The measurement industry diagnosed the speed problem correctly. They got the root cause wrong. It’s not fragmented data or slow reporting cycles. It’s C-Suite quarterly pressure forcing marketing to chase short-term revenue over long-term value. Measurement won’t fix this. Organizational restructuring will. Most brands won’t do it.
The measurement industry is having a moment of clarity. At the IAB Measurement Summit, something shifted. The old debates. MMM versus attribution. Incrementality testing best practices. Channel-specific measurement philosophy. Those conversations are being crowded out by a more urgent question. How do we make decisions when execution is moving exponentially faster than measurement can keep pace?
That’s progress. That’s also why measurement still won’t solve your problem.
The consensus from the summit is sound. Measurement can’t be perfect. It has to be fast. Outcome ownership needs to be explicit. Data fragmentation is killing optimization. Agentic systems are creating measurement blind spots. All of this is true. And it’s not why most brands are underperforming.
The real constraint isn’t the measurement system. It’s the organizational structure around it.
What the Industry Got Right (And What They Missed)
The IAB is addressing the right problem. Execution speed has outpaced measurement capability. Agencies are running hundreds of creative variants simultaneously. AI is making buying decisions in real time. By the time traditional measurement reports arrive, the campaign has moved on. That gap is real.
Project Eidos, the IAB’s multi-year initiative to standardize measurement across channels, addresses real fragmentation. Harmonizing definitions. Creating common incrementality methodologies. Building interoperable attribution frameworks. This matters. Standardized measurement would reduce reconciliation overhead and increase time spent on actual optimization.
But here’s where the industry’s analysis stops. When they talk about outcome fragmentation, they’re treating it as a behavior problem. It’s not. It’s structural.
Brands know what success looks like. It’s profitable revenue over time. EBITDA. Long-term business value. But C-Suite is asking marketing to hit a unit number this quarter. So marketing briefs agencies to chase awareness metrics. Agencies deliver reach targets because that’s measurable in 90 days. Finance approves because it aligns to quarterly revenue targets. Three months later, the quarter hits but brand equity dropped. Twelve months later, that equity deficit becomes a revenue problem and finance blames marketing for not sustaining growth.
This is how valuable businesses get killed. Not by bad decisions. By quarterly decisions.
The measurement industry treats outcome fragmentation as misalignment between teams. It’s actually structural pressure from the top. When C-Suite demands quarterly unit velocity, it forces the entire organization to chase signals that return tomorrow. Those signals often destroy signals that return in 12 months. Brand building doesn’t work on quarterly cycles. Customer lifetime value doesn’t compound in 90 days.
This isn’t marketing’s fault. Marketing didn’t invent quarterly earnings pressure. They’re reacting to it. Finance created the structure. C-Suite enforces it. Marketing is just trying to survive within those constraints.
The outcome alignment problem isn’t between departments at the brand. It’s between what finance is demanding right now, quarterly revenue, and what actually drives business value. Long-term profitable revenue over one to three years. Those are two different games.
That’s the core issue. You’re optimizing for unit movement when you should be building long-term value. Unit movement is part of the process. But when you sacrifice 12-month profitability to hit this quarter, you’re harvesting the business. You’re not growing it.
The Actual Problem No One Wants to Name
The humans and structures around measurement are built for a different era. Finance teams are governed by quarterly incentives. Marketing doesn’t work on quarterly cycles. Brand building happens over 18 months. Modern media requires thinking on both timelines simultaneously. Most organizations aren’t structured to support that.
There are too many handoffs. Marketing recommends strategy. Agencies interpret it. Tech teams execute it. Each handoff introduces friction and deflection. The media underperforms. The agency blames distribution. Distribution blames strategy. Nobody owns the outcome because nobody defined it upfront.
Agencies are actively making this worse, not on purpose though. They’re simultaneously selling data-driven optimization while delivering commodity services. “AI-native” means they added a feature. “Agents” means they run optimization more frequently. But they’re trained on the same incomplete, fragmented data your organization has always had. They’re just training on garbage faster. You’ve automated failure. Sounds too harsh? Look at how long brands stick with an agency partner before the next RFP. This isn’t a partnership model, it’s people chasing shiny pennies.
An agent optimizes toward whatever signal it can see. If it can only see clicks, it optimizes for clicks. None of that data is clean. Most of it is wrong. It’s all incomplete. But the agent doesn’t know that. It moves money toward what looks optimal in a broken measurement system. You’ve created a very fast way to be systematically wrong.
