By Charles Lee Mathews. The marketing industry’s fixation on “big data” has created a dangerous blind spot. Without proper context and structure, this data obsession is leading companies down a path of opaque AI-driven decisions that can’t be measured, justified, or trusted.
“Everyone talks about big data, but data as a standalone construct is useless to most people. What’s critical is how one plans for the data to be converted into information that can be used for insights and actions,” says Ryan Sauer of Redwood Analytics, who firmly believes that we need to learn to walk before we can run.
“Many agencies aren’t considering this from a structural perspective because it requires a significant amount of work, effort, and training. However, if businesses and their agencies understood that their entire marketing operation is a moment of data, we could have much better solutions for the boardroom. We could show up much better,” he says.
Sauer says that raw data is useless, but makes the point that insights are valuable. “The knowledge and the wisdom that comes from information is where marketers can own their craft, and this is all about marketing effectiveness,” Sauer says.
Data, data everywhere
“The industry has become overly obsessed with the idea that data is a function of performance and a function of digital. We are all moving towards AI, and we all want to reach the endpoint too quickly. However, we don’t realise that in every moment of media, branding, marketing and advertising, there are data touchpoints we can intentionally create that can be used in the future to enhance marketing effectiveness,” says Sauer.
“What I mean by this is that every job bag, every creative asset, every media campaign name, every radio spot can be tagged or identified as a data touch point,” he says.
Sauer doubled down on his thinking in a recent LinkedIn post. “Data is useless if you use it in its rawest form. I hear marketers at conferences refer to ‘data’ all the time. It’s a trigger word to me, as I feel it does not explain what marketers and organisations are trying to achieve,” he writes.
Can the clichés
What does he hear people mouthing at events and meetings?
- “We have all this data at our fingertips”
- “We live in an age of data overload”
- “Data-driven marketing is driving performance”
- “Big ideas eat big data for breakfast”
Sauer advises that if marketing wants to leverage AI, it first needs to trust and understand the data being used in its business before taking the following steps.
Graphwise, a company that helps businesses use AI more effectively, explains the relationship between data and insights perfectly in the diagram below. The pyramid illustrates how data transforms into wisdom through successive layers of meaning and context.
Graphic via Graphwise.
Starting with raw data, each step (information, knowledge, and finally wisdom) adds value by answering more profound questions about the data. By enriching data with context, media and marketing professionals can gain insights that enable more informed and intelligent decisions. At the top, wisdom reflects the application of these insights to guide action.
Media buying’s ‘black box’
“When you apply data to algorithms or learning models without fully understanding how it was collected or stored, you risk creating what many now call the ‘black box ‘ of AI,” says Sauer, explaining that in this case, AI becomes an opaque system where inputs and outputs are difficult to trace.
“We’re already seeing this in advertising, where tools like PMAX handle performance marketing with little visibility. Many brands and media agencies accept this trade-off — performance at any cost — without questioning how the data is being used or what decisions are being made,” he says.
PMax, or Performance Max, is a goal-based Google Ads campaign type that leverages AI to help advertisers maximise their reach and drive conversions across all of Google’s advertising channels. It combines smart bidding and targeting features into a single campaign, optimising ad performance in real-time.
Context is everything
“This raises a real issue. If data is unprocessed or lacks context, the outputs become just as opaque and hard to interpret. And they become even harder to reverse engineer,” Sauer says. “AI is already a black box for many, and if businesses don’t start understanding the inputs, which are the context and structure of their data, they won’t be able to measure, justify, or trust the outcomes. This is why contextualising data is essential,” Sauer adds.
“You often hear brands and agencies say things like ‘we’re a data-driven business’ or ‘we make data-led decisions’. But in reality, what they’re often describing is the use of retrospective data, looking backwards to make forward-looking decisions. That’s not the whole picture,” he explains.
“True data-driven decision-making means working with historical data, real-time data, and predictive models. It’s about integrating past performance, present signals, and future scenarios. We need to move beyond just hindsight. Real data decision-making requires a full-spectrum view: past, present, and future.”
Charles Lee Mathews is a senior editor to MarkLives MEDIA and a senior writer to MarkLives.com, as well as co-founder of The Writers, a writing consultancy.
