Guide

How many customers asked for this feature, and what are they worth?

The question behind every roadmap decision. Here is how to answer it by hand, why the answer is usually wrong, and how to get it in one question.

The short answer: collect every request from calls, tickets, chats and notes, group requests that mean the same thing, match each one to an account, then count customers (not mentions) and add up what those accounts are worth. Evermuse does all five steps continuously, so the answer is one question away.

The manual way, step by step

Step 1

Pull every request

Search call recordings, support tickets, Slack channels, CRM notes and survey answers for the feature. Export what you find into one sheet with the date, the source and the exact words.

Step 2

Group what means the same thing

Customers rarely use your feature's name. Merge requests that describe the same need in different words, and split ones that only sound alike.

Step 3

Match each request to an account

Look up who said it and which account they belong to in the CRM. Note whether the account is a customer, a lost deal, an open deal or churned.

Step 4

Count customers, not mentions

One champion asking five times is one customer. Count unique accounts, then keep the mention count next to it.

Step 5

Add up what they are worth

Sum the revenue (or deal value) of those accounts, and split it by segment, plan or region so the number means something in the roadmap meeting.

Why the manual answer is usually wrong

Where it breaks	What happens
🗂️ Coverage	You only search the sources you remember, so quiet channels and old calls drop out.
💬 Wording	Keyword search misses requests phrased differently, so counts come out low.
⏱️ Recency	By the time the sheet is done, new calls have happened. The count is stale the day it is shared.
🔍 Trust	Nobody can check the number, so the loudest anecdote still wins the meeting.

What about asking ChatGPT or Claude?

A chat assistant can summarize the documents you paste in. It can't count across thousands of conversations it has never seen, and it will happily give a confident estimate. If the answer has no count and no links, it is a guess.

The Evermuse way

Ask: “How many customers asked for bulk export, and what are they worth?” Evermuse answers with:

The count

Unique customers and accounts, plus total mentions.

The accounts

Which accounts asked, with their status and value from the CRM.

The quotes

Every request, each linked to the call, ticket or chat it came from.

The trend

Whether the need is growing, and what changed this week.

Ask it in the Evermuse workspace, or from Claude, ChatGPT or Cursor through MCP. Product, sales, CS and leadership get the same number.

FAQ

Can I include churned or lost accounts?

Yes. Counting by account means you can split requests by current customers, churned accounts and open deals. Churned accounts that asked for the same thing are often the strongest signal.

How are duplicates handled?

The same customer asking three times in three channels counts as one customer and three mentions. You see both numbers, and every mention links to its source.

Which CRMs does it work with?

Evermuse links requests to accounts from your CRM so you can see what a need is worth. Attio and Close are listed on our integrations page. Ask us about others.

What if requests are phrased differently?

That is the main reason spreadsheets fail. "Export to CSV", "download my data" and "we need this in Excel" are one need. Evermuse groups them by meaning, and you can review and adjust the grouping.

How long does the manual way take?

For a single feature, usually a day or more of searching call notes and tickets. For a whole roadmap, most teams give up and rely on the loudest voices.

Ask how many customers asked. Get a number.

Connect your calls, tickets and chats. Evermuse counts every request, ties it to the account, and links every answer to its source.