Scoring Real Estate Lead Generation Agencies: No Ranking, Just Evidence

What to take away

Score each agency against the job you are buying. No provider wins across all five categories.
Four tests carry most of the decisioncost predictability, lead provenance, account ownership and exit.
Fixed fees behave differently from commission shares. A 25% to 30% referral fee is a variable cost.
Verdicts below are per category and conditional. No overall number one is named here, and placement is not for sale.
Two finalists can be scored with one table in about an hour.

The four tests, and what each one asks

The four tests

Question

Cost predictability: Is the fee fixed, per lead, or a share of commission?
Lead provenance: Can origin, permission and transfer of every record be rebuilt?
Account ownership: Who holds the ad account, pixel, CRM and creative?
Exit: Can records be exported, suppressed and deleted?

Pass condition

Cost predictability: You can state the monthly worst case before signing
Lead provenance: Source page and consent are retrievable on request
Account ownership: Client logins work without vendor help
Exit: Export and deletion support appear in the contract

Cost structure decides the first test more than sales skill does. A marketplace charging a share of commission cannot quote a fixed monthly figure, because your commission is not fixed. That is not a fault in the provider. It is a mismatch with the buyer.

For a fair comparison of two finalists, the real estate lead generation pilot design sets out an eight-week test against ad hoc outreach.

Example: HomeLight and Real Geeks on the same four tests

HomeLight charges a referral fee typically 25% to 30% of the agent's commission, so cost predictability fails on the first line: the monthly figure follows closings, not a rate card. Request the source record and the export clause before comparing it with anything else.

Real Geeks plans typically start around $249 per month plus ad spend. That passes cost predictability for a solo agent, because the floor is known. Ownership is the test to press, since the site, CRM and lead records sit inside the vendor platform.

A fixed-fee website vendor and a referral marketplace do not compete for the same budget line.

HomeLight and Real Geeks

Providers

Paid media: Ylopo, Real Geeks
Lead marketplace: HomeLight, Zillow Premier Agent
Full funnel: BoomTown, CINC
Predictive seller data: SmartZip, Offrs

Fits

Paid media: Ylopo for teams with an existing CRM; Real Geeks for solo agents and small teams
Lead marketplace: Listing agents in active markets where speed to lead decides
Full funnel: Teams and brokerages above $1,000 per month
Predictive seller data: Listing agents working one farm area or ZIP code

Condition before signing

Paid media: Ad account and pixel stay in the client's name
Lead marketplace: Fee or per-lead price confirmed in writing
Full funnel: CRM export rights confirmed
Predictive seller data: Model inputs and territory limits explained

Ylopo typically runs $500 to $1,000 per month plus ad spend. BoomTown and CINC plans typically cost $1,000 or more per month. SmartZip's SmartTargeting typically costs $500 or more per month per farm area. Zillow Premier Agent lead prices vary by ZIP, often $20 to $75, and Flex referral fees typically run 30% to 40% of commission.

The follow-up system behind those numbers decides the outcome. The best practices checklist from form to follow-up marks the handoff points where paid leads usually go cold.

Build your shortlist in five steps

Write the job in one sentencebuy traffic, transfer inquiries, set appointments, or run the full funnel.
List the providers that do that job and nothing else.
Apply the four tests and mark each one pass, fail or unknown.
Request one dated evidence sample per finalista report, a source record, an export file.
Set a written stop condition and a pilot ceiling before signing.

Unknown is the usual result, and it is not a failing grade for the provider. It is a question to send in writing.

What the evidence has to show

A source page and consent record for two sample leads.
A report that states the denominator, not only a conversion percentage.
Named staff for the account, plus any subcontractors.
Client-owned logins, tested before signature.
Export, suppression and deletion terms inside the contract.

Ask for the denominator in writing, then compare it against something outside the vendor's own deck. The guide to reading lead generation benchmarks sceptically explains why ranges beat single headline numbers.

Bought placement breaks the exercise. The FTC guidance on soliciting and paying for reviews treats conditioned incentives and undisclosed ties as deceptive, so a list that will not show its scoring should be read as advertising.

Settle evaluation factors and their weight before proposals arrive, the discipline behind proposal evaluation under FAR 15.305, though private brokerages are not bound by it. Vet the technology layer and its subprocessors as NIST SP 1326 describes.

Fees, ZIP coverage and platform rules move every year. Re-score the shortlist on a fixed date, using the marketing strategy checklist that catches what audits miss as a second set of eyes on the same contract.

Common questions

Which agency is best?

None overall. The question hides a category mismatch, because a referral marketplace and a monthly software vendor are bought with different money and carry different risk.

Are these verdicts paid?

No provider paid to appear, and each fit above is a condition to verify rather than a bought rank.

What if two agencies score the same?

Run the pilot on both. Score reply speed, record quality and cost per appointment instead of case-study percentages.

What should be checked before a long contract?

Export rights, deletion terms and one tested login in your own name. Then run a 30 to 60 day pilot with a written ceiling.