The AI Pitch Is Real. The Problem Might Not Be Yours.

I got the same pitch three times in 30 days.

Different vendors, different demos, different words. All three made the same claim: your facility is leaving money on the table because it is not smart enough yet.

The demos are good. AI answering after-hours calls, flagging delinquency before the curve spikes, adjusting pricing in real time based on competitor moves. You watch it and think: I can see how that earns its cost.

Then you get home and realize you cannot remember if you checked your delinquency report this week. Or whether your missed-call rate is 10% or 40%. Or whether the last three move-ins came from Google or the listing service you are paying $200 a month for.

None of that is an AI problem. It is a paying-attention problem.

Self-storage is a simple business. Someone calls, they rent a box, and hopefully they pay on autopay. That is the whole loop.

Almost everything that goes wrong is one of those three things failing: a call nobody answered, a unit that never rented, a payment that did not process. The data telling you which one broke is already in your FMS.

At Capital One I watched engineers build elaborate systems for problems that did not need them. We called it resume-driven development. The sophistication was the point, and the system existed to justify itself.

That is what is happening to operators right now. The vendor is the one padding the resume, and your facility is the line item that pays for it.

A lead comes in on Instagram. Another on Messenger. Three more on SMS.

Your team switches tabs, repeats answers, and loses context while hot leads wait hours for replies. At 2am, nobody responds at all.

That’s not a people problem. It’s a process problem.

Wati brings Instagram DM, Facebook Messenger, TikTok, WhatsApp, SMS, RCS, and web chat into one AI-powered inbox. Automations instantly respond, qualify leads, and route conversations to the right person, 24/7.

Your team stops firefighting. Your leads stop waiting. Your pipeline starts moving.

See How Wati Works

The vendor will not say "API" or "MCP" to your face. They say "always on," "AI-powered," "80% of your interactions handled automatically." Strip the marketing off and every one of these products is doing one of two things.

Moving your data around. Pulling inventory, gate codes, and contact info out of your FMS and showing it to a customer in a chat window. That is an API, and it is plumbing.

Deciding what to do with that data. Answering a question nobody scripted, flagging a tenant for collections, choosing what to say at 11pm. That is the AI layer, and MCP is just how it reaches into your systems to act.

The plumbing is cheap and it mostly works. The judgment is the expensive part, and it is only worth buying when there is more judgment happening than you can personally cover.

So when a demo dazzles you, the acronym under the hood does not matter. The question is which of those two things the product is doing, and whether that thing is a constraint you actually have.

Most of what gets sold into this industry sits on top of a problem your phone log and delinquency report already surface. The data is in your FMS right now. The real question is whether you are using the reports you already have.

The need does not scale with revenue. It scales with how many decisions per week you cannot personally see.

One facility. You are the system. You answer the phone, see every late payment, and know when something feels off before the report surfaces it.

There is no signal in that building that does not eventually reach you. AI here is a solution looking for a problem.

Fix the plumbing first: autopay adoption above 80%, every call going to a number that gets answered, one clean report you trust and look at weekly. When those three are working, you will know exactly what constraint is left. It is probably not one that requires a monthly SaaS subscription.

Two to five facilities. This is where the seam appears. You cannot personally see every signal anymore, but you still own every decision.

A call that went to voicemail at your second location is invisible to you unless something surfaces it. A delinquency spike at your third location is sitting in a dashboard you checked four days ago.

The gap between what is happening and what you know is happening starts to cost real money. This is where routing and alerting start earning their cost. Not AI in the full sense, just automated checks that pull the numbers you would have pulled manually and push them to you before you had to ask.

We operate three facilities in Aiken and this is exactly where we live. The value is not intelligence. The value is not having to remember to look.

Six or more. Judgment has to get delegated to something, people or software. There is no version of this where you personally review every tenant interaction, every pricing signal, every after-hours inquiry.

The question stops being whether to buy and starts being what to buy: which decisions are safe to automate and which still need a human in the loop.

Before you sign, ask what it replaces. If the answer is a report you already generate but do not read, you are paying for a more expensive version of something you already have. If the answer is a decision you currently cannot make because the signal never reaches you, that is worth buying at almost any price.

Here is the part no vendor puts in the demo. The AI is only as good as the data you hand it.

You already know this if you have ever onboarded one. Somewhere in week one, the tool made you fill out a long questionnaire: your gate hours, your lien timeline, your unit types, your late-fee rules. That was the vendor discovering that all of your operating knowledge lived in your head and across four systems, not in one place a machine could read.

The scattered data is the real constraint, not the model. Before you pay anyone to add intelligence on top of your operation, find out whether your operation is even legible to it.

Paste this into ChatGPT or Claude before your next vendor call.

The output is your readiness check. If the AI needs ten things and eight of them live in one person's memory or differ at every location, the product will underperform no matter how good the demo looked. Fix the source of truth first, then let the intelligence sit on top of it.

Works in ChatGPT or Claude.

What have you been pitched in the last 90 days? And did you buy it?

Hit reply. I read every one.

If you seek tranquility, do less. Or (more accurately) do what's essential — what reason requires of a social being. Do it as nature demands. Which means honestly, and with all your heart. Both ways bring satisfaction: doing less, better.

Marcus Aurelius