Most SaaS marketing teams do not have a dashboard problem. They have a tile problem. When five people are responsible for pipeline, content, paid acquisition, and retention, a thirty-tile kpi dashboard becomes a monument to good intentions. Nobody checks it, nobody trusts it, and the metrics that actually demand action this week get buried under vanity charts.

The discipline is knowing what to cut. Dashboard value is inverse to tile count for a small team. A kpi dashboard should only show metrics someone can act on this week; everything else belongs in a drill-down. That single rule determines which tiles survive contact with a five-person marketing team and which ones get archived.

This guide walks through the sections ahead: why a small team should count tiles rather than metrics, the Act This Week test that decides every tile, and how review cadence becomes the real cut line. You will see the weekly tiles that earn their place, including trial or signup conversion rate and new paid conversions with MRR added, plus what moves to the drill-down. A worked dashboard example and guidance for keeping the dashboard alive with five people follow. The cut line in one sentence closes it out.

Why a Small Team Should Count Tiles, Not Metrics

My position, stated as an argument rather than a finding: for a five-person team, dashboard value runs inverse to tile count. No benchmark in this piece measures that relationship.

A tile is not a number. It is a commitment: one named reader, one owner, one recurring decision. Thirty tiles on a five-person team means twenty-five numbers with no reader.

The trust cost compounds. One metric nobody acts on teaches the team that the whole page is decoration, and the two tiles that matter stop getting read too.

Large teams can absorb a wide surface because analysts maintain and interpret it. A small SaaS marketing team has a generalist who also writes the launch email. That is the case for cutting hard, as argued in this seven-number dashboard guide.

Tooling will not save you: Funnelkeeper lets teams build dashboards and track funnels, attribution and SEO, but tile discipline is a human decision, not a product feature. It will happily render thirty tiles.

The Act This Week Test: One Rule Decides Every Tile

If tile discipline is a human decision, it needs a written rule. Here is the one this piece uses: a tile earns a spot only if a named person on the team can take a different action this week because the number moved.

Every candidate metric then has to survive three questions:

Who owns it?
What decision changes when it moves?
What would we do differently if it swung 20 percent either way?

Anything that cannot answer all three fails immediately. Total sessions, impressions, follower counts, arbitrary vanity scores and any number whose only possible response is "noted" are out. This is the same subtraction logic behind what to put on a SaaS marketing dashboard and what to leave off, applied as a test rather than a preference.

Two metrics pass and get tested in detail later: trial or signup conversion, and new paid conversions. Both respond to changes the team controls, such as pricing, onboarding or follow-up timing.

Write the rule on the dashboard itself, as a subtitle or header line. The next person who wants to add a tile then has to argue against the rule, not against taste.

Cadence Is the Real Cut Line

The act-this-week rule has a second edge: a number must also move in seven days without a new quarter starting. Weekly tiles are operational, monthly and quarterly tiles strategic. Mixing them in one fresh-looking grid makes old data read as signal.

CAC payback is the offender. My inference, not sourced: sales and marketing spend that closes new ARR was incurred earlier, and that timing mismatch is a plausible reason payback is reported too low. Lagged metrics belong in a monthly or quarterly review. Median was 16 months in 2025, down from 18, top quartile 6 months or less, bottom 24 or more (Benchmarkit, 198 companies, cited by Digital Applied; Aleph's 342-company sample echoes it). A 16-month number does not move between Monday and Friday. Our guide to CAC payback period explains why calculation and interpretation diverge.

Retention resolves over months. Median NRR was 102 percent in 2025. Verify GRR against the original Benchmarkit report before quoting: the source gives 84 percent down from 88, while its own text also says 88.

Lagged context can appear weekly only if clearly fenced off; otherwise the team debates a quarter-old payback number in Monday standup.

The Weekly Tiles That Earn Their Place

With cadence settled, the landing view can be small on purpose. Cap it at two or three weekly tiles, each with a named owner and a threshold that triggers a conversation rather than a shrug.

Before a tile goes live, insist on three definitions: the exact denominator, the data source, and the comparison window (last week or a rolling average). A tile without all three is a debate waiting to happen.

Two candidates pass the act-this-week rule, and both get their own sub-section below: trial or signup conversion, and new paid conversions.

Dedupe is the quiet killer. Count a trial start and a signup from the same person as one win, not two, or the page loses credibility with the team that has to trust it.

Source splits stay off the tile. Put them in drill-down, where funnelkeeper's funnel and attribution views handle the breakdown, including the affiliate traffic patterns covered in our guide to measuring affiliate marketing for SaaS.

The first of the two candidates is trial or signup conversion rate: the share of signups or trial starts in a period that become paying customers within a fixed window. For free plans, that window is six months, and it is the same six months every time.

Set thresholds against data, not aspiration. Median free-to-paid conversion was 8 percent within six months, while trials requiring a credit card converted at 30 percent (ChartMogul, January 2026, survey of 200 products). Card requirements change the denominator, so those are two funnels, never one. Our conversion optimization benchmarks show where the gap between average and top performers actually sits.

It passes the act-this-week test because onboarding emails, in-product activation prompts, pricing page copy and sales follow-up timing can all change within days.

