I once watched a VP of sales open a dashboard, stare at it for maybe four seconds, then turn to his ops lead and ask for the one number that wasn’t on the screen. That is the failure mode of most sales dashboards. They show whatever was easy to pull out of the CRM, not the number the person actually has to act on.
A good sales dashboard is boring in the right way. It answers one question, fits on a screen, and tells the reader where to push before the meeting starts. What follows is a build guide, not a gallery. You leave with four formulas, real thresholds, and a layout you can copy on Monday.
What a sales dashboard tracks
A sales dashboard tracks the health of the pipeline and the efficiency of turning it into closed revenue. That is the whole job. If a tile does not help someone decide who to coach, which deals to save, or whether the quarter is real, it is taking up space that a working number could use.
Rank the tiles by who is reading them, because the audience determines the metric selection. A rep wants their own quota attainment and the deals slipping this week. A manager wants win rate and cycle length broken out by rep. A founder wants two things, coverage and forecast, and gets suspicious when you hand them twenty.
The core set is short:
- Quota attainment, actual revenue measured against the target for the period
- Pipeline coverage, open pipeline divided by the quota still to close
- Win rate, the share of opportunities that finish as closed won
- Average deal size, revenue per deal won
- Sales cycle length, the days a deal spends from open to close
- Pipeline velocity, how much revenue the pipeline generates per day
- Forecast accuracy, last period’s prediction against what actually landed
Notice what is missing. Calls made. Emails sent. Meetings booked. Activity counts feel like productivity, and they belong in a coaching view where a manager works with one rep. On the headline dashboard they are noise dressed as signal. A rep can dial two hundred numbers and move nothing. If you want the argument for why raw counts mislead, that is the whole point of KPI vs metric, a metric quantifies an activity while a KPI measures progress toward a goal.
Here are the four formulas the rest of this page leans on, with a defensible starting number for each.
| Metric | Formula | Healthy number |
|---|---|---|
| Win rate | Deals won / (deals won + deals lost) | Around 19% to 21% across all B2B deals, near 29% on qualified only. Benchmark your own. |
| Average deal size | Revenue won / number of deals won | No universal target. It should drift up as you move upmarket. |
| Pipeline coverage | Open pipeline value / quota for the period | 3x to 5x. The 3x floor is just 1 divided by a 33% win rate. |
| Pipeline velocity | (open opps × win rate × avg deal size) / cycle length in days | No fixed target. Watch the trend line, not the single value. |
Pipeline coverage and velocity
Three times. That is the coverage ratio most sales leaders quote before they believe a quarter, and most of them cannot tell you where it came from. It comes from the win rate. Pipeline coverage divides open pipeline by the quota you still have to close, and if you win one deal in three, you need three dollars of open pipeline for every dollar of target just to break even on the math.
So the honest range is 3x to 5x, and it depends on your own conversion. High-velocity SMB teams can run near 2x to 3x because deals move fast and win rates hold. Enterprise teams carry 4x to 5x because long cycles let deals rot. A word of caution that took me years to internalize: 5x coverage full of stale, unworked opportunities is worse than 3x that moves. Coverage counts dollars. It says nothing about whether those dollars are alive.
That gap is exactly what pipeline velocity closes. Velocity folds four numbers into one and reports revenue per day, so a pile of dead deals stops flattering you. The formula reads: open opportunities times win rate times average deal size, all divided by the sales cycle length in days.
Open opportunities: 120
Win rate: 22% (0.22)
Average deal size: $9,500
Sales cycle length: 45 days
Pipeline velocity = (120 x 0.22 x 9,500) / 45
= 250,800 / 45
= $5,573 per dayNow the lever becomes obvious. Cut the cycle from 45 days to 38 and velocity jumps past $6,600 a day without a single new lead. That is why velocity earns a headline tile and a raw pipeline dollar figure does not. One tells you the machine is speeding up. The other just tells you the machine is big.
Win rate and cycle length
What counts as a win? The question sounds trivial until two teams report win rates that are four points apart and neither is lying. Win rate measures the conversion of opportunities, and the fight is always about the denominator. Won divided by won plus lost is the clean version. Won divided by every opportunity that ever entered the pipeline, including the 60 deals still open, produces a softer, prettier number that means nothing.
Pick the closed-deal denominator and pin it in the tile subtitle so nobody relitigates it in the meeting. Expect roughly one in five across all deals and closer to three in ten once you filter to qualified. Small deals under $50K convert higher, often 25% to 35%. Six-figure deals drag lower, sometimes 12% to 22%. A blended company win rate hides all of that, which is why the manager view splits it by segment and by rep.
Sales cycle length counts the days from open to close. Track the median, not the mean, because one nine-month whale distorts an average and the median does not flinch. Then watch the two numbers together. A win rate climbing while the cycle stretches is often a wash, or worse. It usually means the team is nursing marginal deals to a yes instead of killing them early and moving on.
One more that belongs near these: forecast accuracy compares the prediction to the result. Average teams land somewhere between 50% and 70% accurate. Genuinely good ones hit 80% to 95%. If your forecast and your actuals disagree by more than about ten points every quarter, the dashboard is not the problem, the pipeline hygiene feeding it is.
A layout that works
Put the status number top left and stop fighting the eye. People scan a screen in an F, so the top-left corner is the most expensive real estate you own. That spot goes to quota attainment for the period, one big number against target, colored by whether you are ahead or behind. Nothing else competes with it.
The rest falls into three bands. The wireframe:
- Top strip. Four KPI cards, left to right: quota attainment, pipeline coverage, win rate, pipeline velocity. Each shows the value, its target, and a small trend arrow. This strip alone answers “are we okay” in under five seconds.
- Middle. The story of how deals move. A sales funnel on the left showing drop-off by stage, a line chart on the right showing bookings or created pipeline over time. The funnel is where a manager spots the leaking stage.
- Bottom. The detail table. One row per rep or per open deal, sortable, with attainment, open pipeline, and last activity date. Detail belongs at the bottom right, where people look only after the headline sends them there.
Two rules keep it readable. Use a single accent color for “off target” and let everything else stay quiet, so the eye lands on the one tile that needs attention. And use the right chart for each job, a funnel for stage drop-off, a line for the trend, a plain table for the numbers people copy into an email. If you are unsure which shape fits which question, how to choose a chart type walks through it. The broader structure sits inside our guide to dashboard design, and the tile choices connect to dashboard widgets.
Sales dashboard mistakes
Every list of mistakes is easy to nod along to and hard to actually avoid, because each one felt reasonable the day it got built. These are the ones I see most.
- The leaderboard-as-dashboard. A ranked list of reps by revenue is a motivator, not a diagnostic. It tells you who is winning and nothing about why, or what to do about the rep in last place.
- Total pipeline as a hero number. A giant “$4.2M in pipeline” tile feels great and predicts almost nothing. Coverage and velocity say what that pile is worth.
- Counting open deals in win rate. Padding the denominator with unresolved opportunities inflates the number and hides the real conversion. It is a vanity metric wearing a KPI costume.
- Mean instead of median cycle length. One monster deal drags the average and you start “explaining” a cycle problem that does not exist.
- Stale data with no timestamp. A dashboard that quietly shows Tuesday’s numbers on Friday erodes trust fast, and trust is the only reason anyone opens it twice. Put the last-refresh time on the screen.
- Twenty tiles, no hierarchy. When everything is important, nothing is. Rank ruthlessly and cut.
Build the one that survives contact with a skeptical VP. Four numbers he trusts, a funnel that shows the leak, a table he can sort, and a refresh time he can see. Everything past that is decoration, and decoration is the thing he clicked away from in four seconds.