Marketing analytics
What should a marketing KPI dashboard track?
A marketing KPI dashboard should track the few numbers that change a decision: leads and qualified leads by source, conversion rates between funnel stages, cost per lead and per qualified lead, pipeline by source, and follow-up that is overdue. Each metric needs a written formula, one source of truth and a named owner.
The short answer
Build the dashboard in five sections, in this order, and define every number before it goes on screen:
- Acquisition: new leads and MQLs by source.
- Conversion: landing-page conversion, lead-to-MQL and MQL-to-SQL rates.
- Efficiency: cost per lead and cost per qualified lead for paid channels.
- Pipeline and revenue: pipeline created and closed-won revenue by source.
- Action items: overdue follow-up and broken records or workflows.
Start with the decisions the dashboard supports
A marketing reporting dashboard earns its place when someone changes a budget, a page, a handoff or a follow-up because of it. List the team's recurring decisions first; a metric that answers none of them stays off the main view. Typical decisions and the KPIs that answer them:
| Decision | Question | KPI that answers it |
|---|---|---|
| Budget split | Which channels get more spend next month? | MQLs and cost per qualified lead by source |
| Page and offer fixes | Where do visitors drop off? | Landing-page conversion rate by page |
| Lead quality | Will sales work these leads? | Lead-to-MQL and MQL-to-SQL rates |
| Follow-up | Which qualified leads are still waiting? | MQLs not contacted within the agreed window |
| Revenue contribution | Which sources create pipeline, not just form fills? | Pipeline created by source |
Marketing KPIs by dashboard section
These are the marketing KPIs I put on a first version, in the order the page should read.
Acquisition: leads and qualified leads by source
Show new leads and MQLs side by side for each source, so a channel with volume but few qualified leads stands out. Take source from a CRM field fed by UTM parameters, not from ad platform conversion counts; the setup is covered in campaign performance measurement.
Conversion: landing-page conversion, lead-to-MQL and MQL-to-SQL
Conversion rates show where the funnel leaks. Landing-page conversion points at the page and offer, lead-to-MQL at targeting, and MQL-to-SQL at the qualification criteria or the sales handoff. If sales keeps rejecting MQLs, revisit the definitions in MQL vs SQL before changing campaigns.
Show the count beside every rate: a rate built on a handful of MQLs is a weaker signal than the same rate on ten times as many.
Decide once whether you report period rates (MQLs this month ÷ new leads this month × 100) or cohort rates (the number of this month's new leads that have since become MQLs ÷ this month's new leads × 100). Period rates are simpler to build but can exceed 100% in a slow month, because some MQLs come from earlier leads. Cohort rates tie each lead to its own outcome but keep changing until the cohort has had time to qualify. The definition table below uses period rates.
Efficiency: cost per lead and cost per qualified lead
Cost per qualified lead matters more than cost per lead, because a cheap lead that never qualifies is not cheap. Take spend from the ad platforms and leads from the CRM, and mark organic channels not applicable rather than zero cost.
Pipeline and revenue by source
Pipeline created by source is the total amount on deals opened in the period, grouped by the original source of the first contact associated with the deal. Put closed-won revenue by source on a quarterly view, since deals often close months after the lead arrived, and write the attribution rule under the chart.
Action items: follow-up gaps and workflow bottlenecks
This section is a list, not a chart: records someone must act on this week. If the same leads keep appearing, the fix is usually upstream, in lead routing.
- MQLs with no sales activity after the agreed response window
- Open deals with no next step, or a close date in the past
- New contacts with a blank source or lifecycle stage
- Failed form submissions or integration syncs
KPI definition table: metric, formula, source of truth, owner
The definition table makes the numbers trustworthy. Agree each definition before you build: if marketing and sales count qualified leads differently, a dashboard can display the disagreement but cannot resolve it, a point made in why marketing reporting still requires so much manual work.
Each metric below gets a formula, one source of truth and an owner who answers when it looks wrong. In a small team one person may own several rows, but every row needs a name.
| Metric | Formula | Source of truth | Owner |
|---|---|---|---|
| New leads | New contacts from a form, booking or chat in the period, excluding customers and test records | CRM | Marketing operations |
| MQLs | Contacts that entered the MQL stage in the period | CRM lifecycle stage dates | Head of marketing |
| SQLs | Contacts sales accepted as qualified in the period | CRM lifecycle stage dates | Sales manager |
| Landing-page conversion rate | Leads ÷ landing-page sessions × 100 | GA4 (form key events and sessions) | Website manager |
| Lead-to-MQL rate | MQLs ÷ new leads × 100 | CRM | Marketing operations |
| MQL-to-SQL rate | SQLs ÷ MQLs × 100 | CRM | Sales manager |
| Cost per lead | Paid spend ÷ new leads from paid sources | Ad platforms (spend), CRM (leads) | Paid media owner |
| Cost per qualified lead | Paid spend ÷ MQLs from paid sources | Ad platforms (spend), CRM (MQLs) | Paid media owner |
| Pipeline created by source | Sum of amounts on deals created in the period, by the original source of the deal's first associated contact | CRM deals | Sales operations or RevOps |
| Closed-won revenue by source | Sum of amounts on deals won in the period, by the same source rule | CRM deals, reconciled with finance quarterly | Sales manager |
| Uncontacted MQLs | MQLs with no logged call, email or meeting after the agreed window | CRM activity history | Sales manager |
An example dashboard layout (illustrative)
Illustrative only: an example, not a client's dashboard. It is a one-page monthly view for a small B2B service business with leads from paid social, search and email, filtered by date, source and campaign. Use this layout and the definition table as a starting template, then rename the rows to match your own stages.
