Sales Performance Dashboard: The Metrics That Belong On It

A sales performance dashboard is the one screen a revenue team uses to decide what to do next. That is the whole job. Not to display the quarter, not to reassure the board, not to prove the team was busy. To put the next decision in front of the person who has to make it, with the number that decides it sitting right there.

Almost no dashboard I am handed actually does that. I have built revenue and forecast models for B2B SaaS companies for two decades, and the dashboard is usually the most crowded artifact in the building and the least useful. Thirty tiles, all green, all true, and not one of them tells the VP of Sales what to change on Monday. The reason is a category error baked in at the design stage: the dashboard was built to answer "how are we doing," when the only question worth a screen this expensive is "what do I do next." Those are different questions, and they produce different dashboards. Here is the case for the second one.

Status Dashboards Versus Decision Dashboards

A status dashboard reports the current state of the business. Revenue to date, attainment to plan, activity logged, leads generated. Everything on it is accurate, and almost none of it is actionable, because state is not the same as direction. Knowing you are at 60 percent of plan in week six tells you where you stand. It does not tell you whether to push, who to push, or what is breaking.

A decision dashboard is built backward from the choices a revenue leader makes on a recurring cadence. Is the target reachable from current pipeline? Where are deals stalling, and in which segment? Is the pipeline we are building the kind we close? Each is a real decision with a real owner, and each has exactly one metric that decides it. Build the board out of those, and a tile turning red is not information to absorb. It is a decision arriving with the action already attached.

Here is the contrarian part, and I will defend it for the rest of this guide. A sales performance dashboard should not try to show you how the business is doing. The instant you optimize a screen for completeness of status, you fill it with lagging, comfortable, after-the-fact numbers, because those summarize state cleanly. The dashboard that helps you hit the number looks underbuilt by status standards. It keeps mostly the early-warning set that no one screenshots for the board because it does not flatter anyone. Underbuilt is the goal, not a flaw.

The Decision-Row Layout

The framework I use to rebuild a bloated board is the Decision-Row Layout, and the rule is one line: every row on the dashboard is a decision, not a topic.

A topic-organized dashboard groups tiles by subject. A pipeline section, an activity section, a revenue section. It reads like a filing cabinet, and it steers like one. A decision-organized dashboard gives each recurring decision its own row, and each row holds a matched pair: the leading indicator that warns you, sitting left, and the lagging outcome that confirms whether your response worked, sitting right. You read left to right, from the signal that buys you time to the result that grades your move.

Decision the row drives	Leading tile (warns)	Lagging tile (confirms)	Review cadence
Is the target reachable?	Pipeline coverage ratio	Closed revenue to plan	Weekly leading, monthly lagging
Where are deals stalling?	Stage conversion by segment	Win rate by segment	Weekly
Are deals getting harder?	Sales cycle length by segment	Forecast accuracy	Weekly leading, quarterly lagging
Is capacity matched to pipeline?	Pipeline velocity	Quota attainment distribution	Weekly leading, monthly lagging

Four rows. Eight tiles. That is a complete operating dashboard for most B2B SaaS teams, and the discipline of the layout is that nothing earns a tile without a decision to its left explaining why anyone would look at it. The cadence column is part of the design, not an afterthought, because a leading tile inspected monthly is wasted and a lagging tile inspected weekly just manufactures anxiety. For the underlying definitions behind each of these, the sales KPIs guide sorts which metrics predict revenue and which only report it, and the 22 sales operations metrics reference covers the wider set you draw from.

The reason the pairing matters more than the individual tiles: a leading indicator with no lagging partner is a hunch, and a lagging indicator with no leading partner is a postmortem. Coverage without closed revenue beside it never tells you whether healthy pipeline became healthy bookings. Closed revenue without coverage above it is the result with no warning attached. Put them in a row and the dashboard teaches you, week over week, which of your early signals actually predict the late ones for your business. That feedback loop is the entire value of a dashboard, and a topic layout hides it by filing the warning and the result in separate sections.

The Vanity Tile Trap

The fastest way a decision dashboard rots back into a status one is the vanity tile. These are the numbers that look like progress, trend reliably upward, and change nothing about what anyone does. Total calls logged. Emails sent. Raw lead volume. Demos booked, counted on its own with no conversion rate beside it.

The test for a vanity tile is the same one I apply to any metric. If this number moved against you next week, what would you do differently? For a demos-booked tile sitting alone, the honest answer is nothing. You would note it and move on. A tile that cannot fail in a way that changes your behavior is decoration, and decoration on a decision surface is not neutral. It is camouflage. Every vanity tile you add makes the one tile quietly turning red harder to spot, because the eye has more green to wade through before it gets there.

