A PLG metrics dashboard tracks eight numbers — traffic, signups, activation, retention, PQL rate, revenue, expansion, and referral — that together prove whether your product, not a sales team, is converting and growing customers. Each number should map to one specific moment on the user's journey, not sit as an abstract funnel-stage average disconnected from what actually happened.
Quick answer: The 8 PLG metrics are traffic, signup rate, activation rate, retention, PQL rate, revenue (free-to-paid conversion), expansion (NRR), and referral (K-factor). Track them in journey order — first visit, signup form, aha moment, feature-gate hit, upgrade click — so a dip always points to one fixable step, not a vague "growth is down."
The Trap: Why Most PLG Metrics Dashboards Are a Wall Nobody Trusts
A metrics dashboard fails the moment it becomes a list of everything measurable instead of a sequence tied to the user's actual path. Most PLG dashboards fail exactly this way — twenty tiles, no order, no owner — which is why a metric moving rarely triggers a specific fix, just a Slack thread asking whether anyone knows why.
Picture the Monday morning after a board meeting: activation rate dipped four points last week, traffic is flat, signups are up, and nobody can say why. Someone pulls up the dashboard — twenty tiles, six data sources — and nothing on the screen orders those tiles into a sequence.
This is the standard failure mode, and standard advice tends to make it worse. "Track more metrics" just adds tiles. "Pick a North Star metric" — a genuinely useful idea popularized by Amplitude's product analytics team — focuses attention on one number, but it can't diagnose where in the journey a problem started.
Generic frameworks fail for a specific reason: they treat metrics as funnel-stage outputs, when each one is really evidence from a specific moment — a form submitted, a limit hit, an invite sent. A number without its moment attached isn't diagnostic. It's decoration.
- Traffic tells you nothing about quality without knowing which channel it came from and what those visitors did next.
- Signups tell you nothing about product fit without knowing what fraction reached the aha moment.
- Referral counts tell you nothing about a real loop without knowing whether invited users activate at the same rate as people who signed up directly.
That correction clause — the "without knowing" part — is a moment. Reframing every metric around its moment is what the rest of this piece works from.
The Mental-Model Shift: Treat Each Metric as a Journey Checkpoint, Not a Funnel Stage
The fix is to stop grouping PLG metrics by data source — analytics tool, billing system, CRM — and start grouping them by the exact moment on the user's journey each one measures. A customer journey map built as a stage-by-stage sequence with an emotion curve is the connective layer most dashboards never build.
Funnel thinking asks what percentage converted at a given stage. Journey thinking asks what happened, right here, at this exact moment, and how the user felt about it. The second question takes more work to answer, but it's the only one that tells you what to actually fix.
Here are the eight numbers, each pinned to its moment. Seven of them are probably familiar; the eighth is the one most dashboards quietly leave out, and it's the one that turns the other seven from a scoreboard into a sequence.
| # | Metric | The Journey Moment | What It Answers | Directional Benchmark |
|---|---|---|---|---|
| 1 | Traffic | Before the first click — discovery | Are the right people finding the product? | Meaningless alone; watch it paired with signup rate by channel |
| 2 | Signup rate | The signup form | Does the pitch match what the product delivers? | Often well under 10% for cold traffic; far higher for referred or high-intent visitors |
| 3 | Activation rate | The first "aha" action | Did onboarding get someone to real value? | Commonly cited in the 20-40% range for self-serve B2B SaaS |
| 4 | Retention (curve) | The return visit, weeks 2-8 | Did the value stick past the first session? | Watch whether the curve flattens, not just the week-1 number |
| 5 | PQL rate | The feature-gate or usage-limit hit | Is usage deep enough to predict payment? | Varies widely by product depth; the trend matters more than the level |
| 6 | Revenue (free-to-paid) | The upgrade decision | Did the paywall land on real, felt value? | Often single digits for freemium; meaningfully higher for trial-based motions |
| 7 | Expansion (NRR) | The seat or tier growth decision | Does the account grow after the first sale? | Best-in-class PLG companies frequently report NRR above 110% |
| 8 | Referral (K-factor) | The invite or share moment | Is the product recruiting its own users? | Above 1.0 is self-sustaining; most B2B products sit meaningfully below it |
Two things matter more than any single row here:
- The moments are in order. This is a sequence, not a category list, and each number is only interpretable relative to the one before it — a great activation rate off a terrible signup rate is a smaller win than the same rate off a healthy one.
- PQL rate is the hinge. It's the metric most generic growth-metrics roundups skip, and it's the one that connects the product side — activation, retention — to the revenue side — conversion, expansion. PQL scoring is what turns product usage into a number a growth or sales function can actually act on.
