Civic Plus

Snapshot

What they do

CivicPlus builds and runs the technology local government depends on — public websites, agenda and records management, HR, emergency notification and citizen request handling — for roughly 2,400 government clients across twenty-plus years. Manhattan, Kansas; about 990 people.

Engagement

HubSpot estate audit and integration architecture, alongside the systems work of absorbing an acquisition.

Focus

Making twelve years of accumulated HubSpot safe to change — and making a business-critical integration something the team could actually maintain.

Delivery

Co-delivered by Value-First Team and USS.

The relationship

Jan 2026

Discovery. Their integration team brought us a business-critical HubSpot-to-Salesforce integration that nobody left in the company knew how to change.

Feb 2026

Integration audit delivered — the undocumented function reverse-engineered into documented behaviour, with an honest verdict on what the native integration could and could not replace.

Feb 2026

Streamline acquisition work begins. The scope is reframed inside the first session, from migrating a portal to designing how a newly acquired brand would operate. Migration guide, risk assessment and property audit delivered.

Mar 2026

The migration is deliberately re-timed to mid-year to protect a period of high business activity — a decision the evidence supported rather than resisted. Slack workflow channel migration executed across 60 workflows.

May 2026

Working sessions with their marketing-side HubSpot admin move to the form library, the workflow processing lag, and the contact-ceiling pressure behind both.

May–Jun 2026

The phantom-form investigation lands a root cause, and the form library audit blueprint sets out a cleanup that can be done without breaking anything live.

Where they started

A HubSpot portal that had been accumulating for twelve years, administered for most of that time by a single contractor who left no documentation, with three departments using it and none of them coordinating.

The sharpest edge was an integration. The custom function carrying form submissions into Salesforce had been built by an engineer who had since left, and the source repository was gone — not merely undocumented but missing, so the code could not be modified at all. It was a black box sitting in the middle of how the business captured demand. The team's hope was to delete it and switch to the vendor's native integration.

Around it, an estate nobody could account for: 942 real forms with twelve years of history, hundreds of workflows and dynamic lists pointing at each other, and no way to answer the only question that mattered — which of these still matter, and what breaks if we delete one? That uncertainty had become expensive. Form lists had grown long enough that workflow processing lagged behind them, so genuine submissions were falling out of the lead-creation path and being pushed through by hand. Their marketing-contact count was approaching the ceiling of their plan. And a further 299 records in the form library turned out not to be forms at all.

Meanwhile they had acquired Streamline, and had to decide what happens to a second live portal — while the acquired business was in its strongest period of demand.

What we delivered

An integration audit that documented the black box — what the function actually did, every field mapping, the product translation table and the business rules inside it — and answered their stated goal honestly: native could not replace this, so the work is to make it maintainable rather than to delete it.
A workflow dependency map across all 696 workflows, so that for any form we could name the live workflows that depended on it before anyone touched it.
The finding that de-risked the entire cleanup: enabled status, not reference count. One sign-up form was referenced by 38 workflows and every one of them was switched off. Of 112 form-referencing workflows, only 29 were live. The library looked far more fragile than it was.
A root cause for the 299 phantom forms — the platform's own form-collection feature, minting a new record on every page render — with the evidence, and a remedy that is a single settings toggle with no loss of contact data.
A form library audit blueprint — the six dimensions that actually determine whether a form still matters, and a surgical sequence: start with the 88 fully orphaned forms, archive the silent historical middle, and rename only what survives a keep decision.
A migration guide, a risk assessment and a property audit for the acquisition — including the silent-failure catalogue: the form embed that keeps accepting submissions into an abandoned portal while showing the visitor a success message, the workflow that reads as healthy and never fires again, the dynamic list that quietly matches nobody.
An executed workflow channel migration — 60 workflows re-pointed and verified down to zero stragglers, not merely documented as a plan.

What changed

They went from not being able to touch their own systems to being able to change them deliberately. The integration stopped being a black box and became documented behaviour with a maintainable path — and the reason it had to exist was finally written down, so the next person to ask "why don't we just use native?" gets an answer instead of a shrug.

The form library stopped being a guessing game. "What breaks if we delete this?" became a question with a checkable answer, and the fear that had frozen twelve years of cleanup turned out to be mostly unfounded — most of what looked load-bearing was already switched off. A mystery that had been quietly inflating their systems for twenty-one months got a named cause and a one-switch fix.

And the acquisition stopped being a deadline to survive. Their own evidence showed most of what everyone assumed had to move did not need to move at all, which let them re-time the work around the business instead of around the calendar — and build forward cleanly rather than carrying a decade of debt into a new system.