AppFolio's AI Is for Apartment Managers. Here's Why That Doesn't Help NNN Landlords.

*AppFolio connected Realm-X to Claude at NAA Apartmentalize last week. The demo looked impressive. It's also entirely built for apartment operators. If you manage NNN retail properties, here's what that announcement actually means for you.*

---

Last week at NAA Apartmentalize in Las Vegas, AppFolio demonstrated Realm-X — their AI agent platform, now connected to Anthropic's Claude. The use cases on stage: leasing performance analysis, marketing copy for vacant units, occupancy trends, maintenance workflow automation. Real capabilities. Real time savings.

For an apartment operator managing 500 units across several multifamily properties, that's a meaningful upgrade. I don't doubt that.

I'm not an apartment operator. I manage NNN retail properties — strip centers, single-tenant net leases, a ground lease or two. I've watched AppFolio closely for years because it's one of the most-recommended platforms in property management and I get asked about it constantly. Here's what I took away from the Realm-X announcement.

The Operational Problem Is Different at the Foundation

Apartment software is built around a specific set of problems: tenant screening, placing residents, tracking maintenance across dozens of units, renewing 12-month leases, collecting monthly rent. The complexity scales with unit count. The lease terms are short. The work is repetitive in a way that AI handles well.

NNN is structurally different. I might have six tenants across three properties. Each lease runs 10 to 15 years. The complexity isn't in the number of tenants — it's in the depth of each lease and what I have to calculate from it every single year.

What I actually spend time on:

• CAM pool calculation. Every year I calculate what each tenant owes toward shared operating expenses — parking lot maintenance, landscaping, property insurance, management fees. That calculation depends on their pro-rata share of leasable area, which expenses are recoverable under their specific lease, and whether they have a cap on controllable costs. Those caps differ by tenant, by year, and sometimes by expense category.

• Gross-up provisions. If the property is below stabilized occupancy, many NNN leases let me gross up recoverable expenses — calculate what costs would be at full occupancy so no single tenant absorbs charges that vacant spaces should share. Every lease handles this differently.

• Year-end reconciliation. Each January through March I send every tenant a statement comparing what they paid in monthly CAM estimates against actual expenses. Underpayment means they owe me the difference. Overpayment means I owe them a credit. Tenants dispute these calculations. I need an audit trail that holds up when someone's lawyer calls.

None of that is in Realm-X. It isn't supposed to be — this is not what AppFolio was designed to do.

What Realm-X's AI Is Actually Doing

Realm-X is an AI layer acting on AppFolio's underlying data model. That model was designed around residential leases: tenant name, unit number, rent amount, lease start and end dates. The workflows the AI can act on are the workflows AppFolio has always supported — leasing, maintenance, resident communication, marketing.

That's not a criticism. AppFolio manages over nine million residential units. Their platform decisions reflect that reality.

The limitation for NNN landlords isn't the AI itself. It's what the AI has to work with. If the software doesn't natively model CAM pools by lease structure, gross-up provisions, or controllable expense caps, adding AI on top doesn't fill that gap. An AI agent can only act on data that's structured and stored correctly in the platform. It can search through notes fields. It cannot agentically run a CAM reconciliation from a notes field.

This isn't speculative. A senior accountant reviewing Yardi Breeze — a platform with more commercial support than AppFolio — described the problem plainly: *"CAM reconciliation is very basic and will only work for small mom and pop leases. Large corporate leases with controllables, caps, and base years require so much manipulation it's easier to do manually in Excel."*

That's the core issue. It's not bad software. It's software whose structure doesn't match the operational reality of NNN leases. AI doesn't change the structure.

What to Actually Ask When Evaluating Software

If you're managing NNN properties and reassessing your software stack — whether because of the Realm-X announcement or otherwise — the right questions aren't about AI capabilities. They're about the underlying lease model:

• Can the software store CAM pool structures with recovery categories and expense caps per tenant, per lease?
• Does it handle gross-up calculations natively, or do you recreate that in a spreadsheet every year?
• Can it maintain a reconciliation audit trail that shows what was billed, what was disputed, and how it was resolved?
• Does it track rent escalation schedules with effective dates — not just the current rent amount — without requiring manual field updates each year?

If any of those require spreadsheets running alongside the software, the software isn't doing the job. The AI on top won't change that.

AppFolio is excellent software for what it does. What it does is apartment management, and Realm-X makes it better at that. NNN landlords are a different operational problem that requires a different foundation underneath — regardless of what AI layer sits on top.

---

I built PigJet because I couldn't find software that handled NNN CAM reconciliation without extensive manual workarounds. If you're managing triple-net properties and evaluating your options, see how it handles CAM reconciliation and year-end reconciliation statements.