Real Estate Paid Media Audit: The Monthly Method

• In Q4 2025, advertisers enforcing quality governance turned 56.7% of programmatic spend into benchmark-qualified impressions. Lower performers managed 37.5% (ANA, release of 25/02/2026). Nineteen points of the same budget separate discipline from its absence.
• Meta redefined what counts as a click in March 2026 and renamed engaged view to engage-through (Digiday, 09/03/2026). Your historical series has a seam there, and year-over-year comparisons crossing it compare two definitions.
• The audit that reconciles numbers proves bookkeeping. Proving the campaign created demand takes one incrementality test per launch, and most developers run none.

The monthly media review follows a familiar script. Chart trending up, cost per lead under target, winning creative on its own slide, and a spend line that closes against the approved budget. All of it true. Next to it, the CRM report shows a record count that does not match the lead count on the slide. Both spreadsheets are correct, they describe different things, and the gap between them is the information the meeting never produces.

Auditing media carries a reputation for distrust of the agency. What the audit actually does is more mundane: it establishes which of two truthful reports describes your launch. It takes ninety minutes a month and returns a single number, tracked over time.

Nineteen points of the same budget

The most useful number about media auditing did not come from real estate. It came from the ANA programmatic transparency benchmark released on 25/02/2026, produced with TAG TrustNet and Fiducia, covering the fourth quarter of 2025. Advertisers enforcing disciplined quality governance converted 56.7% of programmatic spend into impressions meeting the study's criteria: fraud-free, measurable, viewable and free of sites built only to host ads. The lower-performing group converted 37.5%.

Two details in the same study change the reading. The share of advertisers above the 50% threshold fell from 50.0% in Q3 to 43.8% in Q4, meaning the average worsened while the best improved. And the benchmark's case studies record reductions of nearly 40% in cost per conversion once optimization moves to quality-adjusted metrics instead of CPM alone.

Launch budgets live in Meta, Google and listing portals, while the study covers open-web programmatic, a different market. What transfers is the mechanism: the gap between the two groups comes from someone checking with a method, not from creative or from inventory choice.

The four files you request before the meeting

Asking for a more detailed slide does not help, because the slide is the finished product of the choices you want to inspect. Ask for the inputs, three business days ahead:

• Raw platform export, CSV, ad level, same date window as the slide, with attribution window and model columns visible. Meta Ads Manager and Google Ads export this in two clicks.
• Platform invoice, or the ad account's card statement, for the same period. It is the document that separates media delivered from media billed and from agency fee.
• CRM export, with creation timestamp for each record, source, campaign and first owner.
• Access map: who holds admin on Business Manager and the Google Ads account, whose entity the pixel sits under, and where the custom audiences live. That answer belongs in the contract before it belongs in an audit, and it is settled in the setup work that happens before the ad runs.

The three reconciliations

Spend. The raw export total, the invoice total and the slide line have to match to the cent. When they do not, the usual causes are campaigns outside the report's scope, another project's budget inside the same account, and platform credits booked in a different month. None is serious, all are relevant.

Volume. Platform-reported leads against CRM records created in the same period. This is where the gap appears, and the temptation is to treat it as an error to eliminate. The objective is different: measure that gap as a percentage, month over month, and keep it as a fixed line in the report. A stable 20% gap is a measurement system with known loss. A gap that jumps from 20% to 60% in one month is a broken integration, and the discovery usually arrives three weeks late.

Quality. Of the records that landed, how many were contacted inside the agreed response window, how many became viewings and how many became offers. That cut exposes the case where media worked and sales follow-up did not, which is the opposite of the diagnosis a media review normally produces.

Forms and chat are two columns

A form lead carries a campaign identifier, reaches the CRM through an integration and reconciles directly. A click-to-chat lead behaves differently: the platform logs the button click, and the conversation may never start. WhatsApp passes no referrer, so tying a conversation back to an ad depends on capturing the ad identifier at the moment the message arrives, which has to be configured deliberately.

The audit question is therefore double: what share of chat clicks became a conversation with a first message, and of those, how many have an ad identifier recorded. A report that adds form submissions and chat button clicks into one column called "leads" presents a total describing two different events. Two columns fix it.

March 2026 put a seam in your series

In March 2026 Meta began requiring an actual link click for click-through attribution, and renamed engaged view to engage-through while widening the metric's scope. Digiday's reporting of 09/03/2026 carries the practitioner reading: the change alters how results are labelled, without touching delivery or optimization, and exists to bring platform numbers closer to third-party tools.

Performance comparisons crossing March 2026 compare two definitions of a click. The audit marks the seam in the spreadsheet, records which attribution setting is live in each campaign, and freezes that setting for the duration of the launch. Changing the attribution window mid-launch improves the report without improving sales.

The argument: reconciliation proves bookkeeping, not cause

Here is the part that splits the room. The three reconciliations prove the numbers are honest. None of them proves the campaign created demand. A large share of luxury launch leads arrive through branded search, listing portals and referral, and paid media captures that traffic competently while billing for it as though it had originated it.

So a complete audit includes one incrementality test per launch rather than per month: a secondary market held out for two weeks, a pause window during the sustaining phase, or the platform's own lift test. The objection is fair, since a launch has a short window and switching off media where inventory sits is an expensive decision. That is precisely why the test runs once per cycle, in the lowest-risk market or period. Without it, the arithmetic measures what you paid to recapture demand your brand had already created. The inventory of where that money tends to disappear sits in where paid media burns budget.

The calendar

Monthly, ninety minutes: the four files, the three reconciliations, and the platform-to-CRM gap written into the same cell every time. By phase: channel allocation, creative rotation and attribution window. By launch: the incrementality test and a review of the access map. The quarter asks for nothing more, and the media review gets its agenda back.

Frequently asked questions

Does auditing media mean replacing the agency?

In most cases, no. The audit produces a conversation about method, and an agency with its house in order hands over the four files without friction. The warning sign is refusal to share the raw platform export, or recurring delay in showing the invoice. A bad number gets fixed through optimization; a missing trail does not.

What gap between platform leads and CRM records is acceptable?

No published reference figure works as a yardstick, because the gap depends on form design, integration quality and how much of the operation runs through chat. What works as a yardstick is stability: your own average of the last three months. A sharp move in that line signals a broken integration, and that is what the number is for.

Can you audit without admin access to the accounts?

Spend and volume reconcile from the exports. What stays out is verification of account structure, audiences and change history, which requires read access to Business Manager and Google Ads under the developer's own entity. Requesting that in the contract costs less than requesting it mid-audit.