Contact Center

Self Storage Inbound Calls: Sorting Caller Intent Before You Measure Conversion

September 28, 2026
5 Minutes

Two managers at the same facility can report call conversion rates four times apart and both be telling the truth.

‍

The difference is rarely the AI, the script, or the team. It is the denominator. One manager divides rentals by every conversation the phone system logged. The other divides rentals by the calls where someone was trying to rent a unit.

‍

Smartlock Self Storage published both versions of their own figure. Counting every touchpoint puts conversion near 4%. Measured against sales calls, roughly 1 in 7 became a rental in May, out of 252 sales calls and 38 rentals booked.

‍

Neither number is wrong. One of them is useful. Getting to the useful one starts with caller intent.

‍

What Caller Intent Means

‍

Caller intent is what the person on the phone came to accomplish. It is not the line they dialed, because a tenant with a past-due balance calls the number on the sign. It is not the agent who answered, because routing is a guess made before anyone speaks. And it is not the outcome, because a call with no buying intent that ends without a lease was never a conversion failure.

‍

Intent is the only one of those that tells you whether a lease was available to win.

‍

Six Types of Caller Intent

‍

Tag a week of calls at any storage facility and the reasons people picked up the phone sort into six groups. The examples are illustrative rather than complete.

‍

1. Buying. Asking what is available, what it costs, which size fits, whether a location has climate control. Confirming a reservation. This is the only group where a lease was on the table.

‍

2. Billing and account. Paying a balance, asking what is owed, updating a card, questioning a charge, asking for more time. These calls protect revenue rather than create it.

‍

3. Access and service. Trouble at the gate, something broken on site, a forgotten unit number. Speed is the whole measure here. A tenant standing at a keypad has a different tolerance for hold time than a shopper comparing facilities.

‍

4. Lifecycle. Giving notice, transferring units, asking how move-out works. Cancellation intent deserves separate attention, because it is one of the earliest reliable signals in move-out probability.

‍

5. Routing. Asking for a person, for hours, for an address. Asking for a human is an intent in its own right, not a failure state.

‍

6. Turns that are not requests. Hellos and goodbyes, thanks, a yes with nothing attached, wrong numbers. Nothing was asked, so nothing can convert. Every real dataset also has an unclassified bucket, and it needs a rule.

‍

Only One Group Usually Converts

‍

A conversion rate computed against all six divides rentals by a pile that includes payment questions, gate lockouts, cancellations, and greetings. The result is arithmetically valid and operationally meaningless. It moves when your delinquency rate moves. It moves when a gate breaks.

‍

Group one is the denominator. Everything else is a different measurement with a different name. Billing and account calls are measured on resolution and recovery. Access and service on time to resolution. Lifecycle on save rate and notice lead time. Routing on transfer rate. Turns that are not requests are excluded from every rate.

‍

Smartlock reported a 7.6% transfer rate on answered collections calls. That is a routing measure on billing traffic. It says nothing about how well voice sells, and it was never meant to.

‍

Which Agent Answered Is Not What the Caller Wanted

‍

This is the part that trips up teams who have already separated sales traffic from service traffic.

‍

Routing sorts calls before anyone speaks. Intent is known after. The two sorts disagree in both directions. Buying intent lands on the support line because a prospect called the main number from Google. Billing and access intent lands on the sales line because a tenant called the number on their gate sticker.

‍

Sorting by which agent answered is a large improvement over counting everything. It is also a proxy, and proxies drift. If your platform reports intent, measure on intent. If it reports only routing, know that your sales denominator carries traffic that was never going to rent.

‍

How to Sort Your Own Calls

‍

Pull 100 recent calls. Tag each with a single primary intent, using the caller's opening request, so a call that starts as a billing question and ends in a transfer stays a billing call. Roll the tags into the six groups. Divide rentals by group one. Compare that to the number you have been reporting, and the gap is the size of the problem.

‍

Two rules keep it honest. Pick the first request rather than the more interesting one, or every ambiguous call drifts toward the flattering category. And decide up front what happens to unclassified calls. Excluding them is defensible. Silently dropping them is not.

‍

Three Questions for Any Vendor Quoting a Rate

‍

What is in the denominator? All conversations, all calls, or calls with buying intent.

‍

What time window? Smartlock's weekly conversions ran 1 to 4 through winter, peaked at 15 in early May, and settled around 5 to 7. Peak week gives a number roughly triple the plateau.

‍

What is the base? Smartlock's voice-attributed share was 91% in March on 20 rentals and 88% in May on 38. The higher percentage sits on the smaller base. A share without the count under it is decoration.

‍

The same three questions apply to containment and resolution rates, which carry the same ambiguity and get quoted with the same confidence.

‍

Doing It at Scale

‍

Sorting 100 calls by hand is a diagnostic. It is not a reporting system, because nobody is tagging 10,000 calls a quarter with a highlighter. Continuous sorting has to happen as the conversations happen, across whichever channel the caller chose. A prospect who calls Tuesday and texts Thursday is one person with one intent, and two disconnected tools will count them twice.

‍

swivl runs across 4,500+ self storage locations and has handled more than 7.5M conversations, with over 85% of those conversations resolved without staff involvement. Reporting is scoped by location, channel, and reason for the conversation, so a conversion rate can be built on the calls where a lease was on the table.

‍

See it in action.

‍

Similar posts

Get started today

See how we can help automate your business today.
Book a demo!