HubSpot Service Analytics

HubSpot service analytics will happily give you forty charts. Five of them change decisions. The rest are wallpaper for a monthly meeting.

What HubSpot service analytics actually gives you

Reporting depth is tier-dependent, and the gap between tiers is one of the real reasons teams upgrade.

Out-of-the-box service dashboards. Volume, response times, agent activity. Fine for a first month.
Custom reports built on ticket properties, with the flexibility and the row limits varying by tier.
Calculated timing properties, such as time to first agent reply and time to close, computed for you rather than derived in a spreadsheet.
Knowledge base and help centre metrics, including article views and searches.
Survey results from CSAT and feedback surveys.
Conversation analysis on the higher tiers, including themes across volume.

The honest summary: adequate at the lower tiers, genuinely good at the top, and never as flexible as a proper BI tool pointed at exported data.

The five reports worth building

Build these, put them on one dashboard, and delete everything else for a quarter. If nobody asks for a deleted chart, it was wallpaper.

Volume by category, weekly. What are we spending our time on, and what's growing. This is your knowledge base roadmap and your automation roadmap in one chart, and it only works if ticket categorisation is disciplined.
Median time to first response. Median, not average. One migration escalation that took nine days will wreck an average and tell you nothing about the customer experience.
Backlog by age. How many open tickets in each age bucket. This catches the slow rot that a volume chart hides completely.
Reopen rate. The purest quality signal you have. A ticket closed twice was not resolved the first time, and rising reopens usually means somebody is being measured on closure counts.
Volume by customer or company. Which accounts are consuming your support capacity. Uncomfortable, occasionally career-defining, and impossible to build without clean company records.

What the numbers hide

Every support metric has a failure mode, and every one of them gets gamed by ordinary people responding sensibly to how they're measured.

Average response time hides the tail. Report the median and the 90th percentile together. The median is the experience, the tail is where the complaints come from.

Resolution time counts waiting on the customer unless your pipeline separates it. If a stage change doesn't stop the clock, half your resolution time is you waiting for a screenshot.

Volume hides duplicates. A queue where one problem arrives twice looks busier than it is, and typically 8% to 20% of tickets are copies of another ticket. Your per-agent throughput, your volume trend and your topic mix all inherit that error.

CSAT is answered by the extremes. Response rates are low and skewed towards the delighted and the furious. Watch the trend, distrust the level, and never rank agents on it without reading the comments.

Closed counts reward the wrong behaviour. Measure a team on tickets closed and you'll get more closures and more reopens. Pair the two numbers or don't use either.

If a metric can be improved without a customer noticing anything, it will be, and nobody involved will feel dishonest doing it.

Dashboard discipline

A dashboard nobody reads is worse than no dashboard, because it creates the impression of measurement.

One dashboard, one audience. The team dashboard and the exec dashboard are different documents with different numbers on them.
Six charts maximum on the one people look at weekly.
Every chart needs an owner and an action. If nobody can say what they would do if the line moved, take it off.
Annotate the anomalies. A spike with a note saying "outage, 14 March" is knowledge. A spike with no note becomes an argument in six months.

When to export instead

HubSpot reporting handles the operational questions well. It handles the analytical ones less well.

If you want to correlate ticket volume with product releases, model support cost per account against revenue, or run cohort analysis, export and use a real analysis tool. Trying to force that shape into the report builder produces a chart that's nearly right, which is the most expensive kind of chart.

FAQ

Frequently asked questions

Where does ticket reporting live in HubSpot?

In the reports tool and on dashboards. HubSpot ticket reporting draws on ticket properties, HubSpot Service Hub reporting adds the prebuilt library, and a HubSpot service dashboard is where the five HubSpot support reports worth reading end up.

Does HubSpot service analytics need a paid tier?

Basic dashboards appear early, but custom reports and the deeper analysis scale with tier. If reporting is a buying criterion, ask to see the report builder on the exact tier you're quoted.

What is the single most useful support report?

Volume by category over time. It tells you what to document, what to automate and what to staff for, which is three decisions from one chart.

Should we report average or median response time?

Median, with the 90th percentile beside it. Averages in support data are dominated by a handful of outliers and describe nobody experience accurately.

Do duplicate tickets distort service reporting?

Yes, materially. Duplicates are commonly 8% to 20% of a queue, and they inflate volume, distort per-agent throughput and skew your topic mix without ever appearing as an error.

Reporting on work that was never real

Duplicate tickets quietly inflate every number on your dashboard. Ticket Merger removes them before they land in your reports, on Zendesk and Freshdesk today, with HubSpot on the roadmap.

Start free trial

14-day free trial. No credit card required.