Insights into Permitting Platforms

In November 2016, staff in Loudoun County, Virginia, put a price on replacing the county's 20-year-old land management system, the software behind every building permit it issues: 2 years and $6 million. The replacement went live in August 2023, nearly 7 years later. By then its budget had reached $13.3 million, and by December the county had hired extra staff to clear a backlog of applications the launch created.

Loudoun's experience is a common one, and so is the decision behind it. About 1 in 10 of the thousands of US cities and counties Builty tracks replaced their permitting platform in the past 5 years, in 2 out of 3 states (counting DC). Half of the switches we could date happened in 2024 or 2025.

Few local government technology projects carry more risk. A permitting platform encodes a jurisdiction's own rules: every permit type, fee formula, review route, and inspection checklist, refined over years, along with a record history that buyers, lenders, and code officials still rely on. Moving all of that into a new product takes years and millions of dollars, so no jurisdiction does it lightly. That so many have says the systems they had were failing them, on age, cost, security, or what the public could see.

What a permitting platform does

Vendors sell permitting platforms as land management, community development, or enterprise permitting and licensing software. A full platform covers the life of a permit: intake, plan review routed to each discipline, fee calculation and payment, inspection scheduling and results, and the final inspection or certificate of occupancy. Most of the larger products also hold planning entitlements, code enforcement cases, and business licenses in the same database, which is what lets a planner see a parcel's open violations or an inspector see its permit history. Staff work in a back-office application, and applicants, contractors, and the public use a portal built on top of it.

Platforms also differ in where they run. On-premise products sit on the jurisdiction's own servers and depend on its IT staff for upgrades and security. Software-as-a-service (SaaS) products run in the vendor's cloud for an annual subscription, and several large jurisdictions have moved from an on-premise install to the hosted version of the same product, among them Hillsborough County, Florida, and Los Angeles County. The rest are homegrown: systems built in-house, sometimes decades ago, or no online system at all.

Why it matters to residents

For residents, the permitting platform is the front door to the building department, and how well it works shows up in their own projects.

• Turnaround. A new system's first months often slow reviews down. In San Francisco, a supervisor said fire and sprinkler permits that used to take 2 days took 8 or 9 after the city's new OpenGov system arrived, and Loudoun County, Fairfax County, and Baltimore all reported application backlogs after go-live. For a homeowner or a builder, each week of delay carries rent, interest, and crew costs.
• Due diligence. The public portal is where a buyer checks whether an addition was permitted and passed final inspection, where a homeowner vets a contractor's permit history, and where neighbors see what is planned next door. How much of the record is public, including inspection results and plan review status, depends on the platform and on the jurisdiction's own configuration, as the scorecard below shows.
• Public money. These projects are paid for from general funds, permit fees, and grants: Honolulu covered the first 3 years of its new system with federal funds, and part of Baltimore's came from a grant. The larger projects run to tens of millions of dollars.
• Payment data. Residents pay permit fees, and often utility bills, through these portals, and card data has been stolen from them before, as described below.

Who is gaining and who is losing

The switching is consolidating the market at the top. The biggest winners were Tyler Technologies' EnerGov, which grew about 20%, and Accela's Civic Platform, which grew about 18%; between them they took the largest share of switching jurisdictions. Smaller cloud vendors grew faster from small bases: iWorQ by about 21%, and GovWell, founded in 2023 and backed by a $25 million funding round in May 2026, by about 45%. GovWell gained jurisdictions without losing any according to our analysis, and neither did the Michigan-based BS&A or Cloudpermit.

The losses fall on products near the end of their life. CentralSquare's Click2Gov lost nearly half of the jurisdictions that ran it, and eTRAKiT lost a quarter. About a third of the jurisdictions on homegrown systems replaced them, most often with Accela or EnerGov. CitizenServe and OpenGov each lost about 1 in 10, though OpenGov also won new ones.

Where switching jurisdictions went. The platform each one left is on the left, and the platform it moved to is on the right, for the 12 platforms with the most switches, sorted by their share of arrivals. Each band is a group of jurisdictions that made the same move, colored by the vendor it moved to. Percentages are each platform's share of all switches. Homegrown means a jurisdiction's own system, or no online system.

The biggest winners and losers. The 5 platforms that gained the most jurisdictions over the past 5 years, and the 5 that lost the most. Light orange bars are jurisdictions lost, and bright orange bars are jurisdictions gained.

