Case study · Real estate · Washington, DC area
A DC-area real estate team: the day the leads disappeared
A top-producing real estate team's Google Ads account looked like it was collapsing. The campaigns had not changed. The tracking had. Here is how JERVO found it, proved it, and rebuilt on a clean signal.
The starting point
The client is a top-producing real estate team in the Washington, DC area, running Google Ads for buyer and seller leads across Washington and Arlington. A lead is a registration on the team's property search site, which feeds straight into the CRM. The account had been restructured in the spring: separate buyer campaigns by market, a seller campaign, a rebuilt brand campaign, a master negative keyword list, and target cost per lead bidding on each.
What happened
On June 11, 2026, the conversion tag on the site's signup form went dead. Nothing in Google Ads changed. Nothing in the campaigns changed. But from that day the account stopped recording leads, conversion rates fell off a cliff, and Smart Bidding, which needs a steady stream of conversions to bid well, started making decisions on a signal that no longer existed.
To the client's website vendor it looked like the campaign changes had broken the account, and they said so. Tag failures do not show up in Google Ads change history, so the campaign is always the first suspect.
What JERVO did
The rebuttal was built from the data. Conversion action status, conversion counts segmented by campaign, and one behavioral tell: branded searches, the people typing the team's own name into Google, were recording zero leads. Brand traffic does not stop converting overnight. Tags do. Anchoring the timeline to June 11 with the evidence laid out visually settled the question.
Then the fix. JERVO coordinated a new tag with a new web team, ran the old and new conversion actions in parallel to prove they counted the same events, and only then made the new action primary and moved the old one to secondary. Reporting during the parallel run was de-duplicated by hand so the client never saw an inflated number. Once the signal was clean, bidding was rebuilt on it and budget moved toward the buyer campaigns, where cost per lead was lowest.
The results
August 2026 against July, both on the clean conversion action: 958 registered buyer and seller leads, up 36 percent from 706. Cost per lead $8.69, down 16 percent from $10.35, on 14 percent more spend. The Washington buyer campaign delivered 553 leads at $4.69 each. Arlington delivered 237 at $11.47. The seller campaign's budget was trimmed and its cost per lead held at about $16.
One honest note: the earlier months are left out of this comparison on purpose. During the tag failure and the parallel run, the counts were either missing or doubled, and comparing against them would make the recovery look bigger than it was. July and August are the first two full months on the clean signal, so that is the comparison we trust.
What's next
The next phase is keyword segmentation on the buyer campaigns, isolating the winners into their own ad groups with their own targets, and closing the last attribution gap between the site's registration form and the CRM so every lead carries its source.