We love to blame walled gardens for grading their own homework. Fair enough. But everyone does it. Agencies, brands and platforms configure measurement based on their objectives. Platforms define engagement and attribution in ways that emphasize value, while agencies and brands structure reporting based on what they can optimize or defend. The result is multiple accurate views modeled based on how each system is configured.
Marketers configure systems differently and create their own categories and definitions based on their needs. Then they blame the industry when the data doesn’t reconcile. They spend too much time pointing out what they can’t control while ignoring what they can.
Marketers don’t control every platform, but they do control the requirements they set and the partners they fund. As you plan for 2027, three things need to change.
1. Build the campaign to support MTA, MMM and incrementality
Your campaign setup limits what measurement can tell you.
Most teams inherit configuration practices and rarely question them. They activate as usual without asking if the campaign will support the business questions that matter.
Problems appear later. Campaign and placement names become strings of code that no one outside of the team that created them can decipher. Creative file names reveal little about the message, format, audience, or version. Tracking pixels, click trackers, and site-served creatives can leave gaps in what ran, where, and what creative was seen.
Multi-touch attribution (MTA) requires actionable exposure, interaction and conversion signals. Marketing Mix Modeling (MMM) requires stable and comparable inputs over time. Incrementality requires a valid comparison designed before media execution. If these requirements aren’t integrated before launch, the data won’t support the questions marketers want to answer later.
Recommendation: Control how campaigns are set up, activated and tracked across platforms, agencies and teams. Examine identifiers, naming rules, conversion definitions, attribution settings, and tracking configurations against the questions MTA, MMM, and incrementality need to answer. Correct structural gaps.
10x your SEO with Semrush for business.
The most powerful SEO platform in the world, created specifically for businesses.
Request a demonstration
2. Harmonize measurement with a common language
When I started Project Eidos at IAB, one of the first priorities was to develop a common taxonomy for advertising inventory and ad types. The agencies said, “We already have a taxonomy.” Then they admitted to maintaining a different version for each customer.
Everyone labels things their own way, and analysts spend hours translating before the analysis can begin.
“We need to customize it to align with the way the customer wants to see it” seems reasonable. But when each customer has a different language, personalization becomes fragmentation. A company may need its own reporting views, categories and levels of detail. These can sit on a common structure. They shouldn’t replace it.
The IAB and IAB Tech Lab have developed campaign data standards and other taxonomies and technical standards covering audiences, content, ad products, inventory and measurement signals. They are not one-size-fits-all and will not complement every data set. But they offer marketers, agencies, publishers and other media owners, platforms and technology providers a common basis for describing and exchanging information.
But standards won’t solve fragmentation if only part of the supply chain uses them. Analysts still have to translate the same information, and the results remain difficult to compare.
Artificial intelligence makes this more urgent. It can combine five different definitions of engagement without knowing whether they represent the same behavior. Unless these definitions and calculations are documented and mapped to a common structure, AI can produce a confident response based on mismatched inputs. Agents need that common language too. Without it, they can’t reliably compare inventory, recommend allocations, or evaluate results.
Recommendation: Treat applicable industry taxonomies and technical standards as reference points for campaign configuration, reporting, and partner data mapping. Add custom fields only where necessary and maintain a clear mapping of the common structure. Require partners to disclose which standards and versions they support, along with any changes or omitted fields. Build these requirements into onboarding and contracts and regularly review compliance.
3. You want transparency and accuracy. Prove it.
Every marketer says they want transparent, independently reviewed measurements. If it’s so important, why isn’t it a routine requirement when partners are selected and renewed?
Certification, accreditation and independent audits can be expensive and time-consuming. Avoiding this work also comes at a cost: Companies apply different methods and filters, then report results under the same metric name. If you’re wondering why measurement stays stopped, start there.
Following an industry standard makes a statement. The external review provides evidence that the relevant implementation or service has been evaluated. Such review may take the form of technical compliance certification, measurement accreditation, or other market- or service-specific assurance.
As of August 2026, the Media Rating Council Digital Accreditation Directory it contained 40 digital service listings from approximately 25 providers. Now compare that to the number of platforms marketers use and the metrics they rely on. Yet marketers continue to fund widely used platforms and rely on their metrics. This is a market failure, not a technical one.
Recommendation: Don’t take “we follow industry standards and guidelines” as an answer. Require partners to document how their counts are calculated, where their methods differ, and exactly what has been independently reviewed. Make applicable certifications, accreditations and other forms of independent assurance part of your partner selection and renewal, or require a clear timeline for obtaining them. If independent measurement is important, fund companies willing to demonstrate their work.
Stop applying digital tape
If nothing changes, next year we’ll blame AI instead of signal loss. Give the AI bad data and it will increase the problem, so it will provide the answer with confidence.
Marketers don’t need to become engineers, auditors, or media operations specialists. But they need enough technical knowledge to ask better questions about campaign setup, partner standards, and what is actually certified or accredited. They must also stay current as standards evolve and interact with the groups that shape them.
This requires time, expertise and investment. But marketers should fund the foundation before purchasing another tool that simply addresses its weaknesses. Otherwise, it adds another layer of digital tape.
You decide which partners renew, which platforms receive next year’s budget, and whether existing standards are actually enforced. You’re writing the check. Set the terms.
Fix the configuration. Unify the language. Request an independent review where it matters. So measurement can move from explaining why numbers don’t work to using them to make better decisions.
