Helen Rose, managing partner, the7stars looks at the enormous potential in the tech shaping our industry, noting that speed isn't the same thing as progress...

For decades, the media industry's transparency debate has centred on money: rebates, margins, principal media, and where advertisers' budgets ultimately go.

Those questions haven't disappeared. But as technology plays a greater role in deciding who brands target, how audiences are constructed, and ultimately where investment is directed, another transparency challenge is emerging.

Advertisers increasingly need to understand not only where their money went, but why a particular planning decision was made in the first place.

That is becoming more important as AI moves deeper into media planning. The industry is racing to automate processes that have traditionally required considerable manual analysis, promising to turn briefs into audiences, recommendations, and media plans in minutes rather than days.

There is enormous potential in that. But speed isn't the same thing as progress.

If technology gets us to an answer faster but makes it harder to understand how we arrived there, we risk trading efficiency for clarity.

More data hasn't necessarily created more understanding

The problem predates AI.

Over the past decade, the volume and sophistication of data available to planners has exploded. We can understand audiences through everything from demographics and attitudes to purchasing, movement, and viewing behaviour.

Yet much of that intelligence remains fragmented.

Different platforms create different definitions of the same consumer. Datasets are interrogated independently. Insights generated in one system don't always translate neatly into another. And increasingly, algorithms sit between the underlying information and the recommendation a planner receives.

The irony is that an industry with access to more audience intelligence than ever can sometimes struggle to explain its audiences simply.

AI could solve some of that complexity. But it could equally compound it.

That's why the next generation of agency technology shouldn't simply be judged by how much of the planning process it can automate. We should also judge it by how much clarity it creates.

Transparency needs to move upstream

Historically, media transparency has largely been something advertisers have demanded towards the end of the process.

What did we buy? What did it cost? Where did the money go? What did it deliver?

Those remain fundamental questions. But advertisers should increasingly be able to interrogate what happens much earlier too.

Why have we defined the audience this way? Which data informed that definition? Why is one group considered a greater opportunity than another? And, where AI has contributed to a recommendation, what information shaped it?

In other words, transparency shouldn't begin at the point of transaction. It should start at the point of decision.

That principle is increasingly influencing the technology we build at the7stars.

Our audience intelligence platform Gravity Connect was originally developed to bring disparate sources of audience and media intelligence together. Its evolution is now increasingly about making the relationships between those signals easier for planners and clients to understand and act upon.

We've recently integrated Samba TV viewing intelligence from 1.8 million UK households into the platform. That means planners can explore the programmes, channels, genres, and times of day that resonate with particular audiences, down to postcode level.

But simply adding another dataset isn't particularly interesting.

The real value comes from connecting that viewing behaviour with broader audience intelligence. Rather than knowing simply that an audience watches a particular programme or genre, planners can interrogate how those behaviours relate to wider characteristics and media consumption patterns.

It is the difference between accumulating data and making it intelligible.

AI should expose the thinking, not hide it

The objective shouldn't be to ask a machine for an audience and accept whatever comes back. The more interesting role for AI is to help planners interrogate information that would otherwise be difficult and time-consuming to connect - surfacing relationships, identifying potential opportunities, and giving planners a starting point they can question and refine.

The media industry is understandably excited about agentic systems capable of automating ever larger parts of planning and buying. But we shouldn't confuse removing friction with removing judgement. The more sophisticated the technology becomes, the more important human interrogation becomes.

A planner needs to be able to challenge why an audience has been recommended. A client should be able to understand which signals contributed to a strategy. And an agency should be capable of explaining the evidence behind a recommendation rather than attributing it to an algorithm.

The goal shouldn't be AI that thinks so planners don't have to. It should be AI that gives planners more to think about.

From information advantage to understanding advantage

This also changes how agencies should think about proprietary technology.

For a long time, the perceived advantage of agency technology was access: more data, more tools, more proprietary information.

But access itself is becoming less differentiating. There is no shortage of data in advertising. The harder problem is connecting it.

The agencies that create the most value won't necessarily be those sitting on the largest number of datasets or deploying the greatest number of AI tools. It will be those that turn complexity into clarity that clients and planners can interrogate and act on.

That requires technology which makes connections visible rather than simply producing outputs. It also requires agencies to resist the temptation to treat complexity as intellectual property. A recommendation isn't more valuable because nobody outside the system understands how it was reached.

Quite the opposite. When advertisers can understand the audience logic, interrogate the data sources, and follow the path from insight to activation, they can make better decisions, challenge assumptions, and invest with greater confidence.

That's not transparency for transparency's sake. It's better planning.

As AI becomes embedded across the media process, our industry's definition of transparency therefore needs to evolve with it.

Advertisers should absolutely be able to follow the money. Increasingly, they should also be able to follow the thinking.