Measurement
Part of Measure proposal software by revenue or by reply, the choice that changes your numbers
Nine steps to build a proposal software reporting dashboard
A nine step guide to building a proposal software reporting dashboard, covering metric choice, data checks, access rules and the reports sales leaders actually use.
What to take away
- Agree five to seven measures, no more, and write a one-sentence definition for each before anyone builds a chart.
- Start from the six questions your sales director asks each Monday, then turn each question into one panel.
- Check the source fields first. If more than one in ten quotes lacks a close date, fix the data before you visualise it.
- Name one owner, set the refresh cadence to match the meeting it serves, and log who exports what.
- Budget about two working days of one person's time for a first version, built from exports of tools you already pay for.
Ask the questions before you draw a chart
A dashboard answers questions. It is not a display cabinet for everything you can count. Write down the six questions your sales director asks every Monday: win rate by deal size, time from quote sent to first reply, value of proposals sitting unsigned. Each question becomes one panel.
Anything that answers no question is decoration and should be cut in the first draft.
Six Monday questions to panels
- Win rate by deal size
- Time from quote sent to first reply
- Value of proposals sitting unsigned
- Each question becomes one panel
- Cut anything that answers no question
Define five to seven measures in a table
Keep the set small enough to read in under a minute. Write each definition as one sentence covering the start point, the end point and who counts. The wider measurement and reporting guide explains why the choice of denominator changes your numbers.
Five to seven measures
- 5-7measures to keep readable in under a minute
- 5typical measures quotes issued, accepted, deal value, reply rate, days to close
- 1sentence definition per measure covering start, end and who counts
Define five to seven measures
| Measure | One-line definition | Source field | Owner | Refresh |
|---|---|---|---|---|
| Quotes issued | Quotes sent to a named buyer during the period | Proposal tool status | Sales ops | Daily |
| Win rate | Accepted quotes divided by quotes issued in the period | Proposal tool plus CRM | Sales ops | Weekly |
| Reply rate | Quotes with any buyer response within 14 days | CRM activity | Sales ops | Weekly |
| Average deal value | Total accepted value divided by accepted quotes | Finance ledger | Finance | Monthly |
| Days to close | Days from quote sent to acceptance | Proposal tool timestamps | Sales ops | Weekly |
Define win rate against quotes issued, not opportunities created. Otherwise two managers will quote two different win rates in the same meeting.
Map where the data lives
List every system that touches a quote. Most teams run a CRM such as HubSpot or Salesforce, a proposal tool such as PandaDoc, Proposify or Qwilr, an e-signature service such as DocuSign, and a ledger such as Xero or Sage. Each system holds part of the answer.
Note the field, the owner and the update frequency for every measure. Most tools export CSV or Excel, which is enough for a first build. Gaps usually sit between sales and finance, where a signed quote becomes an invoice. org/wiki/Configure,_price_and_quote) formalises that join, though a spreadsheet works to start.
If you are still choosing tools, the best tools data and sources note sets out what to compare.
Run the data quality check
Pull a month of records and count the blanks. If more than one in ten quotes has no close date, no chart will rescue the number. Agree a rule for incomplete records, such as excluding them from win rate while still counting them in pipeline. Store that rule beside the definitions.
The nine build steps in order
The nine build steps
- Write the six questions and five to seven measures in one shared document. Owner: sales lead. Half a day.
- Confirm each measure has a source field, an owner and a refresh frequency. Owner: sales operations. Two hours.
- Clean the current month of records and record your exclusion rules. Owner: sales operations. Half a day.
- Build one panel per question, largest number first, four charts per screen at most. Owner: sales operations. Half a day.
- Test the totals against a manual spreadsheet for two consecutive weeks. Owner: finance. One hour a week.
- Set access so account executives see their own book and managers see the team. Owner: sales operations. One hour.
- Schedule the refresh and print the cut-off time on the dashboard itself. Owner: sales operations. One hour.
- Circulate a one-page reading guide with the definitions and exclusions. Owner: sales lead. One hour.
- Review the measures each quarter and retire any that nobody has quoted in a meeting. Owner: sales lead. Ongoing.
That is roughly two working days of effort for the first version. Before you set targets, read the benchmark research notes on how other teams frame theirs.
Set refresh, access and ownership rules
Daily refresh suits pipeline reviews and weekly suits board reporting. State the cut-off on the page so nobody argues about a number that moved overnight.
Cut-off: data as at 06:00, Monday to Friday. Exports taken later will not match these panels.
Access should follow the same logic as your CRM. Sellers see their own quotes, managers see the team, finance sees accepted values. Where proposals carry client personal data, the ICO data protection guidance sets expectations for retention and export. Log who exported what.
Where quotes turn into invoices, reconcile accepted values with the treatment published in GOV.UK VAT rates before finance signs off the panel.
Common questions
How many metrics should a dashboard show?
Five to seven. More than that and readers scan rather than act. If a measure has not been mentioned in a meeting for a quarter, retire it.
Should the dashboard refresh daily or weekly?
Match the cadence of the meeting it serves. Daily for pipeline reviews, weekly for management reporting, and always print the cut-off time on the page.
Who should own the dashboard?
One named person, usually in sales operations or finance. Shared ownership means nobody fixes broken fields, and the numbers drift within weeks.
Can we build it without a data team?
Yes, if the measure set stays small and the sources stay few. Most quoting tools export to CSV or a spreadsheet, which is enough for a first version.
