How to track lead source ROI
The BrokerFold Team — Product + Engineering
Last reviewed
On this page
Tracking which lead sources make money takes three moves: stamp every lead with its source at the moment of capture, carry that stamp untouched through to closed deals, and judge each source on cost per closing and time to close rather than lead count. Simple to describe, rarely done — and the practices that do it consistently discover their budget has been funding the wrong things.
Stamp the source at capture, automatically if possible
Attribution dies at the entry point, so fix the entry point first. Every route a lead can arrive by gets a source value the moment the record is created: website form, portal inquiry, Google ad, social campaign, open house, sign call, referral, sphere. Automatic stamping beats manual every time a human is busy — which is always — so let technology do what it can: leads captured by your own website should arrive in the CRM already tagged with where they came from, and ad-campaign leads should carry their campaign. The manual remainder (referrals, phone calls, conversations at the rink) survives on one habit: no new contact gets saved without a source. An “unknown” pile that grows past a few percent quietly invalidates the whole exercise, because it is never a random sample — it is usually your best, most-organic sources going uncounted.
Carry the stamp to the closing table
The stamp only pays off if it survives the whole journey, and real estate gives it months of chances to get lost — reassignments, nurture loops, a lead who goes quiet and returns through a different door. Two rules protect it. The source field records where the contact first came from and never gets overwritten; later touches can be noted, but first-touch is the value you will judge spend by. And closed deals get a closing pass: when the commission arrives, confirm the source on the record is real before it enters your numbers. That sixty-second ritual at the moment of celebration is the cheapest data-quality program in existence.
Judge on cost per closing and time to close
Lead volume is the vanity metric of this business — a source producing two hundred cheap leads and zero transactions is an expense wearing a costume. The numbers that decide are: cost per closing (everything a source costs per year, divided by closed deals it produced) and time to close (how long that source’s leads take to become revenue). The second is the one agents skip and shouldn’t: a source that closes in six weeks and a source that closes in fourteen months can have identical ROI on paper and wildly different effects on your cash flow and your patience. Run your own numbers through the lead ROI calculator — cost per lead, conversion rate, average commission — and the comparison between sources usually stops being close. For the spend side of the same decision, see how much a realtor should spend on leads.
What honest tracking tends to reveal
Patterns recur across practices. Referral and sphere deals dominate profitability while receiving the least deliberate investment — usually an argument for spending more attention there, not less. Purchased leads look terrible on conversion rate and merely mediocre on cost per closing, which reframes the buy-versus-generate decision as a math problem instead of a religious one. And your own website typically shows the best economics of any paid channel once it produces steadily, because the recurring cost is flat while the lead flow compounds with your content and your market presence. Whatever your version of these patterns turns out to be, the response is the same: move budget toward your best cost-per-closing, and give every verdict enough closings to be a verdict.
Keep the system maintainable or it isn’t a system
The failure mode of attribution is ambition. A solo agent does not need multi-touch models or a taxonomy of forty sources — a dozen source values, first-touch credit, one closing-time confirmation, and a twice-yearly review outperform any elaborate scheme that gets abandoned in March. Where your website is the hub, its analytics add the upstream half of the picture — which pages and campaigns create the leads your CRM then follows to closing. Together they answer the only question the spreadsheet exists for: where should next year’s dollar go?
Frequently asked questions
What about deals with more than one source — an ad click, then a referral?
Pick a consistent rule rather than a perfect one. First-touch (credit the source that created the contact) is the easiest to maintain and the right default for judging spend. Note assists in the record if you like, but resist building multi-touch attribution for a solo practice — the model you maintain beats the model you abandon.
How long before source data is worth acting on?
Two or three closings per source, which for most practices means judging patiently — often a year. Real estate's long cycles and small numbers make quarterly verdicts noisy: one lucky referral can crown a source, one slow buyer can damn one. Act early only on the extremes, like a source producing nothing but fake numbers.
Do I need special software for this?
No — you need a source field that gets filled at capture and read at closing, which any CRM has. Automation helps accuracy more than sophistication: website leads that arrive already stamped with their source remove the human forgetting step, which is where most attribution actually fails.
Keep reading
- Lead ROI calculator Cost per lead through closings and GCI — what your leads actually return.
- How much should a realtor spend on leads The 10%-of-GCI rule of thumb, commonly cited monthly ranges, and the deals-backwards worksheet that turns a target year into a defensible lead budget.
- Buy leads or generate your own? Buying is speed with a tax: shared attention, rising prices, no equity. Generating is a compounding asset with a ramp. The math and the hybrid that works.
- What website analytics should realtors track Five numbers over dashboards: visitors, lead conversion rate, leads by source, alert signups, and cost per lead by channel. Everything else is commentary.