ArcGIS Enterprise Upgrade Program

Accela's civic platform integrates deeply with Esri's ArcGIS Enterprise, and Esri ships on its own cadence. That created a standing tension: upgrade too aggressively and customer integrations break; lag too far behind and customers lose vendor support for the GIS stack their permitting and service-request workflows depend on.

There was no systematic answer. Compatibility questions were handled reactively — a customer hit an issue, a ticket got escalated, and someone reverse-engineered which version combination was at fault. Every customer was a special case, and nobody could tell an agency GIS manager when it was safe to upgrade.

I started with the unglamorous part: building the compatibility picture. Which platform versions, which integration endpoints, which customer configurations — mapped against Esri's release and support timelines.

From that matrix I sequenced upgrade waves: which customer configurations could move when, what had to be validated first, and what the platform itself needed to support next. I wrote the customer communication plan so agencies heard about compatibility windows before they planned upgrades, not after something broke, and aligned engineering and services on a single rollout order instead of ad-hoc exceptions.

The deliberate cost of the program was speed on new capability. Sequencing for stability meant customers on older configurations waited longer for new Esri features. I made that tradeoff explicit rather than letting it happen silently: in enterprise government software, stability buys the trust that features spend.

That framing also changed the roadmap conversation. Once dependencies were visible in one place, sequencing conflicts surfaced before they cost time — and the compatibility matrix, not the wish list, became the starting point for GIS roadmap decisions.

The program replaced emergency escalations with a predictable compatibility posture customers could plan around, and it became the standing playbook for how the platform tracked Esri's cadence.

What I took away: in platform products, the roadmap is downstream of the dependency matrix. The highest-leverage strategy work I did here looked like maintenance from the outside — but it made every subsequent decision cheaper, faster, and defensible to both engineers and agency GIS managers.