Ad-tech tools pile up. Most marketing teams add them faster than they retire them, and every new pixel, SDK, or supply partner drags extra cost and risk in behind it. An audit is how you catch leaking budget, unauthorized sellers, and privacy gaps before results start to slip. The good news: the process is repeatable. The less comfortable news: your stack is probably hiding something.
Why an Ad-Tech Stack Audit Pays for Itself
So what is an ad-tech stack audit? It’s a structured review of every platform, vendor, data connection, and supply path that touches your media spend. Three questions drive the whole thing. Where does each dollar go? Who handles your data along the way? What do you actually get back?
Here is the thing: the stakes are measurable. The ANA’s programmatic transparency benchmark estimates that $26.8 billion in global media value is still lost each year to inefficiencies. In an earlier supply chain transparency study, the ANA found that the $88 billion open web programmatic ecosystem held as much as $20 billion in waste, or 23 percent of total investment.
Waste isn’t the only problem, though. An unvetted SDK can leak user data. An outdated tag can quietly break attribution. And an unauthorized reseller might sell you inventory that never ran where it claimed to. A solid audit surfaces all of it in one pass, so you can chase the most expensive problems first.
How often should you audit? Once a year for the full version, plus a lighter quarterly check on spend quality and supply paths. Pull one forward any time you switch agencies, add a major channel like CTV, or sign a new measurement partner.
Step 1: Build a Complete Ad-Tech Inventory and Data Map
You can’t audit what you can’t see, so start with a list. Write down every tool that touches paid media across web and app:
- Buying platforms: DSPs, ad networks, self-serve platforms, and any managed-service resellers your agency uses.
- Supply partners: SSPs, exchanges, and private marketplace (PMP) deals.
- Measurement and verification: mobile measurement partners (MMPs), analytics suites, brand safety and fraud vendors, and attribution tools.
- Data and identity: CDPs, clean rooms, data providers, and identity solutions.
- On-property code: website tags, pixels, server-side containers, and every SDK bundled in your iOS and Android builds.
For each entry, record the owner, contract terms, monthly cost, the data it collects, and where that data goes. Then check the list against what is actually firing. Crawl your site with a tag auditing tool. Ask engineering for a dependency report on each app build.
What we see most often is a gap between what you pay for and what is running, and that gap is where the first wins tend to hide. Think orphaned pixels from old campaigns, two vendors doing the same job, or an SDK nobody remembers adding.
Duplication deserves special attention. Plenty of brands run overlapping verification or attribution vendors that report conflicting numbers. Teams end up arguing over which one is right, and automated bidding gets less reliable too. If you use self-learning campaign optimization, the algorithm is only as good as the conversion signals you feed it.
Step 2: Verify Supply Chain Transparency With ads.txt, app-ads.txt, and sellers.json
Next, confirm that the sellers you buy from are actually authorized to sell that inventory. IAB Tech Lab maintains the standards that make this possible. Its ads.txt and app-ads.txt specifications let publishers declare their authorized sellers, and app-ads.txt extends the standard to apps distributed through mobile and connected TV app stores. Version 1.1 added OWNERDOMAIN and MANAGERDOMAIN values, which let buyers see who owns a piece of inventory and who manages it.
These standards see wide use. IAB Tech Lab’s supply chain validation run covered over 3 million websites using ads.txt and 400,000 apps using app-ads.txt. Adoption isn’t accuracy, though. Tech Lab warns that incorrect or missing entries leave the door open to unauthorized sales.
Here’s how to put this into practice:
- Pull log-level data from your DSP showing the domain or app bundle, the seller ID, and the SSP for every impression.
- Cross-check each seller ID against the publisher’s ads.txt or app-ads.txt file and the SSP’s sellers.json file.
- Flag RESELLER paths where a DIRECT path to the same inventory exists. Each extra hop skims a fee.
- Block unauthorized or unverifiable paths at the DSP level, then move high-value publishers into direct or PMP deals.
On mobile, check that app bundle IDs match the store listings. Spoofed bundles are a known trick for passing low-quality inventory off as premium apps. CTV needs the same care: confirm that app-ads.txt files exist for the streaming apps you buy. This matters even more if you use CTV advertising for app installs.
Step 3: Measure Programmatic Waste and Media Quality
Once your supply paths are clean, the next job is measuring how much of your spend becomes quality impressions. The ANA’s latest quarterly benchmark findings show that advertisers with disciplined quality governance converted 56.7% of programmatic spend into benchmark-qualified impressions, while lower performers converted just 37.5%. “Benchmark-qualified” means fraud-free, measurable, viewable, and not on made-for-advertising (MFA) sites.
Run the same four filters in your own audit:
- Invalid traffic: bots, data center traffic, and spoofed devices. Look at both general and sophisticated invalid traffic reports from your verification vendor.
- Measurability: impressions your verification tag couldn’t measure at all. A high unmeasured rate often points to in-app or CTV environments without Open Measurement support.
- Viewability: impressions that didn’t meet any reasonable viewability threshold.
- MFA exposure: sites built mainly to arbitrage ad revenue, usually with heavy ad clutter and auto-refresh.
MFA deserves a hard look. The ANA Q2 benchmark found that median MFA spend fell to 0.8%, yet top-quartile marketers still spent up to 28.7% on MFA domains. Bottom line: don’t assume you sit at the median. Go check.