Kyle Poyar, who built much of OpenView Partners' PLG benchmarking work and now writes the Growth Unhinged newsletter, has pointed out that most PLG teams under-invest specifically in that connective layer — they'll instrument activation carefully and revenue carefully, and leave the metric in between as a spreadsheet formula nobody quite trusts.
Activation rate specifically depends on what happens in the first few minutes after signup. What needs to happen in the first ten minutes of PLG onboarding covers the design side of that number, since it's arguably the single highest-leverage moment on the whole list.
Why Retention Belongs in the Middle, Not the End
Retention often gets treated as a lagging report-card metric, checked monthly, long after the moment it describes has passed. That undersells it. Retention is really asking whether the activation moment produced a habit, or just a one-time transaction.
Growth advisors including Casey Winters and Brian Balfour of Reforge have both written about "smile curve" retention analysis — cohort curves that dip after signup, then flatten rather than continuing toward zero. A flattening curve, even at a modest level, says more about product-market fit than almost any other line on the dashboard, because it's the only metric on this list that requires the passage of real time to fake.
Build the Growth Dashboard: 5 Steps From 8 Numbers to a Story
Building a real PLG dashboard is less about new tracking and more about sequencing numbers you already have: anchor each metric to a moment, order left-to-right by journey stage, pair every raw number with its ratio, assign one owner per metric, and set a matched review cadence.
Five moves, in order, turn a wall of disconnected tiles into something a team can actually act on by Monday afternoon.
- Anchor every metric to one journey moment before building a single tile. Sit down with the actual product-led growth funnel and write, next to each of the 8 metrics, the literal screen, action, or event it corresponds to. If you can't name the moment, you don't have a metric yet — you have a query.
- Order the dashboard left-to-right in journey sequence, not alphabetically and not by whichever data source is easiest to pull from. A reader should scan left to right and follow the same path a real user takes: discovery, signup, activation, retention, PQL, revenue, expansion, referral.
- Pair every raw count with its ratio. A raw number — total signups, total PQLs, invites sent — tells you volume. A ratio — signup rate, PQL-to-paid rate, referred-user activation rate — tells you whether the step is actually working. Track both, side by side, never the raw number alone.
- Assign one owner per metric, not one owner per dashboard. A dashboard with a single owner becomes their personal report. A dashboard where activation belongs to onboarding, PQL belongs to growth, and expansion belongs to customer success gets checked by the people who can actually move each number.
- Set a review cadence that matches how fast each moment actually changes. Signup rate and activation can move week to week; retention curves and NRR need a longer lens, since reviewing them weekly mostly adds noise and invites overreacting to a blip.
A dashboard that can't answer "which moment caused this" in under a minute isn't finished. It's just been reformatted from a spreadsheet into tiles.
The Raw Number vs. the Ratio That Actually Matters
David Skok, the venture investor behind the widely read For Entrepreneurs SaaS metrics essays, has long argued that a raw growth number is close to meaningless without the ratio sitting behind it — the ratio is what shows whether growth is durable or just this month's bigger bucket. Every one of the 8 PLG metrics has a vanity-metric shadow version worth naming explicitly.
| Metric | The Vanity Version | Track This Instead | Why |
|---|---|---|---|
| Traffic | Total visitors | Visitor-to-signup rate, by channel | Raw traffic hides which channels send people who never convert |
| Signups | Total signups | Signup-to-activation rate | A signup spike without activation is often bad targeting, not growth |
| Activation | Activated users (count) | Activation rate + time-to-value | The count rises just because signups rise |
| Retention | Logged-in users this week | Week-4/8 retention curve shape | One week's number hides whether the curve is flattening or still falling |
| PQL | Total PQLs generated | PQL-to-paid conversion rate | More PQLs mean nothing if growth or sales can't act on them fast enough |
| Revenue | New MRR this month | Free-to-paid conversion rate | MRR can rise from a big traffic month even as conversion quality falls |
| Expansion | Expansion revenue, in dollars | Net revenue retention (NRR) | A dollar figure hides whether expansion is outpacing churn |
| Referral | Invites sent | K-factor / referred-user activation rate | Invites sent measures a button click, not a working loop |
Common Mistakes That Quietly Turn PLG Metrics Into Vanity Numbers
The same handful of mistakes recur across most PLG dashboards, and all of them share a root cause: measuring a step instead of the moment the step represents. Fixing the moment usually fixes the metric; fixing the metric in isolation usually doesn't.
