Property Management Software Problems: The 7 That Cost You Owners, and How to Avoid Them

Why does "replace AppFolio" end up in the scope so often?

It gets in because the frustration is real and the diagnosis is wrong. Everyone in the office says the same sentence, which is that AppFolio does not do what we need, and the natural conclusion is to get rid of it. But sit with what actually hurts and it is never the rent ledger. It is that the maintenance line rings forty to sixty times a day and almost none of what happens next lives in the platform. It lives in a shared inbox, a Google Sheet, and the memory of a coordinator who knows which plumber answers on a Saturday.

This is specific to property management because the platform is genuinely good at one half of the job. AppFolio and Buildium do rent collection, trust accounting and owner statements properly, and those are the parts your regulator recognises and your auditor expects. What they cannot do is encode your rules, because your rules live in three hundred different management agreements. That distinction gets lost in a scoping meeting where everyone is describing a feeling rather than a workflow.

The fix is a rule you set before you talk to anyone: in phase one, the ledger stays where it is. Write it into the request for proposal. Then ask every developer where trust accounting should live, and treat any enthusiasm for rebuilding three-way reconciliation on day one as a disqualification. Full replacement is a conversation for past several thousand doors, and only after an operational layer has run for a year.

What goes wrong when you pull maintenance and owner history out of AppFolio or Buildium?

The unpleasant surprise is that the historical record you want mostly is not structured. Work order history exists, but the part that matters, meaning what was actually wrong, what the vendor did, what the owner approved and when, is in note fields, email threads and a spreadsheet. Importing it produces rows that look complete and answer nothing. Teams then lose weeks trying to reconstruct approvals that were only ever text messages.

The second problem is vendor and owner identity. The same roofing company appears three times because someone typed it differently, and the same owner holds properties under two entity names. Land that unreconciled and your first spend-per-door report is wrong, an owner spots it, and trust in the new system is gone in week one.

What works: import open work orders, current vendor records with their insurance and tax documents, owner and property records with the approval thresholds transcribed from the management agreements, and roughly twelve to eighteen months of completed work orders for context. Everything older stays queryable in the platform where it already sits. Then plan for a real reconciliation pass on vendors and owners before go-live, run by your office manager rather than by the developer, because she is the only person who knows that two of those roofers are the same firm.

Why do the AppFolio, Buildium and accounting integrations break after launch?

Because data access here is tier-dependent and quietly asymmetric. Buildium exposes an open interface covering most core objects. AppFolio restricts programmatic access depending on which tier you are on, so builds there frequently combine whatever interface access exists with scheduled report exports landed in a warehouse you own. That arrangement works, and it also has more moving parts than anyone budgets for.

The failures are predictable. A report export changes column order after a platform update and a nightly job silently imports garbage. A work order syncs back with a status the ledger does not recognise and lands in limbo. A vendor payment gets re-issued and the sync treats it as a second payment, so a property shows double the spend and an owner queries the statement.

Three questions to ask before signing. Which objects does my tier expose, by name. What syncs hourly versus nightly, and what happens to the nightly job when the export shape changes. And what does the system do when the same record arrives twice. The right answers involve validation on import that fails loudly rather than importing partial data, handlers that are safe to re-run, and an exceptions queue with a named owner. Vague answers here become six-figure surprises later, and this is the most common single cause of a property management build losing its users.

What happens when management agreement terms and vendor insurance are not modelled?

These are the two compliance gaps that turn a workflow inconvenience into a liability.

The first is the not-to-exceed threshold. A tenant reports a leaking water heater, the repair quote comes in at $850, and that owner's agreement caps unapproved work at $500. If the threshold is not stored as data on the owner record, the check depends on a coordinator remembering, and coordinators remember correctly right up until the day they do not. Then the work happens, the owner disputes the invoice weeks later, and you either eat it or start a conversation that ends with them moving twenty doors elsewhere. At an eight percent fee on $1,800 rents, twenty doors is roughly $34,000 a year in management fees walking out over a reporting and approvals problem.

The second is vendor insurance. The certificate is a document in a folder and the expiry date is a field nobody reads. You dispatch a roofer whose general liability lapsed in March and you find out during the injury claim. The fix is one line of logic and it is worth an entire module: a hard block that makes it impossible to dispatch to a vendor with lapsed cover, plus automated renewal requests thirty days before expiry and lien waiver collection above a dollar threshold. Off-the-shelf tools will store the date. They will not enforce it, and enforcement is the whole point.

Should you build custom or configure what you already own?

Under roughly five hundred doors, stay on the shelf and mean it. AppFolio's entry pricing lists at $1.40 per unit per month with a $298 monthly minimum, and Buildium's top tier lists around $375 a month, so even an imperfect fit costs less than a build by a wide margin. Before you spend anything on custom work, genuinely try the higher tier and the maintenance contact centre add-on for six months. Both platforms answer the phone competently, and if your after-hours problem is mostly volume rather than rules, that may be the whole answer.

The signals that configuration has run out are operational rather than emotional. You hire another maintenance coordinator for every four hundred doors and margins do not improve. Spreadsheets have become the actual system of record. Owners are leaving over visibility rather than over performance. And your management agreements contain terms no configuration screen can express, which is the decisive one, because per-owner approval thresholds and per-building vendor rules are not a setting anywhere.

Around eight hundred doors is where the maths usually flips. Even then the answer is a layer, not a replacement: keep the platform as the ledger of record, build the operating system around it.

How do hidden costs get into the quote?

Five ways, and four of them are avoidable by being explicit in the brief.

Payments and trust accounting, if anyone lets them creep in. Say in writing that they are out of scope for phase one.

Voice intake instead of guided text and web intake. Voice sounds better in a demonstration and costs considerably more to build and tune. Most after-hours volume resolves fine through a guided flow with photo capture, with true emergencies escalating to the on-call rotation with full context attached.

Subsidised housing workflows. Inspection and recertification processes for Housing and Urban Development or Section 8 units are a distinct body of rules, and they belong in the brief as a named item or they will arrive as a change request.

Multiple markets. Different states mean different notice periods, different security deposit handling and different rule sets, which multiplies the logic rather than adding to it.

Two-way synchronisation. Every system that must write back as well as read costs more than one that only reads, because now you need conflict rules. For anchoring, our bands: a focused first release covering guided intake, rules-based triage, dispatch and owner approvals runs $40,000 to $90,000 in ten to fourteen weeks, and a fuller platform adding the turn pipeline, owner portal, analytics and vendor compliance runs $100,000 to $250,000 over six to nine months.

What separates a build that works from one that fails here?

The coordinator is in the room from the start. If the proposal you receive does not mention not-to-exceed thresholds, vendor rotation or the after-hours flow, the developer never interviewed the person who runs the operation and has written you a generic ticketing system with your logo on it. Require that interview before quoting, not after.

The first release is one workflow, shipped inside four months. Maintenance intake through to owner approval is the right one, because it is where the daily pain is and it produces a number you can point at. The turn pipeline and the owner portal follow, once coordinators trust the system enough to stop keeping a parallel sheet.

Someone writes down the baseline before anything is built: doors per coordinator, average days vacant on a turn, and how many owner emails a property manager answers with screenshots each month. Those three numbers are the return, and without them the project gets judged on how modern it looks.

And the contract settles ownership before kickoff. You own the source code, the database and the cloud accounts registered in your company's name, with assignment written in. That is what makes the build an asset that transfers when you sell the company, rather than a subscription you did not know you were signing.