Doors per employee going from 104 to 180 on automation. Do renewals survive?

Triage is where I'd expect it to break and mostly it hasn't, with one condition. If the intake bot's job is to ask the four questions a dispatcher would ask and attach photos, it does that better than a coordinator at 7pm, because it never skips the questions. If its job is to decide whether that's a supply line or a drain, it will send me to a job that needed a different truck. The two wrong dispatches I got last quarter both came from a system trying to classify rather than collect.

That slide looks to me like 104 to 180 is real for the tasks and fantasy for the accountability, at least after five years running a service business. The hours you save are the answering and the typing. The hours you don't save are the judgment calls, the owner who wants to argue about a $900 invoice, and the tenant situation where somebody has to decide something and own it. Those don't scale, and at 180 doors a head they land on fewer people who are already at capacity. What snaps back is usually a person leaving, not a metric moving.

@solder that split between collecting and classifying is the clearest way I've heard it put.

Screening is the one I'd be most nervous about automating and I notice it's on almost every one of these slides. If a model is scoring applicants, who explains the decline reason when it's challenged? Isn't that exactly where an automated output becomes someone's legal problem?

Numbers from a small shop, 214 doors, four full time plus me. We were at 61 doors per head two years ago and we're at 71 now. That's it. Real, worth having, nowhere near the deck.

What moved it: automated rent reminders and delinquency sequencing, application intake and document collection, work order intake with photo requirements, and monthly owner statement drafting. What didn't move it at all: renewals, anything involving a deposit dispute, and vendor scheduling in a market where my good plumber doesn't read email. Anyone claiming 180 has either much larger buildings, one owner per fifty doors instead of one owner per two, or they've reclassified a call center in another country as not being headcount.

Question I've been sitting on while I finish licensing coursework. If an automated system is answering leasing inquiries, quoting rent, and discussing lease terms with a prospect, is that activity that needs a licensed person behind it? I've read the answer depends on the state and on whether it's a percentage-of-rent management arrangement, and I can't get a straight answer from my course material.

@halyard licensing triggers for property management activity vary state by state, and some states require a real estate license to lease and collect rent for an owner while others have a separate registration or an exemption for on-site staff, so that's one to confirm with your state commission in writing rather than from a course book. Practically, every shop I know keeps a licensed person named as the responsible party regardless of what the software sends.

@thatch I'd push back a little. Renewals absorbed automation for us better than anything else. Sequenced outreach starting 120 days out with a rate already calculated moved our renewal rate about five points, because the failure mode before was a coordinator forgetting until day 45. The automation isn't persuading anyone. It's making sure the conversation happens on time, and half of renewal loss is timing.

@beacon on screening, we automate the collection and the verification steps and a human signs the adverse decision. Nobody I'd trust lets a model issue declines unreviewed.

The part of that deck I'd want to see is the record retention. Automated communication generates an enormous volume of written statements to tenants, all discoverable, none reviewed. A shop at 180 doors a head has more written output and less oversight of it than a shop at 104, and if any of it contradicts the lease or a required notice, the efficiency slide never mentions the cost. I'd ask the seller for their complaint log and whatever their counsel has said about the templates.

Underwriting a management company as a borrower, the doors per employee line is the one I discount hardest. What I'd want instead is revenue per door split into base fee and ancillary, trailing 24 month owner retention by cohort, and the concentration of doors under the top five owner accounts. Efficiency gains that came from automation stay when a key person leaves. Efficiency gains that came from one heroic operations manager do not, and the deck can't tell them apart.

I own land and a handful of houses and I've had two managers in twelve years, so take this as an outsider's note. Both times I left, it was because a person stopped answering. Nothing about that problem sounds like it improves when there are fewer people.

Working @cobalt's numbers against the deck. 214 doors at 71 per head is five people. The same portfolio at 180 per head is 1.2 people. That isn't an efficiency claim, it's a different business, probably large multifamily or single-owner institutional portfolios where one relationship covers hundreds of doors. @vellum if the target portfolios are scattered single family with one or two doors per owner, the owner communication load alone caps you somewhere well under 100 regardless of what the software does, because owners aren't the party the automation is aimed at.