Real estate has the fastest-rising cost-per-click of any measured industry — and refuses to average its two conversion rates

Real estate carries a published cost per click of $3.22, up 27.27% year on year, which is the largest single-year rise of any industry in the measured set, against a cost per lead of $102.51. That headline hides two businesses. Residential agents convert paid search clicks at 1.29% and pay $157.59 per lead. Property management, inside the same published vertical, converts at 10.24% and pays $76.71. Website conversion for real estate sits at 2.8%, the lowest of the thirteen industries in that dataset. On Meta the published real-estate cost per lead is $16.61, the lowest of any vertical there, with a 3.75% click-through rate, the highest.

A residential agent pays $157.59 per lead at a 1.29% conversion rate. Property management, inside the same published vertical, pays $76.71 at 10.24%.

1. 01

   The vertical is two businesses wearing one label

   An eightfold difference in conversion rate and a twofold difference in cost per lead is not variance around a mean, it is two distinct demand types filed under one category name. Any budget planned against the blended $102.51 is planned against a figure that describes neither business, and the model separates them before it prices anything.

2. 02

   The cheap channel and the expensive channel are not substitutes

   A $16.61 cost per lead on Meta against $102.51 on search is not a sixfold efficiency advantage. It is a difference in what the word lead refers to. A form fill served into a feed and a search for a named neighbourhood are different objects with different downstream behaviour, and the model prices and reports them separately rather than comparing them.

3. 03

   A rising click price against a flat conversion rate is a margin problem, not a bidding problem

   At +27.27% year on year, the largest rise of any measured industry, an account holding its conversion rate steady is losing ground every quarter without doing anything wrong. The available response is on the conversion side, which is also where the published website rate of 2.8% says the most room exists.

4. 04

   There is no credible published rate for a real-estate lead reaching a closed transaction

   This is worth stating plainly because the figure is quoted constantly. No credible published source establishes it, and the numbers in circulation trace back to vendor marketing pages rather than to any dataset. The model therefore stops at cost per appointment held and requires the brokerage's own closed-transaction records to go any further. It does not estimate the gap, and any model that does is estimating from nothing.

5. 05

   The internet finds the property; the agent is found elsewhere

   52% of buyers found the home they purchased on the internet and 88% used an agent. Media aimed at being discovered as an agent is competing for a different and smaller behaviour than media aimed at surfacing property, and the two deserve separate budgets and separate measurement.

Residential agent conversion **1.29%** vs property management **10.24%** (~8× apart) · overall real-estate CPC **$3.22 (+27% YoY)** · CPL $102.51 · residential-agent CPL $157.59.

The model splits the account in two before it does anything else, because residential and property management demand behave differently enough that a shared structure ensures one of them subsidises the other invisibly. Each channel is then priced to its job: search takes high-intent named-neighbourhood and property-management demand, social takes reach and listing exposure, and neither is judged on the other's cost per lead. Landing experiences are built per property type rather than per service, since the published website conversion rate is the constrained step. Tracking runs through to appointment held rather than to form fill, and reporting stops there by design, because the step past it has no published benchmark to lean on and must come from the brokerage's own records.

1. 1

   Weeks 1–2

   Separate the two businesses

   Residential and property management split into distinct accounts, budgets and reporting

   Owner: OmniFlow

2. 2

   Weeks 2–4

   Channel pricing

   Search and social given separate objectives, separate targets and no shared cost-per-lead comparison

   Owner: OmniFlow

3. 3

   Weeks 4–8

   Landing experiences

   Pages built per property type, against the published 2.8% website conversion rate as the constrained step

   Owner: OmniFlow

4. 4

   Weeks 6–10

   Track to appointment

   Instrumentation through to appointment held, not to form fill

   Owner: OmniFlow + brokerage

5. 5

   Month 3 onward

   Own close rate

   Closed-transaction records supplied so the model can extend past the appointment

   Owner: Brokerage + OmniFlow

6. 6

   Month 6

   Review

   Cost per appointment held, reported separately for each of the two businesses

   Owner: OmniFlow

The model reports cost per appointment held for each of the two businesses separately and refuses to publish a blended figure, because blending is what produced the misleading $102.51 in the first place. It would expect residential and property management to diverge sharply on every line. Past the appointment the model goes silent until the brokerage supplies its own closed-transaction data, since no credible published figure covers that step and inventing one would make every number above it unreliable too. None of the figures quoted here are a forecast for a particular brokerage; they are the published starting frame that the account's own data replaces from month two.

Modelled. Inputs: WordStream by LocaliQ, Google Ads Benchmarks 2026 (13,474 US search campaigns, April 2025 – March 2026); WordStream by LocaliQ, Facebook Ads Benchmarks 2025 (data April 2024 – June 2025); Ruler Analytics, Conversion Rate Benchmarks 2026 (110M+ sessions, 5M+ conversions); National Association of Realtors, 2025 Profile of Home Buyers and Sellers. Modelled outputs are not a forecast or a guarantee of results. Cited: LocaliQ *2026 Real Estate Search Benchmarks* (~900 campaigns).

Every figure below is published research, not a client result. They are the inputs to the arithmetic above, listed so it can be checked rather than taken on trust. The bracketed number points to the full citation at the end of this page.

These are the surfaces this engagement is run and measured from, shown with representative figures built around the benchmarks cited on this page. Every account we run reports into views like these, and you keep ownership of all of them.

Each figure on this page, the system it is read from, and the definition and window it is measured over.

Not applicable — this is a modelled engagement. Its weakest input is the one that does not exist: no credible published figure links a real-estate lead to a closed transaction, so the model cannot cost anything below the appointment. Until a brokerage supplies its own closed-transaction records this is a media plan rather than a revenue plan, and it should be read as one. The second weakness is that the published Meta and search figures are not measuring the same object, so the channel comparison is directional at best.

A modelled capability, not a client account. Figures illustrate what the model produces and are labelled as modelled wherever they appear. See all our work

Full citations for everything cited on this page, with the sample and period each study covers, so you can go and read the original.