Then look at how broad your reach really is. The ANA’s supply chain study found that the average campaign ran on 44,000 websites, and it noted that advertisers can reach a high percentage of their target audiences using a few hundred. Export your domain and app list and rank placements by spend and outcome. Build an inclusion list from the top performers.
Will cutting cheap inventory raise costs? CPMs may go up, but cost per outcome usually comes down. In the ANA’s Q4 benchmark, advertiser case studies found that optimizing toward quality-adjusted metrics delivered nearly 40% reductions in cost per conversion, even when nominal CPMs increased. So judge success by cost per real outcome, not cost per thousand impressions. The same logic applies to creative. Low-quality, high-frequency placements speed up ad fatigue without adding meaningful reach.
Step 4: Audit Data Privacy, SDK Risk, and Mobile Measurement
A stack can run efficiently and still expose you to legal and reputational risk. This part of the audit asks what data leaves your properties, and on what legal basis.
For web campaigns:
- Confirm that your consent management platform blocks tags until users give valid consent, wherever the law requires it.
- Check that consent signals actually reach your DSPs and SSPs instead of stopping at the banner.
- Review server-side tagging setups. They boost performance, but they can hide data sharing from routine browser-based checks.
For mobile apps:
- List every third-party SDK in each build, along with the data it collects and the domains it calls.
- Make sure your App Store privacy disclosures and Google Play Data safety answers match what your SDKs really collect.
- Confirm that iOS tracking only happens after App Tracking Transparency consent, and that your SKAdNetwork or AdAttributionKit setup is configured correctly in your MMP.
- Remove SDKs that duplicate functions or no longer serve an active campaign. Every unused SDK adds app size, crash risk, and privacy exposure.
Measurement health belongs here too. Compare installs and conversions across your MMP, ad platforms, and internal analytics. A gap of a few percentage points is normal. A large or growing gap usually points to broken postbacks, misconfigured conversion values, or attribution windows that don’t line up across partners. If AI agents or automation tools can access your ad accounts, as described in our AI campaign automation guide, document their permissions and log what they can change.
Step 5: Lock In Ad-Tech Governance, Contracts, and Ongoing Monitoring
An audit only matters if the fixes stick, and governance is what turns one-time findings into habits.
Start with contracts. The ANA’s study advises marketers to hold direct data access contracts with their main supply chain partners (DSPs, SSPs, and verification vendors). It also urges them to find out whether their agency acts as an agent or as a principal. That distinction matters because an agency acting as principal has no legal duty to be transparent unless the contract demands it. Review your agreements for:
- Rights to log-level data from DSPs and SSPs
- Full disclosure of fees, rebates, and principal-based inventory
- Audit rights and data retention terms
- Clear ownership of audiences, conversion data, and seat-level access
After that, consolidate supply on purpose. The market is already moving that way. The latest ANA benchmark reports that private marketplaces accounted for over 92% of median spend across environments. In the Q2 report, median SSP count fell from 19 to 17. Curation is fine, but it carries a price premium: the ANA Q2 report put average PMP CPMs at $7.15, compared with $4.41 in the open marketplace. Whatever you keep, hold PMP deals to the same quality and outcome standards you apply to open exchange buys.
Finally, set up recurring monitoring:
- Monthly: invalid traffic, viewability, MFA exposure, and discrepancy reports.
- Quarterly: supply path review, inclusion list refresh, SDK and tag inventory check, and vendor scorecards.
- Annually: a full stack audit, contract review, and comparison against industry benchmarks.
Give each tool and each metric a single owner. Shared ownership tends to end up as no ownership. If your team lacks the bandwidth or specialist knowledge, an experienced mobile growth agency can run the audit and build the governance framework with you.
Frequently Asked Questions
What is an ad-tech stack audit?
It’s a systematic review of every platform, vendor, tag, SDK, and supply path in your advertising setup. The goal is to uncover waste, unauthorized inventory, data privacy risks, and measurement errors. Then you fix them in order of financial impact.
How long does an ad-tech audit take?
A focused audit of a mid-sized stack usually runs four to eight weeks. Getting log-level data from partners is often the slowest part, so ask for it on day one.
What is the difference between ads.txt, app-ads.txt, and sellers.json?
Ads.txt lets website publishers list their authorized sellers. App-ads.txt does the same for mobile and CTV apps. Sellers.json comes from SSPs and exchanges, and it identifies the sellers they represent. Used together, they let buyers verify a supply path from start to finish.
What are made-for-advertising (MFA) sites?
MFA sites exist mainly to monetize ad impressions, often through heavy ad clutter, auto-refresh, and arbitraged traffic. They can post acceptable viewability numbers while delivering little real attention or business value.
Do I need a third-party auditor?
Not always. Internal teams can handle inventory mapping and supply path checks. An independent auditor earns their fee when you need to verify agency fees, reconcile log-level data across many partners, or prepare for a contract renegotiation.
Which KPIs matter most after the audit?
Track the share of spend that reaches fraud-free, measurable, viewable, non-MFA impressions. Pair it with cost per real outcome (installs, purchases, or qualified leads). Those two metrics tell you whether your fixes are working.
Conclusion
Auditing your ad-tech stack comes down to five moves. Map every tool and data flow. Verify authorized sellers. Measure waste against quality benchmarks. Close privacy and measurement gaps. Then lock in governance through contracts and recurring reviews. The takeaway is simple: judge every partner by the share of spend that reaches real, viewable people, and cut anything that can’t prove it.
More from Orad
Sign up to our newsletter
The next guide and the next episode, in one weekly heads-up. No noise.