Synthetic data might sound like something from a sci-fi film, but really it’s already been shaping fields as diverse as healthcare and finance for a long time and we’re seeing it becoming more and more a topic of discussion in “mainstream media planning circles”. At its core, synthetic data is information generated artificially by algorithms rather than collected directly from “real people”. It’s designed to mirror the statistical properties and patterns of real data, without being tied to real individuals.
This makes it valuable in areas where privacy is paramount due to GDPR etc, or where collecting enough real-world data is too expensive or impractical. Originally pioneered in statistics and medical research (simulating patient scans to train diagnostic AI’s) it’s being increasingly explored by media agencies and that raises an important question, “What role can synthetic data play in behavioural-led media planning, and what could brands realistically expect from it?”
Where Synthetic Data Comes From
As mentioned, the idea of synthetic data isn’t new. Statisticians in the USA began experimenting with “fake” census records to protect citizens’ privacy while still giving researchers something useful to analyse back in the 90’s. It wasn’t long after that, that medical researchers adopted similar techniques using artificial CT scans, ECGs and other diagnostic records to train algorithms without exposing confidential patient details or if there weren’t enough records to actually make statistical sense of.
What It Is and What It Isn't
Before getting carried away, it’s worth being clear about what synthetic data actually is.
It is artificially generated data that mimics real-world distributions and patterns. Think of it as a “flight simulator for data”. A safe environment to model and test scenarios without exposing live systems or sensitive customer details.
It isn’t a perfect substitute for reality. While it can approximate patterns, it won’t always capture the quirks, edge cases, or outliers that make real human behaviour so complex.
Importantly, it isn’t automatically private either... Poorly designed synthetic datasets can still leak insights about the real people they were based on, so the right safeguards like bias checks, deletion of originals, and independent audits are essential.
Before we get “It’s the year of Mobile” carried away with this, let’s be clear, synthetic data is a tool, not a silver bullet. Used wisely, it can open new doors, used carelessly, it risks creating a model of the world that only exists in theory.
From Healthcare to Media Buying
So what does a medical AI scanning fake X-rays have to do with media planning? More than you might think.
In healthcare, synthetic data allows researchers to train and test algorithms where real-world data is scarce. In media trading and planning world and as part of the House of Communication we can use the principal to try understand audiences or behaviours that are either under-represented in available datasets or hidden behind privacy barriers.
Synthetic data could, for example:
Model the behaviours of emerging consumer segments (say, sustainability-first shoppers)
Create realistic scenarios for how new products might be adopted in unfamiliar markets
Stress-test campaign plans in privacy-restricted environments where third-party cookies or personal identifiers are disappearing.
Synthetic data makes it possible to experiment and plan more flexibly, without always needing vast quantities of personal data up front, and in a world where 1st Party Data is next to impossible to get our hands on, it changes the game a bit for brands willing to be brave.
Why This Matters for Behaviour-Led Media Planning
At Mediaplus UK, our focus has always been on behavioural-led planning. We’re less interested in blunt demographic labels and more interested in the nuanced patterns behind decisions. How people choose, why they change, and where they can be influenced.
Synthetic data supports that mission in three key ways:
New markets – When launching into a market with little historic customer data, synthetic modelling can provide a working behavioural baseline that gets refined once real results come in.
Emerging behaviours – Whether it’s a shift towards plant-based diets or AI-powered shopping assistants, new consumer habits often lack reliable historic datasets. Synthetic data helps us model them responsibly.
Privacy-first planning – With GDPR and the decline of third-party tracking, synthetic approaches can provide a privacy-preserving alternative to filling every gap with personal data.
That said, synthetic data isn’t going to replace real-world observation, the richest insights still come from people and our other tools like surveys, panels, first-party analytics, and lived behaviours, but synthetic data allows us an additional layer, enriching and extending the picture when traditional data is thin.
The Ethical Imperative
For all its promise, synthetic data needs to be handled with care. Algorithms can reinforce the same biases that exist in the original data or worse, “hallucinate” patterns that don’t exist at all.
A model trained entirely on synthetic data might look convincing on paper but fail in the real world.
That’s why the most responsible use is a hybrid approach where we combine synthetic data with anonymised real-world data, testing assumptions against reality, and keeping human judgement at the heart of planning as far as possible.
As an agency, we see synthetic data not as a shortcut but as an ethical way to explore possibilities and help brands innovate without compromising privacy, accuracy, or trust.
The thing is, synthetic data isn’t (as happens far too much in our industry) just a buzzword, from its beginnings in medical research to its growing role in media planning, it offers a way to simulate, test, and imagine behaviours in ways that respect privacy and unlock new opportunities.
It’s also not magic and works best when treated as one part of a broader behavioural toolkit, one that blends real-world insights with smart, ethical modelling.
We believe that’s the sweet spot and using every tool available to understand not just who people are, but why they behave the way they do helps us pin our plans and media trades down to the most effective strategies.
Want to explore how behavioural media planning and responsible use of data can unlock growth for your brand? Contact Mediaplus UK
