Notes for owners · CRM, automation and RevOps
How to get your records ready for a diagnostic, even if they live in a notebook
Export what you have, gather quotes and call logs, and turn a paper notebook into a usable sheet. A step-by-step way to prepare records without cleaning everything first.
The GullySales team · Updated 17 Oct 2026 · 6 min read
On this page
Gather six things for a diagnostic: an export of your CRM or enquiry sheet, your call logs, read access to your ad accounts, the quotations you sent, your invoice or sales register, and a sample of WhatsApp conversations. Do not clean them first. Send them as they are, with a few lines saying what is missing, so the analysis can work out how far the numbers can be trusted.
Start from the question, not the folder
Before collecting anything, write the one question you want answered, such as "why do enquiries from IndiaMART not turn into orders?" or "which salesperson's quotes go cold?" That sentence decides which records matter most. Collect the rest if easy, skip it if hard.
Choose a window next: the last six months is a good default for a business with steady enquiries, a full season for a seasonal one. Use the same dates across every record, or the comparisons will not line up. Include the first enquiry date that is still open, since an order from an April enquiry may only be invoiced in September.
The checklist, record by record
| Record | How to get it | Mark these problems |
|---|---|---|
| CRM or enquiry sheet | Export to Excel or CSV, with every column | Fields people skip; stages renamed over time; old test entries |
| Call logs | Phone system report, or itemised mobile bill for the sales numbers | Calls made from personal phones; numbers not saved with names |
| Ad accounts | Add the analyst as a read-only user. Do not share your password | Accounts held by an old agency; ads paused and restarted |
| Quotations | Folder of PDFs or the quote register, with dates and values | Quotes sent on WhatsApp only; revised quotes with the same number |
| Invoices or sales register | Accounts export by date, customer and value | Cash sales not invoiced; invoices raised in a different name |
| WhatsApp chats | Export ten to twenty recent chats, from enquiry to outcome | Chats on a personal number; voice notes with prices |
Write one line beside each: "kept up to date", "partly kept", or "not kept". That honest line is worth more than a tidy file.
If your records are in a notebook
A notebook is workable. The aim is to make one sheet, not to digitise years of history.
- Take the last sixty days of entries only. Older rows can wait.
- Make columns for date, name, phone, what was asked, source, who handled it, outcome, quote value.
- Type one row per enquiry. If you cannot tell the source, write "unknown"; do not guess.
- Where the notebook has quotes or orders but no enquiry, still add a row, and mark source "repeat" or "unknown".
- Photograph the pages as well, so the source stays with you.
- Mark any row you are unsure about with a question mark in a spare column. Doubts are information.
Type in short sittings, not one long evening, so the rows stay careful.
For example: a building materials trader in Vijayawada
Take an invented trader in Vijayawada selling cement, steel and plumbing fittings to contractors, with two counter staff and a notebook for enquiries. Orders are written in a delivery book, and invoices are raised in accounting software.
The owner's first attempt at preparing records is to retype everything into one tidy sheet, correcting what looked wrong: merging two spellings of a contractor's name, filling gaps with guesses. It takes two weekends and erases the evidence. The analysis needs to see that the same contractor had been recorded three times, with three different quotes, and that two of the quotes had no entry in the delivery book.
The better version is quicker, though not painless. The owner photographs the last sixty days of notebook pages, types them as they stand into a sheet (a day's work, with a handful of rows too faint to read and marked as such), exports the invoice list from the software, and puts "notebook, not always filled in" beside the first and "complete" beside the second. The mismatch between the two lists then becomes the first finding, not a mistake hidden.
What not to do
- Do not delete duplicates, dead leads or embarrassing rows. They show how records are really kept.
- Do not fill blanks from memory, since an invented source is worse than "unknown".
- Do not send passwords. Add people as read-only users instead.
- Do not wait for perfect. A diagnostic can say what the records can and cannot support; it cannot do that if you send nothing.
When to clean first anyway
There is one case: if your CRM holds thousands of unusable rows, such as ten-year-old imports with no dates, a quick archive of the dead part helps, provided you keep a copy untouched. For that kind of tidy-up, see how to clean and organise CRM data and our CRM data cleansing, deduplication and governance service.
Sending it across
Put the files in one shared folder with a cover note. This layout is enough:
Question: why do IndiaMART enquiries not turn into orders?
Window: 1 April to 30 September
CRM sheet: partly kept, source filled for about half the rows
Call logs: itemised bill for two sales numbers; one salesperson uses a personal phone
Quotes: folder of PDFs, some sent only on WhatsApp
Invoices: complete, from the accounts export
WhatsApp: twelve chats exported, ten with outcomes
Not available: anything before April
Start tonight with one thing: photograph the last sixty days of the notebook, or export the CRM, and write the "kept up to date" line beside it. The paid analysis is described on the diagnostic page, and the FAQ has our answers on what is needed to begin. When your list is ready, tell us what you need.