| Row | What it shows | Visual | Notes |
|---|---|---|---|
| 1. Headline | New leads, MQLs, cost per qualified lead, pipeline created | Number tiles | Versus last month and target |
| 2. Acquisition | New leads and MQLs by source | Grouped bars, last six months | Sorted by MQLs |
| 3. Conversion | The three conversion rates | Funnel or table | Counts beside rates |
| 4. Efficiency | Spend, leads, MQLs and both cost metrics per paid channel | Table | Organic marked not applicable |
| 5. Pipeline | Pipeline created and closed-won revenue by source | Bars, quarter to date | Attribution rule under the chart |
| 6. Action list | Uncontacted MQLs, stalled deals, missing sources, failed syncs | Table with owner and age | Oldest first |
| 7. Footer | Last refresh time and a link to the definition table | Text line | Links to channel detail pages |
Building it in HubSpot, Power BI, Tableau or a spreadsheet
The tool matters less than the definitions. Use the one that holds most of the data and that the team already opens. GA4 on its own is a source, not the dashboard: it can record sessions and form submissions but not lifecycle stages or deals.
Whichever tool you choose, connect the sources once so the dashboard refreshes on a schedule instead of from manual exports; that is the core of automated marketing reporting.
| Tool | A good fit when | Watch for |
|---|---|---|
| HubSpot dashboard | Contacts, lifecycle stages and deals live in HubSpot, with ad accounts connected | Available report types depend on your subscription |
| Power BI | Data comes from several systems and the team works in Microsoft tools | Needs a data model, a scheduled refresh and sharing licenses |
| Tableau | An analyst maintains it and people need drill-downs | The same modeling work, plus licenses for viewers |
| Excel or Google Sheets | A first version or low lead volume | Manual exports go stale; automate the pull with Make or Zapier |
Building a HubSpot dashboard
In HubSpot, the usual gap is data quality, not reporting features: blank source fields, lifecycle stages set by hand, deals with no contact. That is HubSpot setup work before it is dashboard work. Once those fields are reliable, build it in this order:
- 01Group new leads by the Original Traffic Source contact property, not by ad platform conversions.
- 02Count MQLs and SQLs by the date each contact entered the lifecycle stage (for example, Became a Marketing Qualified Lead Date), so each contact is counted in the month it reached the stage.
- 03Build pipeline created from each deal's create date and amount, and closed-won revenue from the close date on won deals.
- 04Use the same date range on every report on the dashboard.
- 05Keep the action list as a saved contact or deal view, sorted oldest first.
Keeping it maintainable: sources, owners, refresh and review cadence
Dashboards rarely break all at once: a field stops being filled, campaign names drift, the builder moves on. Four habits prevent that:
- One source per number. Each metric comes only from the source named in the definition table; change the table before changing a chart.
- An owner per metric. When a number looks wrong, its owner checks it that week.
- A visible refresh timestamp. Show when the data last updated, and by whom if it is manual.
- Clean inputs. A dashboard inherits every gap in the CRM, so if source or lifecycle fields are unreliable, audit your CRM data first.
Review cadence
Review at three speeds, and date-stamp any definition change so a jump in a trend is not misread as a change in performance.
| When | What to review | Typical outcome |
|---|---|---|
| Weekly | The action list and any sharp moves | Follow-ups assigned; broken syncs fixed |
| Monthly | Source trends, conversion rates, cost per qualified lead | Next month's budget and campaign changes |
| Quarterly | Pipeline, revenue and the definition table itself | Channel mix; unused metrics retired |
What to leave off the dashboard
Move these to a channel detail page, or drop them:
- Reach and engagement with no link to leads: impressions, followers and likes.
- Email open rates: mail privacy features can record opens that never happened, so use clicks and replies.
- The same metric from two sources: show one lead count from the CRM, not the CRM figure beside each ad platform's.
- Every campaign in full detail: give campaign scorecards their own page.
- Anything without a formula or an owner.
Questions this guide answers
What is a marketing KPI dashboard?
A marketing KPI dashboard, sometimes called a marketing reporting dashboard, is one view of the few measures a team uses to judge whether marketing produces qualified leads, pipeline and revenue. It pulls CRM, ad platform and web analytics data together, with a written formula, source and owner for each number.
How many KPIs should a marketing dashboard have?
Only as many as the team makes recurring decisions about. A practical test: if the headline row needs more than four or five tiles, it is answering too many questions. Move supporting metrics to detail pages. The layout above has four headline tiles; the other metrics in the definition table sit in the sections below them.
Can I build a marketing KPI dashboard in HubSpot?
Yes, if your contacts, lifecycle stages and deals live in HubSpot. A HubSpot dashboard can show leads and MQLs by source, stage conversion and pipeline, plus ad spend from connected ad accounts. Check which report types your subscription includes before you plan the layout.
Why don't my dashboard numbers match the ad platform?
Ad platforms count conversions with their own attribution windows, may include view-through conversions, and often claim leads another channel also claims. The CRM counts contacts, so use it as the source of truth for leads and keep platform conversions on channel detail pages.
Rather have this dashboard update itself?
Bring the report you rebuild every month to a 20-minute call. I'll show you which numbers can refresh on their own and which need cleaning up first.
Book a 20-minute dashboard reviewMichelle Ivanova runs imivs consulting, building marketing operations, CRM workflows and reporting automation in the tools a team already uses.