Most comparisons of real estate lead sources rank channels by volume, or by a cost-per-lead figure someone estimated. Neither answers the question an agent actually has, which is: given my situation, my cash, and my time, which two of these should I do?
Three axes settle that. What it truly costs including your hours, how long before it pays, and whether anything remains if you stop.
The axis nobody uses, and the one that matters most
Ask what is left when the activity ends.
Portal leads, paid ads, and sponsored placement leave nothing. Stop paying and the pipeline empties as the current contacts work through. Rented.
Referrals, reviews, your own pages, and local reputation keep working after you stop actively feeding them. They decline slowly rather than switching off, and they travel with you between brokerages. Owned.
That distinction is the one that reorganises the whole list, and it is the one channel comparisons usually omit, because it makes the fastest options look worse than their headline numbers suggest.
The eight channels, honestly
Past clients and people who already know you. Lowest cost, highest conversion, slowest to feel like work. Owned. Consistently underrated because it resembles admin rather than marketing.
Referrals from other professionals. Lawyers, mortgage brokers, contractors. Moderate effort, long build, genuinely durable once established. Owned.
Portal and paid leads. Immediate volume, shared with competitors, priced privately, gone when you stop. Rented. Covered in detail in buying leads versus owning your own source.
Paid ads you run yourself. More control than portal leads, same rented economics, and the return depends heavily on the landing experience, which is where most agents lose it.
Search visibility. Slow, compounding, and the only channel where a solo agent can beat larger competitors by being genuinely more specific. Owned.
AI search visibility. Behaves like an extension of the search channel, since it draws on the same assets: your site, reviews, and consistent business facts. Fewer visits, warmer, arriving near a decision.
Geographic farming. Expensive in both money and years, and among the strongest positions available once established. Owned, provided you publish something that captures the verification moment.
Open houses and cold outreach. In-market people, physically present, and almost entirely dependent on follow-up. Most agents collect names and never contact anyone again, then judge the channel by the day rather than by the year.
Where the market actually gets its agents
Worth anchoring all of this in the survey data rather than in vendor claims. Per Virginia REALTORS' summary of NAR's 2025 profile, 43% of buyers used an agent found through a referral, and another 18% reused an agent from a past transaction. NAR's own takeaways report 91% of sellers used an agent, with FSBO at an all-time low of 5%.
The largest channel in this business is personal recommendation, and it is the one nobody can buy directly. The closest thing to buying into it is being easy to verify when your name gets passed along, which is why the owned assets matter beyond the traffic they generate.
Counting the cost honestly
A channel that costs nothing and consumes ten hours a week is not free. Price your hours at whatever an hour of your work is worth and add it.
Two things change when you do. Cold outreach and social posting look considerably more expensive than they appear on a budget spreadsheet. And some paid channels look cheaper than the unpaid alternative for agents whose time is genuinely scarce, which is a legitimate reason to buy leads that has nothing to do with laziness.
Your mix should depend on where you are
Most advice is written as though every agent should run the same channels, which is why so much of it fails.
With no past clients and no reviews, the owned channels have nothing to compound from yet. The first job is manufacturing the minimum version: telling everyone who knows you, becoming verifiable, learning one area properly. We laid that out in what actually works in year one.
With a database and reviews and no system, you are almost certainly leaving the cheapest business in your market untouched while paying for strangers. Fix the referral habit before buying anything.
With both working and cash to invest, paid channels become a reasonable accelerator rather than a foundation.
The constraint is attention, not budget
The instinct with money available is to run everything. It rarely works, because most of these channels need consistency rather than funding. Recognition comes from repetition to the same audience, and five channels sampled shallowly reach nobody often enough to be remembered.
Two channels held for three years beat six held for three months, every time, and the arithmetic does not care how well funded the six were.
Judging a channel before it has produced a closing
The practical difficulty with owned channels is that the feedback arrives after most people have already given up. A year is a long time to continue something with no closings attributable to it.
So judge on leading signals instead, which appear far earlier and are specific to each channel.
For referrals and past clients: how many individual conversations you had this quarter, and whether anyone described you to a third party. The second one is the first real evidence, even when it produces nothing.
For search and AI visibility: whether you appear at all for your own name plus your area, and whether the enquiries you do receive mention something specific you published. A visitor referencing your neighbourhood page is the channel working, months before it shows in closings.
For farming: whether people in the area recognise your name before you introduce yourself.
For paid channels the leading signal is simply cost per closing, because they produce fast enough to measure the real thing.
The rule underneath this: if the leading signals are rising while closings are flat, the channel is working and has not paid yet. If the leading signals are flat after two quarters of consistent effort, something is wrong with the execution rather than with the timeline, and that is the moment to change approach rather than to wait longer.
The distinction matters because both situations feel identical from inside them. An agent nine months into building an owned channel and an agent nine months into doing it wrong have the same bank balance and the same doubts. Leading signals are the only thing that tells them apart before the year is over, which is why choosing them in advance is worth the ten minutes it takes.
The takeaway
Compare channels on what they cost including your time, how long they take to pay, and what remains when you stop. Rented channels buy speed at a permanent cost and belong in a business that can afford them. Owned channels demand patience and then keep paying, and the NAR data says they describe how most clients already choose their agent. Pick two that match where you actually are, hold them longer than feels comfortable, and make sure the verification layer underneath them is solid, because every channel on the list eventually routes through someone checking you online. Two channels, chosen honestly and held for three years, is a strategy. Eight, sampled, is only a list of things you tried.
