Services · 04
Data & analyticsEvery report from one place, and it matches the godown
One truthful picture of stock, sales and money, checked against reality before anyone relies on it.
One place every number comes from.
When two reports disagree, the meeting stops. The sales figure from the billing system does not match the one in the spreadsheet, the stock report has never once matched the godown, and month-end drags on because someone copies between systems by hand. The decision waits, or gets made on whichever number looks better.
The fix is one place every number comes from, in your own accounts, with the rule for each number written down and checked against reality before anyone relies on it. Tally stays the book of record. Excel stays for the analysis. What stops is the copying, and with it the disagreement.
- Team
- The named owners, for the whole engagement
- Runs in
- Your own accounts, from day one
- Pricing
- Fixed price, billed by milestone; a fixed itemised quote follows the first call
- Reports that disagree depending on who pulled them
- Month-end that takes a week of copying between systems
- Wanting automation or AI, but not trusting the data under it
- One place every number comes from, in your own accounts
- Proof, run in parallel, that the new numbers match the old reality
- A monthly running cost you saw in writing before the build
- The owner’s morning page: stock, sales, money
Four gates. You accept each one before the next starts.
The same four as every engagement, as they look for this work.
Diagnose
We take the three reports you argue about most and trace each number back to where it was born. The findings document shows where the copies diverge and which numbers have no single source at all, ranked by how much money rides on them.
Prove
One report, rebuilt from the one place, run side by side with the old one until they agree or the difference is explained and signed off. The stock figure is checked against a physical count in one godown. The running cost is written down.
Build
The rest of the reports, fortnightly, each one reconciled against the old before it replaces it. The owner’s morning page appears: stock, sales, money, from the same source as every other report. The rule for each number lives in a document your team owns.
Hand over
Your team runs it for two weeks while we are still on call. Each report has a written definition, the monthly check against the godown is a procedure rather than a favour, and the review says which numbers you can now trust and which still need a person.
Eleven stores, three online channels, one set of numbers each morning.
A fashion brand with eleven stores in four cities and three online channels. Month-end took a week of reconciling three reports, and they still disagreed.
The month-end close
from the last day of trading to a set of numbers the owner will act on

- Report lines that disagreed between channelssame day, same product, two sources, two answers
- Lines out of stock at one store while another held thema size a customer asked for and did not get, while it sat unsold elsewhere
What owners ask before they call.
Why do our reports disagree depending on who pulled them?
Because each report starts from a different copy of the truth. One is pulled from the billing system on the day, one from a spreadsheet updated on Fridays, one from an export someone cleaned by hand. Each copy is a little stale and a little edited, and the differences compound. The cure is not a better spreadsheet; it is one place every number comes from, with the rule for each number written down once, so two people asking the same question get the same answer. Why do our reports disagree? takes the four causes in turn, starting with the one nobody looks for.
Our stock report never matches the godown. Is that a software problem?
Partly, and the software part is the easier half. Returns booked late, transfers between godowns that exist on paper for a day, damaged stock nobody records, and a count taken at a different moment from the report all put a gap between the system and the shelf. We measure that gap first, godown by godown, then close it from both sides: the system records the events it was missing, and counting becomes a short routine with a number attached rather than a yearly ordeal.
Our stores, the marketplaces and the website each report separately. Can they be one number?
Yes, and it is the most common version of this job. Each channel keeps selling the way it sells; what changes is that every sale lands in one place in the same shape, so a product is one product whether it went out of a shop, a marketplace or your own site. Returns and cancellations are counted the same way everywhere, which is usually where the old numbers diverged most. The D2C fashion brand had eleven stores, three online channels and a month-end that took a week and still disagreed.
Do we need a data warehouse for reliable reports?
At the size of most businesses we work with, no, and we will say so. What is needed is one place the numbers are collected, in your own account, that every report reads from, and a written rule for each number. That can be a modest database and a nightly job. The heavy machinery exists for businesses with far more data and far more people asking questions of it; if you are one, we say that too, and the running cost is in writing before the build.
Can the numbers stay in Tally and Excel?
Yes. Tally stays the book of record for the accounts, and Excel stays what it is good at: looking at a number from a new angle. What changes is where Excel gets the numbers from. Instead of a paste from three exports, it reads from the one place, and so does every other report. Nobody copies, so nothing drifts, and the chartered accountant sees the same figures the owner sees.
How do we know the new numbers are right?
By running the old report and the new one side by side, on the same period, until they agree or the difference is explained and signed off by the person who owns that number. That parallel run is a gate: nothing replaces a report you already trust until it has earned it. For stock, the check is physical: a count in the godown against the system on the same day, repeated monthly as a routine, with the difference tracked as a number of its own.
What does a reporting system cost to run each month?
We do not publish prices for the build, and there is no hourly billing. The running cost is a different thing: the servers and services the system uses at your volumes, every month, whoever runs it. That figure is worked out and written down before the build, reviewed with you, and checked against the real bill after go-live. The point is that you saw it first. A cost model is a deliverable explains why.
For your technical adviser
Lakehouse-first warehouses, pipelines and governance. Vendor-neutral and cost-modelled before build: lineage map, parallel-run reconciliation, FinOps alerts, decommission schedule.
- Lineage map and migration plan with costs
- Iceberg or Delta lakehouse with IAM and governance
- Parallel-run reconciliation reports
- FinOps alerts and a decommission schedule
Stack Apache Iceberg · DuckDB · dbt · Spark · PostgreSQL · Airflow · Terraform
A cost model is a deliverable
Why the monthly running cost is written down before the first report is built.
Write a paragraph. A founder replies within one working day.
A 30-minute call with both founders follows. No deck and no proposal template, just a first read of whether we are the right shape for the problem.
Or by phone or WhatsApp: +91 94330 54299 · WhatsApp (opens in new tab)