Why cities switch

Most switches start with a system that has reached the end of its life. Des Moines replaced Tidemark, which it had run since 2001, because it was "end of life, no longer supported." San Diego's permitting ran on a homegrown Project Tracking System developed in the 1990s until the city replaced it with Accela. Many of the jurisdictions leaving eTRAKiT and Click2Gov are moving off products of the same generation.

Security is another reason, and no vendor has been immune:

• Attackers took more than 300,000 payment card records from cities' Click2Gov payment portals in 2017 and 2018, and a second wave in 2019 hit 8 more cities. Some of the affected cities moved to other vendors.
• In November 2025, researchers disclosed 3 vulnerabilities in eTRAKiT and its phone inspection line through CERT/CC, including a SQL injection that could expose the database and a bypass of administrator sign-in (CVE-2025-64280, CVE-2025-59491, and CVE-2025-64281). These were disclosed vulnerabilities, not a reported breach.
• Accela reported to the California Attorney General that an unauthorized party copied files from one of its secure file-transfer portals on December 11 and 12, 2025.
• Tyler Technologies disclosed a ransomware attack on its internal systems in September 2020, and said it had detected no compromise of the client systems it hosts.

Ownership has changed across the market, too. Private equity firms Bain Capital and Vista Equity Partners merged Superion, the maker of eTRAKiT and Click2Gov, into CentralSquare in 2018. OpenGov bought ViewPoint Cloud in 2019, and Cox Enterprises took majority ownership of OpenGov in 2024. Berkshire Partners acquired Accela in 2017, and Francisco Partners invested later. A new owner can bring new priorities, new pricing, and product retirements.

The training burden

The large platforms are built to be configured to almost any process, and that flexibility moves work onto the jurisdiction. Someone on staff has to know how every permit type, workflow, fee formula, and inspection checklist is configured, and often how to script the exceptions. Tyler's standard quotes assume the jurisdiction's own staff put in 70% of the implementation hours. The vendors sell the expertise back as training: Accela's 2025 course list for administration and scripting came to about $3,600 and 70 hours per person, and Tyler priced its EnerGov administrator track at 10% of a city's yearly maintenance fee, with a $5,000 minimum. BS&A, CentralSquare, and OpenGov run their own courses and certifications, and the largest vendors hold a user conference every year.

Larger jurisdictions hire for it. Job postings for Accela and EnerGov administrators list top salaries above $100,000, and some ask for ASP.NET, JavaScript, SQL, and Python. Others buy the expertise from consultants: Berkeley's contract for help configuring and scripting its system reached more than $1 million over 9 years, and Placer County keeps a consultant on call because its vendor contract "does not include specialized assistance or configuration."

Some vendors take the opposite approach. iWorQ includes training in the subscription, and Granicus lists unlimited access to its learning system as included with SmartGov. Newer platforms pitch configuration staff can do themselves: Clariti says permit types can be added "using clicks, without IT support," GovWell lists jurisdictions that went live in 3 to 6 weeks, and Cloudpermit offers an AI assistant to staff and applicants. AI is already in plan review, too: California and Los Angeles deployed Archistar's AI plan check for wildfire rebuilds in 2025. With AI assistants able to walk staff through a configuration screen, a multi-day course for each module is getting harder to justify. Course prices, job postings, and consulting contracts for each vendor are in the appendix.

What a new system costs

Costs run from a few thousand dollars to more than $100 million. iWorQ quoted a Utah town of under 2,000 people $2,000 a year plus $3,350 to set up. Austin approved up to $59.3 million for a new system in September 2026, and New York City's DOB NOW has run to about $140 million across its integration contracts. Most county and big-city projects we found fell between $2 million and $20 million.

Software licenses are usually the smaller part of the bill. In Phoenix's 2019 approval, licenses were $11.6 million of $31.6 million; the rest was implementation and hosting. Beyond the vendor's invoice come systems integrators, project managers, and backfill for the staff pulled onto the project: Loudoun's 2016 estimate set aside money for a contract project manager and backfill staffing alongside the software.

The first estimate is rarely the last. Chesterfield County, Virginia, planned on 19 months and $2 million; its system launched fully more than 5 years after the project started, at $3.8 million. Some projects never launch: Montgomery County, Maryland, terminated its contract in 2023 after spending about $2.1 million, because the new system "provided less functionality than the legacy systems it was replacing." Others go to plan: Dallas switched in May 2025, ahead of its July target, at 4% over its contract.

