A programmatic media audit is an independent review of how your display, video and CTV budget is actually being bought: where the impressions ran, what each layer of the supply chain kept, how the targeting was built and whether the reported results hold up. Most brands commission a programmatic media audit at one of three moments: performance has stalled, a new CMO or CFO wants to know what the company is paying for, or the agency relationship is up for renewal. For luxury and high-consideration brands, where customers are worth $5,000 or more and sales cycles run longer than 30 days, a second opinion is often the cheapest way to recover wasted budget before deciding whether anything bigger needs to change.
What does a programmatic media audit actually cover?
A useful audit looks at six areas. A weak one looks only at fees or only at a dashboard.
- Account ownership and access. Who owns the DSP seat, the data, the audience segments and the historical logs.
- Fees and working media. How much of each dollar reaches publishers after agency, technology, data and verification costs.
- Supply quality. Which domains, apps and publishers carried the impressions, and how much ran on low-quality or made-for-advertising inventory.
- Audience and targeting setup. Whether the affluent segments, first-party audiences and exclusions are built correctly and still sized sensibly.
- Campaign configuration. Frequency caps, pacing, bid strategies, optimization goals and attribution windows.
- Measurement. Whether reported outcomes reflect incremental revenue or simply credit the media for sales that would have happened anyway.
Each area can hide significant waste on its own. Together they explain why two brands with the same budget and the same DSP can see very different returns.
When should a luxury brand get a programmatic media audit second opinion?
The common triggers:
- Cost per acquisition is rising without a clear reason while reported impressions and clicks look healthy.
- Reporting is aggregated and the agency resists sharing domain-level or log-level data.
- The brand does not have login access to its own DSP seat or cannot export its own audience data.
- Senior management is questioning the channel. A finance team asking "what does programmatic actually do for us?" deserves an evidence-based answer.
- Before renewing or re-tendering an agency contract. An audit gives the review a factual baseline; our guide to running a media agency review for a luxury brand covers the wider process.
- After a merger, rebrand or new market entry, when legacy settings may no longer match the business.
A second opinion is not an accusation. Many audits confirm that an account is in good shape and identify a handful of fixes. That is still worth knowing.
Who owns the account, and why does a programmatic media audit start there?
Before analyzing performance, confirm you can see everything. Ask for read access to the DSP seat, the verification vendor dashboard, the ad server and any clean room or data onboarding accounts. If the agency operates the seat under its own contract, find out what happens to your audiences, logs and learnings if you leave.
This matters because an audit built on screenshots and exported summaries can only test what the current partner chose to share. Our article on who owns the DSP seat and data when you hire an agency explains the ownership models and what to put in the contract.
How does a programmatic media audit check fees?
Fees are where many audits begin, because the numbers are concrete. A typical programmatic dollar passes through several layers:
| Layer | What it pays for | Typical range seen in audits | Red flag |
|---|---|---|---|
| Agency fee | Planning, trading, reporting | 10-20% of media, or a flat retainer | Fee not disclosed or bundled into CPMs |
| DSP fee | Platform access | 10-20% of media | Undisclosed markup above the platform's rate card |
| Data costs | Third-party audience segments | $0.50-$3.00 CPM, often higher for affluent data | Data layered on every line, including retargeting |
| Verification and brand safety | Fraud, viewability, MFA avoidance | $0.05-$0.30 CPM | Paid for but not switched on pre-bid |
| SSP and exchange fees | Supply side take rate | 10-20% of the bid | Many reseller hops for the same inventory |
| Working media | What the publisher receives | 45-65% of total spend in a well-run account | Below 40% without explanation |
These are Stillwater planning ranges observed across audits in 2025 and 2026, not industry standards; your contract and DSP determine the real figures. For context, the ANA's Q4 2025 Programmatic Transparency Benchmark found that quality-led advertisers got 56.7 percent of spend into fraud-free, measurable, viewable, MFA-free impressions, versus 37.5 percent for lower performers. Our dedicated guide to programmatic fee transparency and agency audits goes deeper on the fee stack, rebates and audit rights.
How does an audit test supply quality?
This is where the largest savings usually sit. The auditor should:
- Pull domain and app-level placement data for at least the last 90 days and rank by spend.
- Flag made-for-advertising, low-attention and unknown domains, and calculate the share of spend they absorbed.
- Check supply paths using the IAB Tech Lab's sellers.json and ads.txt standards to see how often inventory was bought through resellers rather than direct sellers.
- Review CTV app-level delivery, including FAST channels, screensaver inventory and device types, so streaming spend is not quietly diluted.
- Confirm private marketplace deals are delivering and that the deal ID inventory matches what was sold.
If the account cannot produce placement data at this level, that is itself the most important finding. Our article on log-level data from a programmatic agency explains what to request. A clean supply policy, covered in our guide to supply path optimization for premium advertising, is usually the first fix.
What should the audit check in targeting and configuration?
Settings drift over time. A campaign built two years ago may still be running with assumptions nobody remembers making. Look for:
- Affluent segments that are too broad. A "high income" segment covering 40 percent of US households is not an affluent audience. Compare segment sizes with the realistic size of your market.
- Missing exclusions. Existing customers, employees, recent purchasers and job seekers should usually be suppressed from prospecting.
- Frequency caps that are absent or too loose. For luxury prospecting, Stillwater planning ranges are 3 to 8 impressions per household per week across display and CTV combined.
- Optimization goals that reward the wrong thing. Optimizing to clicks or cheap CPMs pushes spend toward low-quality supply.
- Retargeting taking credit. If retargeting absorbs more than 30 to 40 percent of budget and claims most conversions, the account may be paying to follow buyers who were already coming.
- Geography and dayparting errors. Delivery outside your service markets, or across hours when your buyers are not watching.
