Municipality Software: Custom Build or Buy Accela and Tyler EnerGov

What Accela, Tyler EnerGov and OpenGov actually do well

Accela and Tyler EnerGov are serious permitting platforms with two decades of municipal practice inside them. They hold records, inspections, licences and code cases properly, and they will not lose your data. OpenGov ViewPoint and CityView are lighter, cleaner and priced for smaller cities. Clariti is a credible newer entrant. On the systems of record side, Esri ArcGIS is the authoritative parcel platform and nothing you build should replace it. Laserfiche does document management and retention well, Granicus handles agendas and public meetings, Cityworks and Cartegraph run public works, SeeClickFix and QAlert take resident reports, and GovQA and NextRequest track public records requests competently.

Now the sentence that costs us business. Most cities should buy, and a specific class of city should buy and then leave it alone. Under roughly 25,000 residents, under about 1,500 permits a year, no full-time information technology staff, no unusual overlay districts or heavily amended code: take ViewPoint or CityView, adopt their process as written, resist customising, and you will be fine. The expensive mistake in this category is not buying the wrong platform, it is buying an enterprise platform and then paying professional services to make it behave like your city.

Buy without building if most of these hold:

• Your permit volume fits comfortably in one product's standard workflow.
• Your fee schedule is a table rather than a set of formulas with tiers.
• One or two departments touch a typical permit, not six.
• Nobody at the city maintains a spreadsheet that a permit could not be issued without.
• Your last two years of vendor spend was licence, not change orders.

Where they stop: routing you cannot edit and fees in a spreadsheet

It is 9:40 on a Tuesday at the community development counter. A contractor wants to convert a retail bay into a restaurant. The permit tech opens the permitting system to start the record, switches to ArcGIS to confirm the parcel and the overlay zone, opens the spreadsheet a senior planner maintains to compute plan check fee from the valuation table, emails the Fire Marshal because Fire review is not in the workflow, and tells the contractor to come back Thursday because the grease interceptor question belongs to the sanitary district.

Both platforms technically support parallel routing. The problem is who can change it. Adding Fire as a conditional reviewer when occupancy exceeds a threshold, or making Engineering parallel instead of serial, is a scripted workflow change, which in practice means a scoped change order and a wait measured in months. So departments route around the software using email, and the record stops being the truth.

The fee schedule is the sharper example, and it is where custom versus off the shelf genuinely separates. Council adopts a fee resolution. It changes the building valuation multiplier, adds a park in-lieu tier and touches three impact fees. The fee logic inside your permitting system is vendor scripted, so the practical fix is that a plans examiner keeps the real arithmetic in a spreadsheet and types the answer in as a manual fee. Your fees are now correct in a file nobody backs up, and unauditable when the annual audit asks how a specific charge was derived. No state-wide product can generalise this, because your fee formulas are your ordinances.

The third break is the address itself. The assessor has a parcel number, ArcGIS has a situs address, utility billing has a service address, and the permit system has whatever the applicant typed. When a lot line adjustment creates two parcels or a subdivision drops sixty new addresses, permitting finds out when somebody notices. So permits attach to the wrong parcel, code cases lose their history, and nobody can answer what has ever happened at this address without three searches and a guess.

The arithmetic: per-permit cost versus the cost to build

Put both sides on a per-permit basis, because that is the number a city manager can defend to council.

Add your annual licence, your hosting or maintenance line, and every professional services invoice from the last two years divided by two. Divide by permits issued annually. A city issuing 3,000 permits a year on a fully loaded $150,000 is paying $50 a permit before a single staff hour. At 1,200 permits it is $125 a permit. Run yours, and include the change orders, because in this category they are frequently the larger half.

Now the build. A focused first release runs $60,000 to $130,000. At the midpoint amortised over five years with year two support, call it $28,000 to $48,000 a year. At 3,000 permits that is roughly $12 a permit, and the marginal cost of the fourth permit type is close to zero once the model exists.

The lines cross near 2,500 to 3,000 permits a year, or roughly 60,000 residents. But the honest signal is simpler than the arithmetic: if your workflow changes require a vendor change order, and you have spent more on change orders and professional services than on licence over two years, you are already paying more than a build costs and receiving less.