Planned and actual cost and time, 8 projects. Hollow dots are the plan, and filled dots are what happened. Start is the contract date, except for Loudoun County, where it is the 2016 estimate that set the budget and timeline. Baltimore, Honolulu, Los Angeles County, and Prince William County are shown without a plan because none was found in their documents, and Los Angeles County's point is its first department going live. Costs are budgets or contract ceilings from council and budget documents.

Audits of the troubled projects point to the same causes, and few are about the software. Montgomery County's Inspector General put it directly: "Many enterprise systems fail due to lack of preparation as opposed to software limitations." San Diego's auditor found that the documentation needed to map the old system's processes did not exist, and the only person who fully understood them, the system architect, retired with the blueprints unfinished. Chesterfield County removed its third-party integrator partway through.

Spending continues after go-live, as jurisdictions add departments, modules, and support. Los Angeles County's contract grew from $2.3 million in 2014 to $18.7 million in 2025 as more departments joined. The full list of projects, with sources, is in the appendix.

What a switch takes

A typical replacement runs through 10 stages and takes 2 years or more from the first needs assessment. Procurement often takes most of a year, and configuration and data conversion take most of the rest. Clariti's buyer's guide puts implementation alone at 12 to 18 months, plus a 60-day hypercare period after go-live.

Most of those stages can be shortened with tools and contracts that already exist:

• Procurement. Cooperative purchasing contracts, such as Sourcewell's, can replace a full RFP where state law allows. The scope has to be settled first: Montgomery County's Inspector General found that buying through a contract bridged from another agency left the county unable to negotiate the changes it later needed.
• Configuration. Permit types, fee schedules, and inspection checklists can be generated from the jurisdiction's existing forms, fee ordinances, and adopted codes, then reviewed by staff, instead of rebuilt by hand in a configuration tool.
• Data conversion. This is where many projects stumble: Honolulu listed data migration errors among its launch problems, and Mecklenburg County was still paying to migrate legacy records 13 months after its switch. AI-assisted mapping can propose field matches between the legacy schema and the new data model and flag the records that do not fit, which turns months of spreadsheet work into a review.
• Hosting. A vendor-hosted system takes hardware purchasing, server setup, and upgrade cycles off the plan.
• Switchover. Leaving permits already in progress in the old system, and opening the new one to new applications only, keeps the intake freeze to days.
• Training. Guidance built into the product, and AI assistants, can replace multi-day classes for most staff and applicants.

A permitting system replacement, typical and modern. Typical durations follow published implementation guidance and the projects in this post; the modern lane shows what the approaches above can save. Stage lengths are illustrative.

What the public can see

Cities on the same platform can show the public very different things. In October 2026 we looked at the public portals of 3 to 5 jurisdictions on each of the 12 largest platforms, as an ordinary visitor without an account, and checked 9 things a resident or contractor might look for.

No platform does everything, and a city's own settings matter as much as the vendor's. Accela portals showed the most complete records, with project values, contractor license numbers, inspection results, and reviewer comments, plus a download of search results. EnerGov, eTRAKiT, and Accela usually keep fire permits alongside building permits, while BS&A, iWorQ, and GovWell portals rarely showed them. GovWell was the only platform where every city let any visitor download its permits, and MyGovernmentOnline was the only one that required an account to search.

Key: ● Yes, in most jurisdictions; ◐ In some, or partly; ○ No, or not public; ? Could not be checked.

Sampled October 2, 2026. Each row summarizes 3 to 5 jurisdictions; a single jurisdiction can show more or less than its platform's row.

Questions to ask before you sign

• Data ownership and export. Who owns the records, and can staff export all of them, including history, attachments, and inspection results, in a documented format without a service order?
• API access. Is there a documented API that other software can use to read and write records, and is it included in the subscription?
• Public portal. Which permit types, details, inspections, and documents will the public see, and can the jurisdiction change that itself?
• Setup. Can staff add a permit type, change a fee, or edit a workflow, or does each change need a paid change order?
• Training. How many hours of training does an administrator need, what does it cost, and is it included?
• Modules. Are fire prevention, planning, code enforcement, and business licensing in the same record, or separate modules priced separately?
• Exit terms. At the end of the contract, how is the data returned, in what format, within how many days, and at what cost?
• Price increases. How much can the subscription rise each year, and what happens at renewal?
• Hosting and security. Where is the system hosted, what certifications does the vendor hold, and what incidents has it disclosed?
• References. Which jurisdictions of similar size went live in the past 2 years, and how long did it take them from contract to launch?

