← blogNovember 13, 202613 min read

How to Build an Email KPI Dashboard That Gets Used

TL;DR. An email KPI dashboard works when you limit the main view to five to seven KPIs tied to one campaign goal, define every formula once with unique counts over delivered, and review it weekly with a single owner. Build it with a fixed field schema and a small sample dataset first, confirm your arithmetic matches your tool's report, then swap in real exports. Write the attribution window on the dashboard so revenue trends stay comparable month to month.

Quick Summary

An email KPI dashboard is a single screen that shows the few numbers that tell you whether your email is working, pulling opens, clicks, replies, bounces and unsubscribes into one view. Build it by fixing a field schema, calculating each KPI with a consistent formula, and reviewing it on a set cadence with one owner.

  • Limit the main view to five to seven KPIs tied to your campaign goal.
  • Define each formula once, using unique counts and delivered as the denominator.
  • Write the attribution window on the dashboard so revenue trends stay comparable.
  • Split results by lifecycle segment so each group is measured against its own baseline.
  • Review weekly and record one number that moved plus the action it triggers.

What an Email Marketing Dashboard Is and Why It Matters

An email KPI dashboard is a single screen that shows the handful of numbers that tell you whether your email is working. It pulls opens, clicks, replies, bounces and unsubscribes into one view, so you can see trends without digging through reports.

Plenty of teams send email and then only check results when someone asks, which tends to mean a month later, in a meeting, with the wrong export open. A dashboard fixes the timing problem more than the measurement problem. When the numbers sit in front of you every week, a slow drift gets caught while it's still a drift.

What you really gain is speed of diagnosis. You stop asking how a campaign did and start asking why it beat, or lost to, the one before it.

Email Marketing KPIs Worth Tracking and How to Calculate Them

A KPI is a metric tied to a goal. Most of what your sending tool displays isn't a KPI; it's just a number it happened to collect.

KPIFormulaFields requiredMeasurement caveat
Delivery rateDelivered / Sent x 100sent, deliveredSent should exclude suppressed and skipped rows, or the rate understates
Open rateUnique opens / Delivered x 100delivered, unique opensApple Mail Privacy Protection and similar prefetching inflate opens; treat as directional
Click-through rateUnique clicks / Delivered x 100delivered, unique clicksBot and security-scanner clicks inflate this on some lists
Click-to-open rateUnique clicks / Unique opens x 100unique opens, unique clicksInherits open-rate distortion, so it moves when opens move, not only content
Bounce rateBounced / Sent x 100sent, bouncedSplit hard and soft; a soft bounce is not a dead address
Unsubscribe rateUnsubscribes / Delivered x 100delivered, unsubscribesOne-click List-Unsubscribe makes leaving easier, so the rate can step up after you add it with no change in content
Complaint rateSpam complaints / Delivered x 100delivered, complaintsSmall denominators make this jump; report on a rolling 30 days. Gmail doesn't report individual complaints back to your sending tool, so check Google Postmaster Tools for its spam rate
Reply rateUnique replies / Delivered x 100delivered, repliesOnly meaningful on 1:1 sends, not broadcasts
Revenue per emailAttributed revenue / Delivereddelivered, revenue, attribution windowThe window you pick changes the number more than the campaign does

Here's a worked example you can check your own arithmetic against. A campaign sends 2,000, delivers 1,940, and gets 611 unique opens, 93 unique clicks, 8 unsubscribes and 2 complaints.

  • Delivery rate: 1,940 / 2,000 x 100 = 97.0%
  • Open rate: 611 / 1,940 x 100 = 31.5%
  • Click-through rate: 93 / 1,940 x 100 = 4.8%
  • Click-to-open rate: 93 / 611 x 100 = 15.2%
  • Unsubscribe rate: 8 / 1,940 x 100 = 0.4%
  • Complaint rate: 2 / 1,940 x 100 = 0.1%

Do that arithmetic by hand once and compare it against your tool's own report. When the two disagree, the denominator is almost always the culprit, because some tools divide opens by sent instead of delivered. That gives a lower rate than yours.

Settle the attribution window before you build the revenue row, not after someone questions a number. A metric nobody has pinned down can't be compared to itself next month.

Reply rate deserves a mention because broadcast dashboards tend to drop it entirely. If you run 1:1 sequences, count replies against delivered sends and put it on the main view.

Keep every formula in one sheet, versioned, with the date it last changed. Six months from now somebody will argue about a number, and that sheet is how the argument ends.

Choosing KPIs That Match Your Campaign Goals

Pick KPIs by goal, not by whatever your tool puts on the default screen. A product launch lives or dies on click-through and conversions. A list cleanup is about bounce and unsubscribe rates, and the open rate going up afterwards is a side effect, not the point.

