NEXSUM_LABS
  1. Home
  2. Work
  3. A B2B manufacturer's marketing team stopped filing HubSpot dev tickets
Book a call

[ Case study ]

ManufacturingHubSpot CMSHubLHubDBHubSpot CRM

A B2B manufacturer's marketing team stopped filing HubSpot dev tickets

The marketing site ran on HubSpot, but every layout change was a developer task: modules were one-off, fields were undocumented, and the marketing team had stopped requesting improvements because the queue was months long.

CLIENT a B2B industrial components manufacturer — FOCUS Agree the content schema before HubL

HubSpot CMS DevelopmentWeb DevelopmentHubSpot CMS DevelopmentManufacturingRepresentative example
Client
a B2B industrial components manufacturer
Industry
Manufacturing
Engagement
7 weeks — experience pod — frontend engineer + automation specialist
Service
Web Development / HubSpot CMS Development
Headline outcome
Monthly dev tickets for page builds, measured across the first quarter after handover: 12 → 0, read from Marketing's request log

Representative examplesEvery case study in this library is an illustrative composite of the kind of engagement we deliver — written to show our method and standards, not to name clients.

Where they started

Family-owned for four decades, this industrial components manufacturer sells through distributors and direct to plant engineers; the printed catalog remains its most trusted document — sales carries it, engineers reference it, and the website is expected to agree with it. Marketing is one person plus agencies. The website runs on HubSpot because the CRM does, and the product pages were built once, years ago, by whoever was available, then left alone.

What it was costing

The marketing site ran on HubSpot, but every layout change was a developer task: modules were one-off, fields were undocumented, and the marketing team had stopped requesting improvements because the queue was months long.

What they could see

  • Every layout change meant a dev ticket, and the queue ran months; even small fixes waited behind other work.
  • The marketing team had stopped requesting improvements altogether — the request log filled with items closed without a build.
  • Product specs on the website disagreed with the printed catalog, and sales heard about it from engineers first.
  • When the one marketer took leave, nothing on the site could change at all.

The constraints we worked inside

  • Product data had to stay consistent with a printed catalog the sales team also used.
  • The CRM was mid-migration; content had to work with both the old and new property sets during the build.
  • Only one in-house marketer owned the site — documentation had to assume zero continuity.

What had been tried before

The marketer rebuilt two key pages herself with HubSpot's stock template library.
Stock templates couldn't represent the catalog structure, sales said the pages looked off-brand, and she reverted them within a month.
Their IT contractor was paid to build custom modules for the product pages.
The modules worked but shipped undocumented and one-off; the next change needed the contractor again, so the queue just changed owners.
A shared document listed every module and field on the site.
Nothing enforced it, so the document described a version of the site that no longer existed by the second quarter.

What we proposed

We proposed staying on HubSpot and rebuilding the way pages get made: an agreed content schema first, then a documented module library that assembles product pages, application notes, and landing pages from safe, editable parts. The marketer composes; catalog consistency is kept by structure, because product data lives in HubDB tables that mirror the printed catalog's columns. Because the CRM was mid-migration, every form and smart-content rule would be wired against both property sets from the start — the site had to survive the cutover without lead routing breaking.

Just as important is what we ruled out, and why:

  • Rebuilding the marketing site on a second CMSA mid-CRM-migration plus a one-person marketing team could not operate two systems; the integration, not the platform, was the failure.
  • Commissioning more custom one-off modulesUndocumented one-offs were the current problem; paying for more of them buys the same queue with a different name.
  • Generating all pages dynamically from HubDB aloneApplication notes and landing pages need editorial freedom that product pages don't; one dynamic pattern would flatten page types that behave differently.

How the work ran

01Agree the content schema before HubL

We defined page types and module fields in writing — product pages, application notes, landing pages — so every module got built once with the right editable surfaces.

02Build a module library, not pages

Reusable HubL modules with documented fields replaced one-off templates, giving the marketer real composition freedom inside safe boundaries.

03Bridge the CRM migration

Smart content and forms were wired with property mapping that worked through both CRM schemas, so no lead routing broke mid-project.

Delivered by the experience pod — frontend engineer + automation specialist over 7 weeks, with working increments reviewed with the client every week.

The stack, and the reasoning

HubSpot CMS
The CRM already runs the company; keeping site and CRM on one platform leaves a one-person marketing team one system to learn rather than two.
HubL
Modules with documented editable fields end the one-off pattern; the marketer composes within boundaries instead of requesting rebuilt templates.
HubDB
Product spec tables mirror the printed catalog's columns, so site and catalog read from one structure and sales stops reconciling the two.
HubSpot CRM
Mid-migration, both property sets had to work; smart content and forms were mapped against old and new schemas so lead routing survived the cutover.

What went wrong

Obstacle

In week four the CRM migration renamed a lead-source property, and smart content on the application-note pages went blank in staging.

Handled: We rebuilt the property mapping to read both schemas, re-tested every smart-content rule against the new set, and the bridge held through their cutover.

Obstacle

Product specs existed in three shapes — catalog PDFs, a dealer spreadsheet, and years of page copy — and the three disagreed in places.

Handled: We reconciled everything against the printed catalog as the source of record, loaded HubDB from it, and left the marketer a written diff wherever page copy had disagreed.

How we worked together

Cadence
A 30-minute call every Tuesday and written progress every Friday; the marketer reviewed module builds on staging between calls and logged requests in a shared board.
Client side
One in-house marketer owned the site and attended everything; a sales operations analyst answered catalog questions; the IT contractor joined for the CRM-bridge work.
Decisions
Page-type and field decisions were signed off in writing before build; mid-sprint calls went to the marketer, with sales escalation for anything touching catalog data.
They provided
The printed catalog as the spec source, dealer spreadsheets, both CRM property sets during migration, and a weekly hour of the marketer's time for editing trials.

What changed

The headline: monthly dev tickets for page builds, measured across the first quarter after handover12 → 0, read from Marketing's request log. A second check: product-page update turnaround at 5 days → same day.

The marketer proposes improvements again — the request log filled with small ideas because small ideas became cheap to ship. Product updates land the same week sales hears about them, and the printed catalog and the site agree, so sales stopped adding check-the-website disclaimers to every quote. The documentation assumed zero continuity and has since proven it: a temp covered the marketer's leave using the module docs alone, and nothing broke.

The result was read from Marketing's request log against the pre-engagement baseline over the stated window, with a guardrail check on product-page update turnaround. Where platform-reported numbers and business outcomes differ, this record says which layer it is quoting.

What they own now

  • The module library with documented fields, each editable surface labeled in plain language.
  • A content schema document covering page types, module fields, and naming rules.
  • HubDB tables mirroring the printed catalog, with a refresh procedure the marketer runs.
  • The CRM property-mapping bridge that works across both schemas, with a post-migration cleanup note.
  • Zero-continuity documentation written for a marketer who has never seen the build.

What we would do differently

We would have scheduled the marketer's walkthrough earlier — the first module revision cycle happened before she had seen the editing model.

Web DevelopmentHubSpot CMS DevelopmentManufacturingHubSpot CMS

Next case study

A commercial cleaning company got same-day quotes instead of five-day ones