Case study · HubSpot
Dashboard Sections: solving a five-year-old request
Customers asked for a clean way to organize HubSpot dashboards for five years. I found out what “good enough” really meant, prototyped it from existing design system parts, and shipped a front-end-first version without waiting on backend capacity.
COMPANY
HubSpot
PRODUCT AREA
Reporting & Dashboards
MY ROLE
Senior Product Manager
REQUEST OPEN
2020 to 2025
THE CONTEXT
Where this started
Why now, after five years?
Sections had been on the wish list since before I joined. Text and image blocks became the workaround, but they took up as much room as a report, and the only way around a big dashboard was scrolling. Technical complexity, and a sense that the workaround was good enough, kept it on the back burner, until we refocused on dashboard polish. In the age of AI, a dashboard has to be worth coming back to.
What was I responsible for?
I owned getting sections out the door: confirm what customers actually needed, design it from what we already had, and find a way to ship it without waiting on backend capacity.
01 · HOW IT CAME TOGETHER
From “good enough” to shipped
Phase 1
Validate
WHAT DID “BETTER” MEAN?
→ We reached out to a few vocal customers who were frustrated with using note cards as headers.
→ We asked why notes weren't good enough and what would make them better.
→ The bare minimum: something much smaller that took up less room. Bonus points if it could be rearranged.
Phase 2
Design and beta
BORROW, THEN POLISH
→ Static sections already existed in our design system, so I borrowed them and built a rough design.
→ Design added the polish: collapsing, and edge cases that needed their own components.
→ The public beta ran about 6 weeks, gated by engineering capacity, not by a lack of proof. Feature CSAT came in at 98%, and only one user opted out.
→ What changed: sections first opened expanded every time. But a dashboard can be huge with only one section relevant to you, so we started remembering each person's collapsed state.
Where AI helped: a rough first version in Lovable, built from existing design system components.
Phase 3
Launch
SHIP SMALL, THEN MEASURE
→ Building it “properly” on the backend would have taken much longer, and our backend team was finishing a larger project.
→ So we shipped a front-end-only version first, with a transition plan agreed with engineering.
→ It rolled out across all hubs and tiers. Success meant a behaviour change: people converting their dashboards to sections, and fewer note cards added as makeshift headers.
→ Lagging indicator: overall dashboard usage for people with sections compared with those without, alongside the Dashboards app CSAT.
02 · THE KEY DECISION
Ship value now. Build it properly next.
IT WASN'T
“Wait for backend capacity and build it all in one go.”
IT WAS
“Ship a front-end-only version now, with a plan to move it to the backend.”
Our backend team was finishing a larger project. Going front end first let us validate and deliver value sooner, and I worked with engineering on the transition plan before we started.
03 · THE TRADE-OFF
What made v1, and what waited
✓ In v1
Create, name, reorder and collapse sections, and move reports between them, built front end only.
→ What came next
Fast follows: adding a description to a section, and better ways to add reports from inside one. Next we're exploring section-level permissions and filtering, plus moving sections onto the backend as planned.
My role: I owned it from validation to launch: the customer conversations, the first prototype, the front-end-first call and the transition plan, partnering with design on polish and engineering on the rollout.
04 · THE OUTCOME
What happened next
5 yrs
the Ideas thread stayed active, 2020 to 2025
98%
feature CSAT in the public beta, with just one opt-out
FE first
shipped without waiting on backend capacity, with a migration plan ready
BEFORE
“Every client portal I open has one dashboard with fourteen reports in no particular order, because reordering them meant going back into settings and fighting with it.”
Casey Hawkins
HubSpot Freelance Consultant · LinkedIn, Sep 2026 →
AFTER
“HubSpot dashboards just got the overhaul they've needed for years.”
Frederik Ramskov Banja
Specialist, Helion (HubSpot Elite Partner) · LinkedIn, Oct 2026 →
AFTER
“Especially useful when several people are working with the same dashboard.”
Lisa Freitas
Senior HubSpot Consultant, Thalox · LinkedIn, Oct 2026 →
05 · WHAT TO TAKE FROM IT
Three lessons for new PMs
1
Delivered isn't solved
“Delivered” on a roadmap isn't the same as solved for the customer. Read the comments after a launch, not just before.
2
Old requests are free discovery
A long-running thread already tells you the job, the pain and what a bad fix looks like.
3
Ship the smaller version first
Going front end first got value to customers while the “proper” build waited for capacity. A transition plan made it a step, not a shortcut.
Sources: Section Headings for Dashboards, HubSpot Ideas (HubSpot Community). Image: HubSpot Knowledge Base.
Your turn.
Want help turning your experience into a story like this?
I mentor career switchers from CS, support, marketing and analytics. Start with a free 30-min clarity call. We'll map your story and your next step. No sales pitch.