Keep the main view to five to seven KPIs. Past that, people skim instead of read, and a dashboard that gets skimmed is worse than three numbers in an email. Everything else goes on a secondary tab where the curious can find it.

Give each campaign type one primary KPI and let the rest be supporting evidence. For a welcome series that might be the share of new subscribers who make a first purchase. Secondary metrics explain the result; the primary one decides whether you call the campaign a success. If you want to turn these numbers into a return figure, how to calculate email marketing ROI walks through the method step by step.

Building an Email KPI Dashboard Step by Step

What follows is reproducible from end to end: a fixed field schema, a sample dataset, the formulas applied to that dataset, and the views you end up with. Build it once with the sample data, confirm the numbers match, then swap in your own export.

Step 1: Fix the field schema before you touch a chart

One row per send. These are the columns, and anything your export doesn't contain stays a blank cell rather than a plausible-looking estimate.

FieldTypeExampleSource
campaign_idtextwelcome_01sending tool
campaign_typetextwelcome, broadcast, sequencesending tool
send_datedate2026-10-04sending tool
segmenttextnew_subscriber, inactive_90dsending tool
sentinteger800sending tool
deliveredinteger784sending tool
unique_opensinteger378sending tool
unique_clicksinteger71sending tool
bouncedinteger16sending tool
unsubscribedinteger2sending tool
complainedinteger0sending tool
repliesinteger0sending tool
revenuedecimal1240.00web analytics or order data

Missing a field? Leave the column out and drop the KPI that depends on it. Half-populated rows cause more arguments than they settle, and you can always add the metric back once the data actually arrives.

Step 2: Load the sample dataset

Four rows, enough to catch a broken formula before it reaches real data. The product_tip_10 row is the campaign from the worked example above, and all four rows feed the sample dashboard further down this article.

campaign_idcampaign_typesegmentsentdeliveredunique_opensunique_clicksbouncedunsubscribedcomplained
welcome_01welcomenew_subscriber800784378711620
product_tip_10broadcastall_active20001940611936082
trial_offer_10broadcasttrial_started150014614089239101
reengage_10broadcastinactive_90d700651120144971

Calculate each KPI in a new column. The product_tip_10 row should match all six figures in the worked example above, and every row should match the sample dashboard below. If they all match, your formulas are sound, so point them at the real export.

Step 3: Clean before you calculate

Most bad rows get caught by the same few checks:

  • Test sends. Anything with a recipient count under 10, or a campaign name containing "test", should be filtered out before it reaches a chart.
  • Duplicates. Deduplicate on campaign_id plus send_date, because a re-export will happily append a second copy of rows you already have.
  • Rows that don't reconcile. On every row, confirm delivered never exceeds sent, and if your tool counts every non-delivery as a bounce, that delivered plus bounced equals sent. Checking only the totals can hide one bad row. A row that fails has a wrong denominator, and every rate built on it is wrong too.

Watch out: paste in a new export whose columns come in a different order, and a single shifted column will drop opens into the clicks field. Every rate on the dashboard is then wrong, and nothing about the output looks wrong, which is the part that gets people. Check the header row of every export against the schema before you load it.

Step 4: Build the four views

ViewChart typeWhat it answersRefresh
Primary KPISingle number with trend arrowAre we on track this week?Daily
TrendLine chart, one line per KPIIs performance moving up or down?Weekly
Campaign comparisonBar chart, sorted by click rateWhich campaign drew the most clicks?Per send
Segment tableTable with exact figuresWhich audience is underperforming?Weekly

Primary KPI goes top left, where the eye lands before anyone decides whether to keep reading. The segment table goes last, since that's the view people open after the trend line moves and they want an explanation. For which chart suits which question, and how to keep axes and time windows honest, see how to visualize email campaign performance.

Step 5: Connect the data sources and set the attribution window

Most teams are pulling from three places: the sending tool for sends, opens, clicks, bounces and unsubscribes, web analytics for sessions and conversions from email, and order or signup data for revenue. Join them on a campaign identifier that exists in all three. If the links in your emails don't carry a campaign parameter (UTM tags), add one before the next send, otherwise the revenue column stays empty and you'll spend a week wondering why. And if your tool rewrites links to count clicks, a custom tracking domain puts those links on your own domain instead of a shared one.

Then pick an attribution window and write it on the dashboard itself. Match it to how long people take to decide: 3 to 7 days after the click for a trial signup, the trial length plus a few days for the paid upgrade at its end, and closer to 30 for a purchase people think over. Setting up an email attribution model goes through the choice in more detail. Don't agonize over it. What matters is that it's the same choice every week, because a window that quietly changes turns your revenue trend into noise.

