Pipeline questions answered from the export you already pull.
Every CRM has a report builder, and every sales lead still ends up in a spreadsheet — because the question they actually have needs two objects, a date range the builder will not express, and a definition of "won" that someone changed last quarter.
The reason pipeline numbers drift is rarely bad data. It is that "win rate" means three things to four people, and whichever spreadsheet is open decides which one you get this month.
The metrics come pre-defined
Every formula, with the assumptions written down — because the assumptions are where these numbers go wrong.
| Metric | Definition | The assumption behind it |
|---|---|---|
| Win rate | won deals / closed deals | Counts deals, not value, and excludes still-open deals from both sides. |
| Won amount | value of deals closed won | Windowed on close date — a deal created earlier counts in the period it closed. |
| Open pipeline | value of deals not yet closed | As-of by nature, but the close-date window still applies, so a deal with no expected close date is not counted. |
Read the win-rate row again, because it is the one that causes arguments. “Of the deals we finished, how many did we win” and “how much of our pipeline converted” are different questions with different answers, and most spreadsheets quietly answer the first while the person reading it believes the second. Quiriz picks one, states it, and computes it the same way every month.
The vocabulary it also understands
Beyond the three governed metrics above, Quiriz knows the language of a pipeline and will work with it: weighted pipeline, bookings, open deal count, average deal size, average sales cycle in days, won and lost counts.
Worth being precise about the difference. The three metrics in the table are compiled definitions — same question, same number, every time. The vocabulary above shapes how a question is understood but is not a compiled guarantee, so treat those figures the way you would treat a good analyst's first pass: right in shape, worth a check before it goes in a board pack.
The exports it reads
- Deals — from HubSpot, Salesforce, Pipedrive or Zoho. Quiriz recognises each one's column names for the same field, so
dealstage,hs_object_idandclosedateneed no renaming. - Owners or users — so results break down by the person or region that owns the deal.
- Anything else at deal grain — a territory mapping, a product list, an ARR spreadsheet the CRM does not hold.
An export, not an integration. No CRM admin has to approve anything, no OAuth screen, no app in your instance. Sales ops pulls the report they already pull. The trade is that figures are as current as the last export, which fits a weekly pipeline review rather than a live leaderboard.
What it breaks down by
Governed answers break down by deal owner region, over any window you name. That is a short list on purpose: a breakdown is published here only once it is implemented end to end, because an answer that quietly groups by the wrong field is worse than no answer.
Other groupings — by stage, by source, by rep — are answered from your own export's columns rather than from a compiled definition, and are labelled as such in the answer.
Why the number holds still
The failure mode with a chat assistant is not a wrong sum. It is that it re-decides what "closed" means each time you ask, so your Q2 win rate moves between the Tuesday you asked and the Thursday you presented. Quiriz routes the question to a stored definition that compiles to SQL, and every answer states whether it came from those definitions or was answered freehand — so a number in a board pack can be traced back to the rule that produced it.
Pricing
You pay for workspaces and the people who build them — not per seat asking questions. Every rep can check their own pipeline without another licence. See pricing →
Your files stay yours, and we do not train on your data. Privacy policy →