The short version
Most sales dashboards fail by being complete. Forty widgets, no hierarchy, and a Monday meeting that devolves into arguing about definitions. I build GTM dashboards around ten numbers in three groups: input health, conversion, and outcome. Every number has one owner and one decision attached to it.
Input health tells you whether the system is fed: target accounts with mapped stakeholders, data completeness, emails bounced, domain complaint rate. Conversion tells you where the process leaks: reply rate, positive-reply rate, meeting-held rate, meeting-to-opportunity rate. Outcome tells you what it produced: qualified pipeline created and pipeline coverage versus target.
Writing the definitions before the dashboard
Arguments in Monday meetings are almost always definition arguments in disguise. Before building a single chart, write each metric as one sentence with a numerator, a denominator and a time window. Positive-reply rate is replies classified as interested divided by delivered first-touch emails in the same send cohort, measured at day 14. Put that sentence under the number on the page.
I also add a data-source line and a refresh time. When someone asks why meetings-held differs from the CRM, the answer should be visible without a meeting: this one counts calendar events marked completed, that one counts activities logged by reps. Two metrics that look similar but come from different sources should either be merged or labelled loudly.
Pairing each number with a threshold and an action
A number without a trigger is trivia. Next to each of the ten, write the amber and red thresholds and what happens when they trip. For example, bounce rate above 2 percent pauses the sending domain and sends the list back for re-verification. Meeting-held rate below 70 percent triggers a review of confirmation and reminder steps. Pipeline coverage below 3x triggers a prospecting sprint.
Choose thresholds from your own trailing quarter, not from a blog. Take the median of the last 12 weeks as the baseline and set amber at roughly a quarter below it. That keeps alerts meaningful. If everything is red all the time, people stop looking, and if nothing is ever red, the thresholds are too loose to catch a problem.
Building it in Power BI or whatever you have
The tool matters less than the layout. Put outcome numbers top left, because that is where eyes land, conversion in the middle row, and input health along the bottom as the diagnostic strip. Show each metric as a value, a four-week trend line and a comparison to target. Avoid pie charts and gauges; they hide the trend, which is the only thing you actually need.
Treat the page as a product with a changelog. Keep a small note at the foot recording the last definition change and why. Hold a quarterly pruning session: any widget with no views in 30 days is deleted. A dashboard that only grows will return to forty widgets within a year, and you will be back where you started.
Worked example (illustrative)
Hypothetical: a team’s Monday meeting used a 38-widget report and ran 60 minutes. They rebuild it into ten numbers with owners and thresholds.
What changed: The gain came from attaching an owner and an action threshold to each number, not from the visual redesign. Faster detection of bounce problems is the clearest sign the page now drives decisions.
Do it: the steps
Tick them off as you go. Your progress stays on this device.
Traps to avoid
- ×
Tracking activity (emails sent, calls made) as a headline number. Activity belongs in the input group, not the outcome group.
- ×
Mixing cohorts and calendar weeks. Reply rate calculated against emails sent this week, when replies arrive over ten days, will look artificially low every Monday.
- ×
Showing averages only. A median meeting-to-opportunity rate hides the fact that one rep or one campaign carries everything, so add a split by owner or segment on drill-down.
What to track
- Positive-reply rate
- positive replies divided by delivered first-touch emails per send cohort; a common planning range is 1 to 4 percent, varying widely by segment.
- Meeting-held rate
- meetings that took place divided by meetings booked; teams often aim for 70 to 85 percent.
- Pipeline coverage
- open qualified pipeline divided by remaining target for the period; 3 to 4x is a typical planning range.
Benchmarks are typical planning ranges, not guarantees. Calibrate to your own data.
Questions people ask
Which ten metrics should be on the page?
A workable set is: mapped-account coverage, data completeness, bounce rate, spam complaint rate, reply rate, positive-reply rate, meeting-held rate, meeting-to-opportunity rate, qualified pipeline created and pipeline coverage. Swap one or two for your motion, for example inbound speed to lead, but keep the three groups of input, conversion and outcome.
How often should the page be refreshed and reviewed?
Refresh inputs and conversion daily if the pipeline allows, and review weekly. Outcome metrics such as pipeline created can be weekly too. Do a deeper review monthly for trends and quarterly for definitions. Real-time dashboards tempt people to react to noise, so a stable weekly cadence usually produces better decisions.