A marketing analytics dashboard exists to explain why a result moved, not just to report that it did. A reporting dashboard says booked calls fell this month. An analytics dashboard lets you find out whether that was fewer leads, slower replies, a weaker channel or a single broken form. This article answers how to build that diagnostic view in Reporting on the levelpartner.ai platform, and how to use it to trace a problem from outcome back to cause in a single sitting.
Reporting versus analytics
Reporting is a fixed view, read on a schedule. It answers "how are we doing". Analytics is a set of views you drill into when something changes. It answers "why".
Most firms of 10 to 250 people have some reporting and almost no analytics. When the monthly figure drops, the meeting speculates. With an analytics dashboard, someone can open it and narrow the cause down in ten minutes.
The four questions an analytics dashboard should answer
- Did volume change? New contacts by source, compared with last period.
- Did quality change? Share of contacts that reached a booked call, by source.
- Did speed change? Time from first contact to first reply, and from booking to call.
- Did conversion change? Stage-to-stage rates through the pipeline.
Each question gets its own group of widgets. When a result drops, you work down the four in order and stop at the first one that moved.
Before you start
- A reporting dashboard that people already read, so you know which outcomes matter.
- Source captured on contacts and opportunities.
- Conversations connected, so reply times are recorded.
- A pipeline with stages that the sales team updates consistently.
How to set it up
- Go to Dashboard, click + Add Dashboard and name it marketing diagnostics.
- Add Contacts Count Over Time; set First Attribution and group by Session Source.
- Add the Funnel widget from General and use Next-Step Conversion.
- Add Average response time trend from Conversations, with SLA switched on.
- Add Appointment Counts by Status to track no-shows.
- Add Stages Distribution from General for the pipeline you use.
- Add Lead Source Report and check the N/A row.
- Click Save Changes and set a comparison range in the Date Picker.
- Write the four questions in each widget's Description.
How to use it when a number drops
Start at volume. If new contacts fell, check which source fell. A single source dropping usually means a broken link, a paused campaign or a lost listing. If volume held, move to quality. If one source's booked call rate fell, look at what changed in that source's targeting or landing page.
If quality held, check speed. Slower replies are the most common hidden cause of falling conversion. If speed held, check pipeline conversion stage by stage. The first stage that fell is where to focus.
Say booked calls fall from 44 to 29 month on month while volume holds at about 180 contacts. Quality by source shows paid search steady but the paid social booking rate halved. Speed shows average response time rising from 12 minutes to 3 hours because the person covering the inbox is on leave. In that scenario a staffing gap, not marketing, would explain most of the drop, and an AI agent that answers and books during cover gaps would be the fix to try.
What it replaces
It replaces the meeting where everyone offers a theory about why numbers fell. It also replaces ad-hoc exports to spreadsheets that someone pivots and re-pivots to find a cause. The views are ready before the question is asked.
Time it takes
Building the diagnostic dashboard takes two to three hours. It does not need a weekly review. Open it when a reporting figure moves outside its normal range. A typical diagnosis takes 10 to 20 minutes.
The mistake most people make
Jumping straight to the channel when a result drops. Teams often cut a channel that looks weak this month, when the real cause was slower replies or a sales process change. Always check volume, quality, speed and conversion in order before blaming a source.
When not to use it
A diagnostic dashboard is wasted effort if nobody reads the reporting dashboard regularly, because you will not notice when to open it. Fix the reporting habit first. It also adds little for very small volumes, where a single deal can move every rate. Below about 30 new contacts a month, read individual records instead.
For the fixed reporting view that triggers diagnosis, see marketing reporting dashboards. To understand stage-level behaviour more deeply, read customer journey analysis. For the metric definitions behind each widget, see the marketing metrics dashboard.
Troubleshooting
Reply time looks too fast. Automated acknowledgements are counted as replies. Filter to replies sent by a team member or an AI agent that actually answers the question.
Comparison periods are uneven. One month has more working days. Compare equal-length periods or use weekly averages.
Stage conversion shows rates above one. Deals were moved backwards and forwards between stages. Ask the team to move deals forward only and mark lost deals as lost.
The no-source count is high. Fix capture before trusting source-level analysis. Every diagnosis that depends on source is weaker until then.
Frequently asked questions
How is this different from a marketing dashboard?
A marketing dashboard reports outcomes for regular review. The analytics dashboard holds diagnostic views you open when an outcome changes and you need to find the cause.
Who should own the analytics dashboard?
Whoever is expected to explain changes in marketing results, usually the marketing lead or an operations analyst.
Can AI help with the diagnosis?
AI tools can summarise what changed between periods. The person reading still needs to check the order of causes, because a summary will not know about a staff absence or a site change.
Do I need separate dashboards for each channel?
No. Add a source filter to each widget so you can narrow the whole view to one channel when needed.