What Counts as a Lead? Why Your Conversion Rate Can't Be Compared to Anyone Else's

Someone in a portfolio review asks the question that sounds simple.

"Is 12% good?"

The room goes quiet for a second, because there are two honest answers and neither is useful. The first is "compared to what?" The second is "12% of what?"

Then somebody Googles it during the meeting, reads out a benchmark from the first result, and everyone moves on feeling slightly better and no better informed.

The number is not wrong. The denominator is undefined.

 Search for the residential lead-to-lease benchmark and the results will not agree with each other. 

One widely cited multifamily figure is 8.7%. It comes from a leasing-funnel analysis of 1.5 million leads across 4,300 properties during the 2024 peak leasing season, and it is a real number from real data. But read what it actually measures. It is 8.7% of prospects who entered the funnel by filling out a guest card, which works out at roughly 12 enquiries per lease. Top performers in the same dataset reached 16.5%, or about six.

Hold onto that phrase: by filling out a guest card. Not by calling. Not by messaging a portal. Not by walking in. One channel, one entry point, one definition.

Keep searching and the ranges widen. You will find figures in the single digits and figures running to 30% and beyond, and the gap is often definitional rather than operational. Different populations, same label.

That spread is not evidence of a confused industry. It is evidence that everyone is dividing signed leases by a different number and calling the result the same thing.

Watch the same month produce four different answers

Here is a month, simplified. Your team signs 25 leases. Every calculation below uses those same 25 leases and the same team doing the same work.

Nothing changed except the denominator. The performance is identical in all four rows.

Now think about what happens when you take row one into an owner meeting and they compare it to a benchmark built on row four. You will spend the meeting defending a number that was never comparable, and you will probably lose that argument, because the other number looks better and neither of you can explain why.

That is the real cost. Not that the benchmark is wrong. That you cannot defend your own number, so you stop bringing it, and a metric nobody reports is a metric nobody improves.

Four things teams count differently

When operators write down how they actually define a lead, the disagreement clusters in four places. Every one of them is defensible. The problem is picking without deciding.

• Duplicates across channels. A prospect enquires on two listing sites and then calls the office. Some systems record that as three enquiry events, because each arrived through a different channel. Ask the leasing manager and they will usually say it is one person who is very interested. Both views are reasonable. Very few operations have the deduplication rule written down anywhere.

• Enquiries about unavailable units. Somebody asks about a unit that was leased last week. They are a real person with real intent who found you and made contact. Do they enter the funnel? If yes, your conversion rate falls every time your marketing is slow to come down. If no, you are hiding a marketing problem inside a definition.

• Non-responders. A prospect submits a form and never replies to anything after that. Some teams count them, on the basis that a lead is anyone who raised a hand. Others exclude them, on the basis that you cannot convert a person who will not talk to you. This choice can move the rate materially.

• Bots, spam and agents. Automated form fills, listing scrapers, and other agents fishing for availability. Almost nobody wants these in the count, and almost nobody has a rule that removes them consistently.

There is not one universal answer here. There is only a consistent one, chosen to fit what the metric is for and applied to every month, so that your line moves for operational reasons rather than definitional ones.

Even with the same definition, the comparison still breaks

Suppose you and another operator agree on all four rules. Your rates are still not comparable, for reasons that have nothing to do with leasing quality.

• Pricing. If your rents sit above the market while comparable properties adjust down, conversion drops with identical traffic and an identical team. The rate is reporting a pricing decision, not a marketing one. That is worth sitting with: your leasing conversion number can fall because of a decision made in a revenue meeting nobody from leasing attended.

• Screening standards. Tighter criteria mean fewer approvals from the same applicant pool. A lower conversion rate can be the intended result of a deliberate risk decision.

• Channel mix. A portfolio leaning on high-volume portal traffic will convert at a lower rate than one leaning on referrals and walk-ins. Neither is better. They are different funnels with different shapes.

• Asset class and submarket. Student housing, affordable housing and market-rate multifamily produce enquiry behaviour so different that a shared benchmark means very little. In heavily oversubscribed affordable housing, a low conversion rate can indicate extremely high demand rather than poor performance.

Four variables, none of which are visible in the number, all of which move it.

Write the four rules down

Start here. The right-hand column is one worked example, not a prescription.

Pick the deduplication window based on how long you consider an enquiry to be the same leasing opportunity. What matters more than the number is that it is written down and dated.

Then stamp the definition on the report itself. One line at the bottom saying what was counted and when the rule last changed. When somebody asks "is 12% good?", that line is your answer to "12% of what?", and you keep the meeting.

That is the easy half.

A lead definition is a capture rule, not a reporting rule

Here is what makes this harder than an afternoon of decisions.

You cannot filter for something your system never recorded. Every one of those four rules has to be enforceable at the moment the enquiry arrives, not at the moment somebody builds the report. Look at what each one actually requires.

• Deduplication requires resolvable identity across channels. This is the one people underestimate. Depending on the portal and how enquiries are routed, prospect contact details may be masked or relayed, which can make the same person appear as two different records. The address attached to a portal enquiry often has no relationship to the mobile number they use when they ring the office on Thursday. Matching on email alone fails precisely where duplication is heaviest, and it fails quietly. You will have a deduplication rule that everyone believes is running.