Where permitting platforms can improve

The switching in this post follows from the problems above: systems that age out, projects that run for years, and training that has to be bought. SaaS permitting platforms, including those named here, could close most of that gap:

• Full self-service export and a documented API, so a jurisdiction can get its records out without a service order.
• Setup staff can do themselves, for permit types, fees, and workflows, without change orders.
• Help inside the product in place of multi-day courses and certifications.
• Fire, planning, code enforcement, and licensing in a single record, rather than modules bought and set up separately.
• Public portals that show the whole record, including inspections, plan review status, and documents.
• Migration tools that carry history across, so the old system can be retired on day 1 instead of kept alive for lookups.
• Published security disclosures and timelines, so a jurisdiction learns about an incident from its vendor first.

Find out how local municipalities are working with Builty to help with their permit data.

Appendix A: How we counted

We identified the permitting platform each jurisdiction runs from its public-facing systems, and recorded a switch when public evidence showed one: a city announcement, a redirect from the old portal to the new one, or an old portal going dark as a new one opened. We rate each switch as confirmed (an announcement, a redirect, or a dated notice), likely (2 indirect signals), or possible (1). The figures include all 3. A jurisdiction that switched twice counts once, from its first platform to its last. A switch is dated by the city's announced launch, or else by the last permit filed in the old system.

Appendix B: What new systems cost

Figures are contract ceilings, appropriations, or budgets from council, audit, and budget documents unless noted, and are not the same as amounts spent.

Appendix C: Training, certification, and consultants

Every large permitting platform runs its own training operation, and the paid courses concentrate in administration, configuration, and scripting:

• Accela's 2025 catalog listed 8 administrator and scripting courses that together come to about $3,600 and 70 hours per person at list price, from Records and Mobile Administration (20 to 22 hours, $1,200) to 3 levels of scripting. Its Level 1 certifications cost $400 per area and require a year of administration experience. Accela also offers free e-learning videos, and runs regional user groups and an annual conference.
• Tyler sold its 4-day EnerGov System Admin Track as a subscription priced at 10% of the annual maintenance agreement, with a $5,000 minimum. Its standard quotes assume implementation hours split 70% client and 30% Tyler, and cap on-site classes at 12 people. Tyler also hosts Tyler Connect each year.
• CentralSquare runs CentralSquare University and sold add-on classes at its ENGAGE conference, including one on configuring reports for eTRAKiT.
• BS&A trains at a 6,000-square-foot facility in Michigan. Quotes in council packets price classroom days at $205 per person and on-site days at $1,100, with report designer training extra.
• OpenGov offers more than 200 modules and certification through OpenGov University. Computronix's 2022 POSSE proposal to Nebraska planned 80 hours of courses, after which agency IT staff "with some training in programming" would take over system development.

Cities hire for it, too. Rochester, Minnesota, posted an Accela administrator role at $82,338 to $121,086 that asks for ASP.NET, JavaScript, SQL, and Python; Santa Clara, California, an Accela operations analyst at $119,107 to $152,152. Cape Coral, Florida, has an EnerGov application administrator classification at $81,474 to $126,360, and Orange County, North Carolina, posted an Enterprise Permitting and Licensing administrator that can require 3 years of EnerGov administration. Smaller places fold the job into another one: Chaska, Minnesota's senior permit technician administers SmartGov, and an Oceanside, California, analyst administers TRAKiT.

A consulting market has grown up around the same work. Accela's partner directory lists firms for configuration, integration, and data migration, and one of them, Avocette, describes Accela as "a highly configurable software solution that requires users to achieve a high level of expertise." Berkeley's contract with TruePoint Solutions for Accela configuration and scripting reached $1,047,200 between 2015 and 2024, and Palmdale contracted 2,400 hours over 3 years for $396,000, including refresher training. Placer County explained why it keeps one on call: its vendor contract "covers standard maintenance and support, but does not include specialized assistance or configuration needed for Placer's business processes." BS&A customers can buy training from partner MuniVate, which Pontiac, Michigan, hired for up to $37,220 of on-site training in 2024.