The worst part. Because you can build campaigns faster, you’re building more campaigns. More variants. More tests. More tuning. From the outside, this looks like rigor. The CFO looks at all this motion, all these optimizations and adjustments and AI-driven sophistication, then looks at the actual P&L and asks one question. If we’re learning faster and optimizing constantly, why isn’t the business actually moving?
That’s because optimization speed and business impact aren’t the same thing. Marketing needs a runway. A campaign needs three to six weeks to build frequency, reach saturation, generate repeat purchase, shift category perception. But organizations are now tuning campaigns weekly. You’re killing campaigns before they have time to work. You’re confusing motion with action and calling it a win because you moved faster than last quarter.
Here’s What’s Actually Happening
Marketing is 80% science now. Maybe 20% art. But organizations are still built for the opposite. They structure around judgment and discretion and the ability to change narratives. Science requires speed, systems, measurement, accountability. Organizations hate that. They prefer ambiguity, but say the opposite; behavior is language.
Most won’t let machines do what machines actually do better. A machine learns that one audience segment drives 40% higher ROI and the marginal ROI shows that incremental investment will drive even more return. The response should be automatic: move the budget there. Instead, organizations ask questions. Run more studies. Get external validation. Hunt for reasons why the machine might be wrong.
This isn’t because machines make better decisions. It’s because humans don’t trust anything they can’t slow down enough to debate.
The companies winning in the near future will let machines do the computational work faster and more holistically. Then they layer human judgment on top of machine output. Not instead of it. Human value lives in strategy and context and creative direction. Not in whether you move $100K left or right in the media mix. Machines are categorically better at that allocation. Not because machines are perfect, but because humans are worse at it, the problem is too complex for a person.
What Actually Needs to Happen
Measurement standardization matters. Project Eidos will reduce overhead. But that’s not where the real work starts. The real work happens here.
Start by defining outcomes with actual honesty. Not what the CEO wants to believe. What actually drives the business forward. Revenue. Profit. Lifetime value. Most organizations have multiple conflicting outcomes sitting on the table. Name them all. Pick one that matters most. Get agreement across finance, marketing, commercial, and sales that you’re optimizing for the same thing. Truly understand causal relationships, if one guy hits his goal and another woman hits her goal, does it lead to a better comprehensive result (spoiler alert, the goals actually counter each other and business stays flat at the P&L level).
Then design the organization around that outcome and most importantly the timeframe you need that outcome. If you’re chasing revenue, you need a completely different structure than if you’re chasing margin. Most companies have org structures left over from when they were optimizing for something else entirely. They’re trying to force new outcomes into old structures and that never works.
Build your decision cadences around when decisions actually need to happen. Weekly for some choices. Monthly for others. Annual for strategy calls. Not around when you finally have perfect data. Perfect data never comes. Design your measurement process for when decisions actually need to flow. Understand that if you need a quarterly unit target and annual revenue target and a three year EBITDA target, you need to look at them all together and not as three different meetings.
Define what humans actually add to this system. Strategy. Context. Creative direction. Not approval gates that slow everything down. If you’re using humans to sign off on budget reallocations after a machine already calculated them, you have a bottleneck, not a safety feature.
Finally, accept that learning happens fast when decisions happen fast. You’ll make mistakes. You’ll move budget to channels that underperform. You’ll miss opportunities. But you’ll learn in weeks, not months. The organizations that beat their competition won’t be the ones with perfect strategies. They’ll be the ones that learn fastest and adjust before the quarter ends; fail forward and fast fast are your friend.
The Real Competitive Advantage
Competitors will implement these same IAB recommendations. They’ll get better measurement. Cleaner data. Standardized attribution. Faster reporting. Some will add agents. And then they’ll do exactly what you do. They’ll use that capability to debate faster instead of decide faster. They’ll have data proving they’re wrong and still choose to be right.
The real differentiator isn’t who builds the best measurement system. It’s who actually uses it. Who trusts it enough to let it guide decisions. Who restructures the entire organization around it instead of forcing outcomes into an existing structure that doesn’t fit.
Most organizations will choose to slow the machines down. They’ll find reasons why the data is incomplete. They’ll commission more studies. They’ll hire another consulting firm.
Then they’ll watch faster competitors take market share and wonder what went wrong.