Add the staff side before you decide. In the intake work we have done for community development departments, the counter step on a commercial application runs 22 to 35 minutes across three systems, and roughly a third of applications get sent back for something the city could have checked in ninety seconds.

Cost to build a custom city platform

A focused first release covering an address and parcel service reading your ArcGIS layer, one permit type end to end with intake, routing, fees, review and issuance, online payment and 311 intake with status runs $60,000 to $130,000 over 12 to 16 weeks. That is enough to show council a cycle time number inside one budget cycle. A full platform covering multiple permit and licence types, inspections with offline field capture, code enforcement, records and payments runs $150,000 to $400,000 phased over 6 to 12 months.

The two lines nobody quotes:

• Data migration: 10 to 25 percent of the build, and often the ugliest line item. Twenty years of permits carry addresses that never matched a parcel in the first place. Budget it as its own workstream with a cleanup decision from your staff rather than as a task inside development, and expect your permit techs to spend real hours on judgements only they can make.
• Year two: 15 to 20 percent of build cost annually. A fee resolution every year, code amendments, a new licence type, and accessibility retesting. Under the Department of Justice rule implementing Title II of the Americans with Disabilities Act, WCAG 2.1 Level AA is the technical standard for state and local government web content and mobile applications, with compliance dates staged by population. Building to it from the start adds roughly 10 to 15 percent to front-end work. Retrofitting costs about three times that.

What pushes price up here: the number of departments in a single workflow, since each is a stakeholder with its own conditions; integration with a legacy financial system that has no interface, where a nightly file exchange plus reconciliation is real engineering; payment card scope, which you contain by tokenising through Point and Pay or Invoice Cloud rather than touching card data; hosting posture if your insurer or state wants StateRAMP or SOC 2; and multilingual resident screens.

The four situations where building wins

• Regulatory fit. Fees have to be defensible. Each fee needs a formula, its inputs pulled from the record, an ordinance citation and an effective date, with every calculation storing the inputs and the formula version so a fee assessed two years ago recomputes exactly as it did then. That single property ends fee disputes and audit findings, and no packaged product will give it to you.
• Scale economics. Past roughly 3,000 permits a year, per-permit platform and services cost exceeds the amortised build, and each additional permit type on a custom model costs almost nothing.
• A workflow that is your competitive advantage. For a city, the competition is the next jurisdiction over. Cycle time is why a developer builds here rather than there, and routing rules your Building Official can edit in an admin screen, rather than request as a change order, are the difference between a six week resubmittal loop and a two week one.
• Integration sprawl across three or more systems. Permitting, the general ledger, work orders, resident reporting, records and payments. Every seam is a person: the clerk re-keys, the inspector re-keys at 4pm from paper cards, the finance analyst reconciles a nightly file. Those staff hours never appear in a licence comparison and they are the real cost.

How to decide in a week

Five days, no consultant. Monday, take one commercial tenant improvement permit issued last year and reconstruct every fee from the ordinance that applied on the application date. Tuesday, time the counter step on the next three commercial applications, counting screens and systems touched. Wednesday, ask your Building Official how long it would take, and what it would cost, to add Fire as a conditional reviewer above a given occupancy. Thursday, pick a parcel that was split in the last three years and ask what has ever happened at that address. Friday, total your vendor professional services spend over two years and set it beside your licence spend.

Wednesday is usually decisive. If the answer is an afternoon in an admin screen, keep your platform. If it is a change order and a quarter, you have found the reason your process lives in email.

Then buy a paid discovery phase before any build. At Digital Heroes that produces a signed product requirements document before code, covering the domain model of parcel, address, record, review, condition, fee, inspection and case, the routing configuration, permissions and acceptance criteria. You own it whether we build it or not, and you can take it to every firm on your shortlist so the quotes describe one system.

We are the wrong firm for you if you want a body shop billing hourly, if you want your general ledger or your parcel platform rebuilt, or if you are a small city that should adopt a packaged process as written. We are more than fifty specialists across India LLP, US LLC and UK LTD entities, so your intellectual property assigns under your own law, and you meet the named team before signing rather than a bench in month two. Ask us about payment card scope containment and records retention as engineering rather than as change orders. Our record is checkable on Clutch, Trustpilot, Fiverr Vetted Pro and our D-U-N-S listing.

Book a 30-minute call with Digital Heroes and get a written plan and a fixed quote within 48 hours.