Vanity tiles are seductive precisely because they are inputs someone started treating as outcomes. Demos booked feels like a result. It is an activity. Lead volume feels like pipeline. It is raw material that may or may not become pipeline. The moment an input gets promoted to an outcome, it stops prompting the question that would make it useful, which is "did this convert," and starts generating the satisfied nod that makes it useless. The fix is not to delete the activity. It is to pair it with the conversion one step downstream, so demos booked rides next to demo-to-opportunity, and the row only looks good when the activity actually produced something. The deeper logic of leading versus lagging selection lives in the sales performance metrics guide, and the productivity inputs most often mistaken for outcomes are covered in sales productivity metrics.

A Worked Example: Marsfield Rebuilds the Board

Numbers below are illustrative, chosen to show the mechanism, not a benchmark.

Marsfield is a mid-market B2B SaaS company running a status dashboard with thirty-one tiles. In week seven of the quarter, the board is reassuring. Closed revenue is pacing to plan. Activity is up double digits. Lead volume hit a quarterly high. Demos booked just set a record, and that tile is the one the team puts in the QBR deck.

Three things are wrong underneath, and the status layout files every one where no one is looking.

First, enterprise stage conversion from proposal to closed-won has slipped from roughly one in three toward one in four. It sits in the pipeline section, blended across all segments, where strong SMB conversion holds the average flat and the enterprise leak vanishes into the mean.

Second, the enterprise sales cycle has stretched by a few weeks. It lives in a different section, two scrolls away from the conversion tile, so no one connects the drift to the leak even though they are the same story.

Third, the record demos-booked number went up because qualification got looser. The proudest tile on the board is the symptom. More demos, worse fit, which is what produced the conversion slip in the first place. On a status dashboard that causal chain is invisible, because cause and effect are shelved in separate rooms.

Marsfield finds out at quarter-end, when closed revenue misses by a margin nobody saw coming, because closed revenue was the only honest tile on the board and it reports last.

Now rebuild Marsfield's board as a Decision-Row Layout. The "where are deals stalling" row puts enterprise stage conversion on the left, reviewed weekly, with win rate by segment beside it. The proposal leak surfaces in week three, not at close, while there are still nine weeks to work it. The "are deals getting harder" row puts cycle length by segment right there too, so the drift reads as drift the moment it starts. And demos booked is not a standalone tile at all. It rides paired with demo-to-opportunity conversion, which flashes the instant qualification loosens. Same company, same quarter, same deals. One board buried three signals under twenty-eight that did not matter. The other had four rows and caught all three in time to act. The difference was never the data. It was whether the layout was built for status or for decisions.

Designing the Board for Decisions

Once the Decision-Row Layout is in place, a few habits keep the board from re-bloating into a status screen.

Start from the decision list, not the metric list. Write down the choices the revenue leader actually makes on a recurring cadence, then attach exactly one leading and one lagging tile to each. If a metric you love has no decision to sit under, it does not go on the dashboard. It goes in a report someone pulls when they have a specific question. Segment before you average, every time. The Marsfield leak existed only because the board blended segments. Win rate, cycle, and conversion behave differently across SMB, commercial, and enterprise, and a blended tile is comfortable and blind. The segmented one points at the fix. This single rule catches more buried problems than any other on the list, which is why it shows up across the sales process optimization work too. Tie every tile to one source and one definition. The fastest way to kill a dashboard is two numbers for the same metric. Win rate computed one way in the board deck and another in the team review means nobody trusts either, and a dashboard nobody trusts is just wallpaper. Pick the definition, write it down, pull it from one place. Re-audit the board quarterly. Dashboards bloat the way closets do, one reasonable-seeming tile at a time. Every quarter, walk each tile and ask what decision it drives and what you would do if it moved. The ones with no answer come off, and they will have crept back on, because the pull toward status reporting is constant.

A Dashboard Is a Decision Surface

If you keep one idea from this, keep the reframe. A sales performance dashboard is not a monitoring screen and it is not a report. It is a decision surface, and a decision surface is judged by exactly one thing: when a number moves, does the right action become obvious to the person looking at it? Every tile that does not pass that bar is borrowing attention from one that does.

That is why the better board almost always looks smaller than the one it replaced. You are not measuring less of the business. You are refusing to let the screen that drives your moves get cluttered with numbers that drive none of them. Build it as rows of decisions, pair every warning with its confirmation, segment before you average, and the dashboard stops being the thing you stare at after the quarter is lost and becomes the thing that warns you while there is still time to catch it. An early signal reads as noise on a static board because nothing connects it to the outcome it predicts, which is precisely the link ORM builds when it pins each leading tile to a live forecast.