On the tile: the number, change versus last week, a 13-week sparkline, and a target line drawn from your own eight-week median rather than an industry figure. Drill-down splits by plan, acquisition source, campaign, and whether the visitor arrived from an AI chatbot or a search engine. That last split matters, since 51 percent of B2B software buyers now start research in an AI chatbot more often than with Google (G2, March 2026, a survey figure, not a measured traffic share).

Tile 2: New Paid Conversions and MRR Added

Where conversion rate measures the funnel's health, this tile measures its output. Count the new paying customers added this week, next to the MRR they bring, with the prior week and a four-week rolling average beside them. It passes the act-this-week test because it is the result every function feeds: onboarding, campaigns, pricing and sales follow-up all trace to this number without a translation step, and it sits inside the four structural levers that actually move ARR.

Noise is the trade-off. A weekly count can double on a single enterprise deal, so let the four-week rolling average drive decisions and treat the raw count as secondary context.

Put the source and channel split in drill-down, using funnelkeeper's attribution view. The landing tile stays one count and one direction of travel.

Spend connects only in the monthly view. Sales and marketing ran at a median 35 percent of revenue in 2025 (Benchmarkit, cited via Digital Applied), a period ratio that tells you nothing useful on a Monday.

What Moves to the Drill-Down

With the two weekly conversion tiles settled, the rest of the dashboard has to justify itself one click down or leave entirely. Move CAC payback to a monthly or quarterly view, as the cadence section argued. Pair it with segmented targets labelled honestly: your own targets for your motion and deal size, not survey data. Retention belongs monthly too: median NRR was 102 percent, and usage-based-only companies reported 108 percent versus 98 percent for seat-based. Show that split in drill-down because it explains which retention lever exists. Channel, campaign, keyword and landing page detail sit one click down so the landing view stays a decision surface. Traffic totals, impressions and follower counts do not move down. They leave. This is the hardest cut. Defend it with: "These stay tracked in the channel report. The tile answers one question: does this change a decision this week? Impressions can rise while pipeline falls."

A Worked Dashboard Example for a Five-Person Team

With payback and retention moved off the landing view, the remaining page is deliberately sparse. A header line carries the act-this-week rule in one sentence, followed by the date the tiles were last pruned, so the page documents its own scope. The top row holds exactly three tiles, left to right: trial or signup conversion, new paid customers with MRR added, and one rolling four-week trend. Nothing else sits there, because the top row is what the team opens every Monday.

The second row is a single funnel visual from visitor to paid customer. Show numbers only at stages that map to a team responsibility, not at every micro-step; how to optimise conversion across your entire SaaS funnel explains why those gaps matter more than traffic totals.

The bottom strip holds three links labelled CAC payback, retention, and channel attribution. Each opens a monthly view or a drill-down rather than rendering a number on the landing page. Described rather than shown, that structure is the whole example: three weekly decisions, one shared funnel, and three deliberate exits.

Keeping the Dashboard Alive With Five People

The prune date in that header is only useful if someone acts on it. Here is the maintenance routine that keeps three tiles from quietly becoming thirty.

Assign an owner to each tile, not to the dashboard. A shared dashboard has no owner. Name one person per tile, so a metric cannot outlive the person who cared about it by two quarters without anyone noticing.
Run a 15-minute monthly prune. Any tile nobody acted on since the last review gets demoted to drill-down. The only discussion permitted is naming the last action it triggered, and if nobody can name one, the decision is already made.
Cap the landing view and apply one in, one out. Every addition requires a removal. Without that rule, tiles return through small, reasonable-sounding requests, five at a time.
Log every add, cut and demotion in the dashboard description. One line each, with a date. This ends the quarterly argument about the same tile.
Set alerts only on tiles that have thresholds. An alert with no defined threshold is noise, and alert fatigue breaks a small team's dashboard faster than missing data ever will.

Governance sounds dull until you skip it. Then the page becomes the thirty-tile dashboard you just spent a post cutting back.

The Cut Line in One Sentence

If the monthly prune is the maintenance habit, the cut line is the principle it protects: a tile earns its place only if a named person can take a different action this week because it moved. Everything lagged, including CAC payback and retention, belongs in a monthly view or a drill-down.

Three things you can do today:

List every candidate metric without filtering.
Run the three-question test on each one: who owns it, what decision changes, what would we do differently.
Set the cap at three landing tiles with one-in-one-out, as the maintenance routine enforces.

One last warning on cadence: dropping a 16-month payback figure into a weekly grid does not make it fresher. It makes the fresher numbers beside it easier to doubt.

If you want the funnel and attribution drill-downs behind those tiles built, funnelkeeper can help with that. Deciding which tiles survive stays a human call.

Conclusion

A five-person SaaS marketing team does not need more metrics. It needs fewer tiles that change decisions. Keep the test simple: a tile earns its place only when a named person can take a different action this week because it moved. Put lagged numbers like CAC payback and retention in a monthly view or drill-down, where cadence matches their meaning. Cap the landing view at three tiles, start with the two weekly conversion tiles, and use one-in-one-out from there. Then protect the dashboard with lightweight governance: log every add, cut, and demotion, and set alerts only where thresholds exist.

Run the three-question test on your candidate list, and if you want the funnel and attribution drill-downs behind the surviving tiles built, funnelkeeper can help with that.

KPI Dashboard for a Five-Person SaaS Marketing Team: What Deserves a Tile