Step 6: Set the review cadence and the owner

One person owns the dashboard and raises the flag when something shifts. Weekly while campaigns are running, monthly for the longer trend. Each review should produce one written line: the number that moved, and what you're going to do about it. If the review never changes what you send next, you've built a reporting habit rather than a measurement one.

If you'd rather the send-side data came from a single place, SEMAOS runs broadcasts and 1:1 sequences in one tool. Each broadcast and sequence can carry a lifecycle stage label, and Analytics has a By lifecycle stage table, on every plan. Paid plans and the trial add a nightly engagement score per contact on a 0 to 100 scale. Two caveats: the Free plan has no engagement score, and replies don't count toward it, so keep a reply-rate row alongside it if replies are what you care about. SEMAOS doesn't track purchases or revenue, so the revenue row still comes from your analytics or order data.

One more thing to know before you trust the send-side columns: platform sends always track opens and clicks, but sends from a connected Gmail or Outlook mailbox track neither until you switch tracking on under Settings, Sending (click tracking there also needs a verified custom tracking domain, a paid-plan feature). Bounces on mailbox sends go back to your own inbox, not to SEMAOS. Replies work the same way on Gmail mailbox sends, so count those by hand, and platform sends only see replies once "Detect replies & auto-stop sequences" is on under Settings, Sending.

A Sample Email KPI Dashboard, Worked Through

The table below shows four campaigns. These figures are illustrative, not pulled from a real account:

CampaignSentDeliveredOpen RateClick RateUnsub RateBounce Rate
Welcome 180078448.2%9.1%0.3%2.0%
Product tip2,0001,94031.5%4.8%0.4%3.0%
Trial offer1,5001,46127.9%6.3%0.7%2.6%
Re-engagement70065118.4%2.2%1.1%7.0%

Read it for patterns rather than for the individual numbers. The welcome series comes out on top, which is what you'd expect from subscribers who signed up days ago. The re-engagement campaign carries the highest unsubscribe rate by some distance, and that's worth attention: either the list is past saving or the offer isn't landing. Its 7% bounce rate is the bigger problem, though. Split hard bounces from soft ones and remove the dead addresses before that list gets another send, and look at why the other three campaigns sit at 2% or above as well.

A simple trigger: any campaign whose unsubscribe rate clears 1% gets an audience review before the next send goes out.

Segmentation, Benchmarks, and Data Quality Pitfalls

Segment before you compare anything. A lifecycle stage framework gives you groups that stay consistent from month to month. An average open rate across the whole contact list mostly tells you about the ratio of new subscribers to dormant ones, which is not what anyone thinks they're reading.

Treat published industry benchmarks with suspicion. A benchmark drawn from a different audience, sending at a different frequency, measured with a different denominator, tells you very little about your own program. Your own results from last quarter are a better reference point.

Data quality problems are duller and more damaging. Duplicate contacts inflate counts, and missing timestamps break cohort analysis outright. A quick monthly check for both is worth the twenty minutes.

Small samples will also lie to you. A 3% click rate on 200 sends is six clicks, so wait for volume before declaring a winner.

Set Up Your First Dashboard Review This Week

Thirty minutes is enough to get the first version standing. Block it out and do the following:

  • Write down your primary goal, then the five KPIs that actually support it.
  • Export the last 60 days of campaign data and run each KPI against it.
  • Finally, note the one number that surprised you, and what you'll do differently because of it.

Then do it again next week, and the week after. A scrappy spreadsheet that someone opens every Monday will do more for you than an expensive dashboard that gets looked at twice a year.

FAQ

How many KPIs should an email dashboard show?

Five to seven on the main view. Past that, people skim instead of read, and a skimmed dashboard is worse than three numbers in an email. Put everything else on a secondary tab for the curious.

Should open rate be calculated on sent or delivered?

Delivered. Some tools divide unique opens by sent instead, which makes every open rate a little lower and means two tools can disagree about the same campaign. When your hand arithmetic disagrees with the tool's report, the denominator is almost always the cause.

What attribution window should I use for email revenue?

Match it to how long people take to decide: 3 to 7 days after the click for a trial signup, the trial length plus a few days for the paid upgrade at its end, and closer to 30 for a purchase people think over. What matters most is that it stays the same every week and is written on the dashboard itself.

Are industry email benchmarks useful?

Treat them with suspicion. A benchmark drawn from a different audience, sending at a different frequency and measured with a different denominator says little about your program. Your own results from last quarter are a better reference point.