judge the work from that.