How does a programmatic media audit evaluate measurement?
For high-consideration brands this is often the most consequential section. Reported results usually rely on last-touch or view-through attribution with generous windows. Those methods can credit programmatic with sales it did not cause.
The audit should answer three questions:
- What attribution model and windows are used, and do they match a sales cycle that may run 30 to 180 days?
- Has the account ever run an incrementality test, such as a geographic or audience holdout? If not, nobody knows the true return. Our guide to holdout testing in advertising shows how to design one.
- Does reporting connect to revenue? For brands with long sales cycles, CRM matchback of exposed households to opportunities and closed sales is far more useful than platform conversions.
Measurement standards bodies such as the Media Rating Council accredit viewability and invalid traffic measurement, which is a useful check on whether your verification data is trustworthy, but accreditation does not prove incrementality. Only a test does.
What does a programmatic media audit cost and how long does it take?
| Audit type | Scope | Typical timeline | Stillwater planning cost range |
|---|---|---|---|
| Rapid second opinion | Fees, top placements, setup review from exports | 1-2 weeks | Often free or under $5,000 from a prospective partner |
| Standard independent audit | All six areas, 90 days of data, seat access | 3-5 weeks | $10,000-$35,000 |
| Forensic log-level audit | Impression-level analysis across a year | 6-10 weeks | $40,000-$100,000+ |
A free second opinion from an agency pitching for the business is useful but not neutral. Treat its findings as hypotheses to verify, and ask for the evidence behind each one. For larger budgets or contentious relationships, an independent auditor with no interest in winning the account is worth the fee.
What are the most common mistakes in a programmatic media audit?
- Auditing fees only. A low fee on badly bought media is still poor value.
- Accepting summary data. Without placement and log-level detail, the audit only checks the partner's own narrative.
- Judging by CPM or CTR. Cheaper impressions and higher click-through often signal worse supply, not better buying.
- Skipping the measurement question. The biggest error in most accounts is not waste on bad sites; it is believing results that were never tested.
- Ignoring CTV. Streaming budgets deserve the same app-level, publisher-level and household-frequency scrutiny as display.
- Not agreeing on a remediation plan. Findings without owners and dates rarely change anything.
- Treating the audit as a one-off. A light quarterly review keeps settings from drifting again.
How should you prepare for a programmatic media audit?
Preparation decides whether the audit takes three weeks or three months. Before the auditor starts, gather:
- Contracts and statements of work with the agency, DSP, data providers and verification vendor, including any rebate or volume terms.
- Invoices for the last 6 to 12 months, so billed amounts can be reconciled against platform spend.
- Read access to the DSP seat, ad server, verification dashboard and analytics, plus any clean room used for matchback.
- A list of active deals, including private marketplace and programmatic guaranteed deal IDs and the publishers behind them.
- Business outcomes by month, such as qualified inquiries, appointments, applications and closed revenue, so media can be compared with results rather than with its own reporting.
- Known changes, such as price moves, product launches, website rebuilds or tracking changes, that could explain shifts in performance.
Agree in writing on what the auditor may share, who sees the findings first and how the incumbent partner gets to respond. A fair process produces better answers and protects a working relationship if the account turns out to be sound. It also helps to name one internal owner, usually in marketing operations or finance, who can chase access requests and answer questions quickly; most audits stall on access, not analysis.
Finally, write down the decision the audit is meant to inform before it begins. "Should we renew?", "Can we cut 20 percent of spend without losing revenue?" and "Is CTV working?" need different depth and different data. A clear question keeps the scope and the cost under control.
What should you do with the findings?
Sort recommendations into three groups. Quick fixes, such as switching on MFA avoidance, tightening frequency or adding exclusions, can happen within days. Structural changes, such as moving prospecting into private marketplace deals or rebuilding affluent audiences, take a few weeks. Commercial changes, such as seat ownership, fee disclosure and audit rights, belong in the contract.
Then run a holdout test on the cleaned-up account. It is the only way to know whether the changes improved revenue rather than just the dashboard.
Want a second opinion on your programmatic account?
Stillwater Media reviews programmatic and CTV accounts for a limited number of luxury and high-consideration brands each quarter, covering seat ownership, fees, supply quality, affluent targeting and incrementality. If you want an evidence-based second opinion before your next budget decision, apply to work with Stillwater Media.
About the author: Stillwater Media is a selective performance media agency for luxury and high-consideration brands, headquartered in Charlotte, NC. The team plans and buys premium CTV, programmatic, audio, DOOH and social media for brands whose customers have a lifetime value above $5,000. Signal. Strategy. Scale.
Frequently Asked Questions
What is a programmatic media audit?
A programmatic media audit is an independent review of how a brand's programmatic budget is bought. It covers account ownership, fees, supply quality, targeting, configuration and measurement.
How much does a programmatic media audit cost?
A rapid second opinion can be free or under $5,000, a standard independent audit typically runs $10,000 to $35,000 and a forensic log-level audit can exceed $40,000, based on Stillwater planning ranges. Scope and data access drive the price.
How long does a programmatic audit take?
A rapid review takes one to two weeks, a standard audit three to five weeks and a forensic log-level audit six to ten weeks. Delays usually come from waiting on data access.
What is a good working media ratio in programmatic?
Stillwater planning ranges put a well-run account at roughly 45 to 65 percent of spend reaching publishers. The ANA's Q4 2025 benchmark found quality-led advertisers got 56.7 percent of spend into fraud-free, viewable, MFA-free impressions.
Should I trust a free audit from a competing agency?
Treat it as a useful set of hypotheses rather than a neutral verdict. Ask for the evidence behind every finding, and use an independent auditor for larger budgets or disputed relationships.
