Most teams don’t have a reporting problem. They have a data trust problem. Reps don’t believe the numbers. Managers pull their own spreadsheets. Dashboards exist but nobody checks them. Sound familiar?
Join us for a Small Group Coaching session
We run a session on HubSpot analytics end-to-end, and three things came up every time that are worth sharing with the community — because they’re not obvious until someone points them out.
Mistake 1: You’re filtering on the wrong thing and don’t know it
Filtering by “lifecycle stage = Customer” shows everyone currently at that stage. Filtering by “lifecycle stage changed to Customer in the last 30 days” shows who moved through it in that window. These are different questions — and most people are running the first one when they actually need the second. If your conversion numbers feel off, check this first.
Mistake 2: Associations are inflating your pipeline totals
One deal linked to three contacts shows up three times in a contacts-based report. Before you share a pipeline total or conversion rate, ask yourself: does this report multiply rows? If you’re not sure, export to a list view and manually check the count against what your report says.
Mistake 3: Starting with the builder instead of the library
The custom report builder is powerful, but it’s not always the right first move. HubSpot’s report library has hundreds of pre-built templates across sales, marketing, and service. Most people never look there first and spend 20 minutes building something that already exists. Start with the library, find something close, then customize.
The framing that actually changes how people build reports
Before touching the report builder, answer three questions: What business question am I trying to answer? How is that data being captured in HubSpot today? Where might there be gaps or inconsistencies in the underlying records? Skipping this step is the reason most custom reports either return weird numbers or get abandoned.
On dashboards specifically
A report answers a question. A dashboard helps a team monitor a set of questions over time. Those are different jobs. If your VP opened your dashboard for 30 seconds, what should they understand immediately? If the answer isn’t obvious, the dashboard isn’t done — regardless of how many reports are on it.
Happy to go deeper on any of this — cross-object reports, attribution setup, AI-generated reports with Breeze, or dashboard automation. What’s the reporting challenge your team keeps running into?