Notes for owners · CRM, automation and RevOps
What to do when your records are too thin to analyse
If your sales records are a phone and a diary, start writing four things down for every enquiry today. Here is the register layout, and what a diagnostic can and cannot say from thin data.
The GullySales team · Updated 17 Oct 2026 · 5 min read
On this page
If your records are too thin to analyse, do not wait for a better system. Start a one-page register today with four columns: the date an enquiry arrived, where it came from, what the buyer wanted, and what happened to it. One named person fills it in the same day. After a few months you will have something to read. Until then, an analyst can only say what is missing and what to start capturing. That is worth knowing, but it is not a finding.
What counts as too thin?
Be honest with yourself on three tests.
- Count. Can you list the enquiries of the last three months, with names, without asking anyone?
- Source. For each, do you know how they found you: search, Maps, IndiaMART, a referral, a walk-in?
- Outcome. For each, do you know whether it became an order, was lost, or is still open?
If you can pass one test, you have a start. If you can pass none, you are at the beginning, and that is a perfectly respectable place to be.
The four fields to write down from today
Keep it on a sheet that opens on every phone in the team. Resist adding columns for the first month.
Date in | Source | What they asked for | Where it stands
12 Mar | IndiaMART | 20 steel racks, Belagavi | Quote sent 13 Mar, call 18 Mar
12 Mar | Walk-in | Counter displays | Lost: bought elsewhere, said delivery
A few notes on each column. Date in is the date the enquiry arrived, not the date you entered it. Source is what the buyer says if you ask once, politely, at the start of the call. What they asked for is a few words, enough to tell products apart. Where it stands is the one that matters: a status, the next date, and, once it closes, the reason in the buyer's own words.
Why these four and not forty? Because they answer the questions that reveal most leaks: where enquiries come from, what they want, how fast you moved, and why you lost. A longer form gets skipped by the second week.
Rebuilding the past, cheaply
You can seed the register with history. Spend one afternoon going through the call log, WhatsApp chats and email for the last two or three months. Add one row for each real enquiry. Leave cells blank when you do not know, and do not guess. A blank is honest data. A guessed source quietly poisons every count that follows.
If the history sits in several people's phones, ask each to send the names and dates, not their opinions about them. For help consolidating, see our note on tidying and organising customer data.
What a diagnostic can still tell you from thin records
A diagnostic, such as the paid one we describe, reads your own CRM, call logs, ad accounts, quotes and sales records. If those are thin, the result is limited, and any analyst should say so before you pay for work that cannot be done.
| Thin record | What can still be said | What cannot |
|---|---|---|
| Call log only | How many calls came in, at what hours, how many went unanswered | Why anyone did or did not buy |
| Quotes in a folder | How many were sent, how detailed they were, how they compare | Which were chased and which were won |
| Sales records, no enquiries | Who bought, when, how much, and who repeated | The share of enquiries that never became orders |
| Ad account, no outcomes | Spend and clicks by campaign | Which clicks became orders |
The right-hand column is what the register fixes. Notice that most of it is about outcomes and sources, the two things that are cheapest to write down at the time and impossible to recover later.
For example: a stationery distributor in Nagpur
Take an invented distributor who supplies schools and offices around Nagpur, with three people taking orders on the phone and nothing written but invoices. The owner wants to know why bulk-tender enquiries seem to be shrinking.
There are invoices, which show who bought. There is no record of who asked and did not buy. The honest answer from that data was that the shrinking could not be measured, only felt.
So the owner starts the four-column sheet, kept by one clerk, and spends an afternoon on call logs to seed it. The first mistake is a "Where it stands" cell that said only "lost". The clerk now has to add the buyer's own words. After a stretch of recording, one pattern shows: most lost rows name the delivery date, not price, though there are only about thirty closed rows, so it is a lead, not a proof. Before the sheet, the whole team had said price. The owner moves the dispatch cut-off by half a day for schools that order before noon, and keeps recording to see whether the delivery rows fall.
Reading the register after a few months
Do not read it for patterns too early. Rows are too few, and one big tender will skew everything. Wait until you have roughly a few dozen closed rows, then count four things: rows by source, rows by outcome, the most common reason given for a loss, and how many rows still say "open" after your normal sales cycle. A pile of stale open rows is itself a finding. It may mean nobody is closing them out.
When not to start with a register
Skip the sheet if you already have a working CRM that nobody uses. The job there is cleaning and adoption, not starting again. The post on whether you need a CRM or a spreadsheet covers when software earns its place.
Start the sheet today
Pick the person. Put the sheet where they will open it daily. Read the last few rows together at a fixed time, and ask for the reason on every lost row. Once the habit holds, you can add columns, or move to a CRM; CRM implementation and customisation is the service for that step. The paid diagnostic can read what you have once there is something to read, and the FAQ covers how we start. When you want a hand setting the register up, tell us what you need.