• The unavailable-unit rule requires the enquiry to be linked to a unit whose status was live at that moment. If availability sits in one system and enquiries in another, you are reconstructing after the fact. By the time the report gets built, the unit's status has changed, and nothing recorded what it was when the person asked.

• The non-responder rule requires you to have logged a two-way exchange. Most operations can tell you what was sent. Far fewer can tell you what was answered, across phone, email, portal message and SMS together. Without that, "non-responder" is a guess dressed as a filter.

• The spam rule requires the exclusion to be logged and reversible. Otherwise you have a filter nobody can audit, and a rate that moved for reasons nobody can reconstruct six months later.

Two of those four can be approximated manually with enough discipline. The other two get considerably harder to apply consistently when enquiry and activity data is spread across separate channels and systems. Not impossible, but the effort scales with every property you add, and manual reconciliation tends to quietly stop happening around the point the portfolio gets busy.

Which reframes the problem. The reason so few operations have a consistent lead definition is not that nobody got round to writing one down. It is that the writing is the last step, and the unified customer view that makes the rules enforceable at capture is the first.

Benchmark against yourself, then against your own properties

Once the definition is fixed and your system can actually apply it, the useful comparison is not the industry. It is your own trend and your own portfolio.

Track the same definition month over month and watch the direction. Six months of your own data on a consistent rule can be more useful for an operational decision than any published figure, because it controls for variables a benchmark cannot see.

Then compare properties inside your own portfolio, where pricing strategy, screening criteria and process are broadly shared. Two similar assets in similar submarkets converting at 8% and 16% is a finding you can act on this week. It is also a much easier conversation than an industry comparison, because both numbers came out of the same system under the same rule.

That last point is doing more work than it looks. Property-level comparison only means something if every property is being counted the same way, which is a question about your reporting layer rather than about your leasing teams.

Where the definition tends to break in practice

• The rule outruns the system. Covered above, and it is the most common failure by a distance. A definition your tools cannot apply consistently is not a definition you have in practice.

• Nobody owns the rule. The definition exists in a document, but each report gets built by whoever has time, and the filters drift. Six months later the trend line is measuring changes in reporting habits.

• The definition changes without a note. Someone adds spam filtering, the rate jumps three points, and the improvement gets attributed to a leasing initiative that launched the same month.

• Stage rates get skipped. The headline number moves and nobody can say which stage caused it. If you are tracking one number, you have a scoreboard rather than a diagnostic. Our guide to improving lead-to-lease conversion covers the stage breakdown and what each rate is telling you.

What good looks like

A leasing report where the conversion rate carries its own definition. A contact rate sitting beside it. A trend line running back at least six months on a rule that has not moved. Property-level comparison inside the portfolio rather than against a figure from someone else's blog.

That is a low bar, and it is surprisingly easy for reporting practice to fall short of it. Which is why the question in the portfolio review keeps going unanswered.

The reason nobody could answer "is 12% good?" was never that the team lacked data. It was that the number arrived without the one line of context that would have made it mean something.

Writing the four rules takes an afternoon. Being able to enforce them takes a system that saw every channel at the moment each enquiry landed. Write the rules first, then test whether your current setup can apply even the first one.

If it cannot, that is worth knowing before the next portfolio review. This is the part RIOO was built around. Leasing management captures enquiries from website forms, listing portals, referrals and walk-ins into the same record set, with the source attached to each one, so the channel question stops being a reconciliation exercise. Unit availability sits in the same system as the enquiry, which is what makes the unavailable-unit rule answerable rather than reconstructed. And dashboards and reports read from that single record set, so property-level comparison is comparing like with like. 

Frequently asked questions

Q1. What counts as a lead in property management?
There is no universal definition, which is the source of most benchmark confusion. A workable one is a unique person, deduplicated across channels, who has made contact about renting from you. Your operation then needs explicit rules on duplicates, enquiries about unavailable units, non-responders and spam, captured consistently at the point the enquiry arrives.

Q2. Why do published lead-to-lease benchmarks vary so much?
Because each divides signed leases by a differently defined lead count. The widely cited 8.7% multifamily figure, for example, measures guest-card entries specifically rather than all enquiries across all channels. Other published ranges run to 30% and above, and much of that spread is definitional rather than operational.

Q3. Should non-responders count as leads?
Either choice works as long as you apply it consistently. If you count them, report your contact rate alongside your conversion rate so the two effects stay separable.

Q4. What is the difference between contact rate and conversion rate?
Contact rate is the share of leads your team actually reached with a two-way exchange. Conversion rate is the share that signed. Reading them together tells you whether a low conversion rate is a reachability problem or something happening later in the process.

Q5. What is a good lead-to-lease conversion rate?
Less useful a question than it appears, because the answer depends on your definition, your pricing position, your screening standards and your asset class. Our guide on improving lead-to-lease conversion covers the directional ranges and the stage-by-stage metrics underneath them.

Q6. How often should the lead definition be reviewed?
Periodically, but avoid changing it in the middle of a reporting period. Every change breaks comparability with everything before it. If you do change it, restate the prior periods under the new rule or annotate the break clearly on the chart.