- Treating activation as binary. "Activated or not" hides that some users barely cross the line and others blow past it. JTBD framing helps here — the aha moment is the point someone completes the job they hired the product for, and depth of completion predicts retention better than a yes/no flag.
- Watching referral before revenue is proven. A K-factor above 1.0 amplifies whatever is already true about the product — and it's rare outside consumer network effects to begin with; Andrew Chen's The Cold Start Problem documents just how few B2B products sustain one. Amplifying a leaky paywall just produces more people who bounce off it faster.
- Blending logo retention with revenue retention. These answer different questions — did the account stay, versus did the account's spend grow — and collapsing them into one number hides expansion problems inside an otherwise healthy-looking retention chart.
- Generating PQLs nobody is set up to act on. A PQL rate climbing while self-serve and sales-assist motions haven't defined who engages a PQL, and when, just produces a growing list nobody works.
- Reviewing every metric on the same cadence. Weekly reviews of NRR mistake normal quarter-to-quarter noise for a trend. Monthly reviews of activation leave a broken onboarding step live for weeks longer than necessary.
The Prodinja Angle: Mapping the Moment Before You Trust the Metric
Every one of the 8 metrics above is only as useful as the map of moments it's pinned to — and building that map by hand is exactly the work most teams skip. Prodinja's Customer Journey tool plots that map directly: an explicit emotion curve marking a buyer's or user's high-anxiety and high-delight moments, step by step.
Most teams that skip this step aren't lazy — they just never had a structured place to put it besides a workshop poster nobody reopens after kickoff.
Prodinja ships today as an interactive prototype. Treat what follows as the capability it's designed to give you, not a result it has already produced for someone else's dashboard.
That's the same moment-level detail a PLG dashboard needs before its numbers mean anything: where hesitation spikes right before the paywall, where confidence peaks right after a first invite lands. Sketching that curve is what turns "activation rate dropped" into "activation rate dropped because step three of onboarding introduced a decision most users weren't ready to make yet."
It's a structured place to sketch that same curve you'd otherwise reconstruct from a support inbox and a hunch — not a system that has already run this analysis for another team.
Key Takeaways
- A PLG dashboard is 8 numbers pinned to 8 moments — traffic, signups, activation, retention, PQL rate, revenue, expansion, and referral — not a scattered set of whatever's easy to pull from an existing tool.
- Order the dashboard in journey sequence, left to right, so a drop in any one number points backward to the step that likely caused it.
- Pair every raw count with its ratio — total signups mean little without signup-to-activation rate; invites sent mean little without referred-user activation rate.
- PQL rate is the hinge metric connecting product usage to revenue, and it's the one most generic metrics lists skip entirely.
- Retention is a mid-journey diagnostic, not a monthly report card — a flattening cohort curve is one of the hardest signals on this list to fake.
- Assign one owner per metric, not per dashboard, and review each on a cadence that matches how fast that specific moment actually moves.
- A journey map built before the dashboard is what makes any of these eight numbers diagnostic instead of decorative.
Frequently Asked Questions
What are the 8 core PLG metrics?
The 8 core PLG metrics are traffic, signup rate, activation rate, retention, PQL rate, revenue (free-to-paid conversion), expansion (net revenue retention), and referral (K-factor). Each maps to one specific moment on the user's journey — from first visit through invite — rather than a generic funnel stage.
What's the difference between a PLG dashboard and a regular SaaS metrics dashboard?
A regular SaaS dashboard typically centers on revenue and churn, reported after the fact. A PLG dashboard adds upstream, product-led metrics — activation rate, PQL rate, time-to-value — because in a product-led motion, usage itself is the qualifying and converting mechanism, not a sales rep's pipeline notes.
Which PLG metric should I look at first?
Start with activation rate, since a broken activation step invalidates almost every metric downstream of it — retention, PQL rate, and revenue all depend on users reaching real value first. Signup rate matters too, but a healthy signup rate feeding a weak activation step just produces more people who bounce.
How often should PLG metrics be reviewed?
Review fast-moving, upstream metrics like signup rate and activation weekly, since onboarding issues compound quickly if left unfixed. Review slower, downstream metrics like net revenue retention and referral coefficient monthly or quarterly, since weekly reviews of those numbers mostly surface noise, not signal.
Is traffic really one of the 8 PLG metrics if it's a marketing number?
Yes, because traffic sets the ceiling for every metric downstream of it, even though marketing usually owns the channel mix. A PLG dashboard tracks traffic specifically paired with visitor-to-signup rate by channel, not as a vanity total, so the product team can see which channels send people who actually convert.