by David Strnad, Chief Product Officer, OOH, Ipsos

Out-of-Home (OOH) is having a measurement moment, and frankly, it’s overdue. As advertisers rightfully demand the same accountability from OOH that they get from other channels, the market is flooded with aggressive narratives suggesting that more data points, specifically from sensors and mobile devices, automatically equal better measurement. But there is an inconvenient truth that many providers still sidestep: more data isn’t better measurement. It’s not about counting connections; it’s about understanding attention and measuring people.

OOH happens in the real world, on commutes, at curbside, in transit hubs, in malls, in lifts, in moments where people are moving, waiting, distracted, or unusually attentive. It’s not about the phone or any other device; it’s about people-centric behavior. And when you’re investing significant budgets, your measurement needs to reflect this reality. That’s why Ipsos has taken a clear position: OOH measurement must be people-centric first and data-centric second.

The “Ping” Is Not the Impression

Yes, mobility data is valuable, even when it’s streaming in real time. But without the hard work, on-site calibration, validation, real-world visibility logic, and proper conversion into human exposure, it risks becoming the modern-day version of “trust me, it’s big.”

The core problem with device-centric measurement is a confusion of presence with audience. A detector logging a signal tells you a device was nearby; it does not tell you whether a person was attached to it, in the right place, or looking in the right direction. A device signal is not a person, and a ping isn’t an impression. A map full of dots is not an audience currency. Even if you arrange individual measurements in time and build a vector from them, it is still only a very rough approximation of the movement of the device and not of the person.

Clients tell us: “Our dashboard shows millions of ‘exposures,’ but are we actually reaching our target audience? Can we deduplicate with our digital spend? How do we calculate meaningful reach and frequency?” Real-time device counts can’t answer these critical business questions. When a CFO asks about media efficiency, when you’re optimizing cross-channel campaigns, and when you need to defend ROI, you need people metrics, not proximity data, regardless of how quickly that proximity data updates.

Scientific evidence supports this. Research studies highlight that sensor technologies provide proxies for exposure and require modelling and interpretation to estimate real human contact. Relying solely on raw sensor counts invites massive inefficiencies, from double counting across multiple detectors to counting devices moving through transit hubs and curbsides at conditions where the advertising asset is physically invisible.

The Pillars of Validated Measurement

Ipsos is a global leader in OOH measurement. Our OOH Advanced Measurement Systems are designed as full-stack currencies: inventory truth + visibility + mobility + modelling + governance. To move from counting signals to measuring audiences, we leverage advanced technology to measure mobility, and we make sure the data reflects real human behavior, not just digital noise. Our global framework and experience rest on three distinct layers:

1) Inventory Truth

Before processing behavioral data, we must understand the physical environment: where the assets really are, how they face, and where they are seen from. For example, if a screen faces north but traffic flows south, a simple detector will still count passing signals as part of the audience. We don’t. We use an auditable inventory layer and consistent visibility adjustment, rather than a black box approach. Our output is a spatially interleaved Realistic Opportunity to See (ROTS) model, which is equipped with an additional verified probability-visibility adjustment (VA).

2) Agent-Based Modelling

We are pioneering the way we measure people on the move, demonstrating that accurate methodology works across all environments — from screens to streets. Rather than relying on sparse or noisy detector signals, we apply agent-based modelling to build high-fidelity population simulations trained on real mobility, survey, and behavioral evidence, where the only synthetic element is the population agents themselves, while identifiers remain anonymised.

By modelling continuous movement instead of isolated location pings, this approach enables modern reach and frequency estimation, deduplication across environments, and credible planning outputs. It is privacy-safe and designed to be compliant with GDPR, CCPA, and other applicable regulations.

Importantly, the value of preserving movement continuity is also supported in adjacent technical fields. For example, in anonymous video analytics benchmarks, Sánchez-Matilla et al. (2020) show that tracking methods which maintain temporal continuity consistently outperform frame-by-frame detections. While originating in computer vision, this finding reinforces a broader methodological principle: modelling trajectories over time provides more reliable population estimates than momentary observations — directly aligning with our movement-based audience modelling framework.

3) Contextual Intelligence

OOH is uniquely sensitive to context: weather, events, congestion, waiting time, crowd density, local business rhythms, even mood and openness to messaging. These aren’t nice-to-haves; they are the difference between a theoretical contact and a meaningful opportunity to be noticed. We analyze dwell hotspots and mobility friction, and we incorporate seasonality, holidays, and other contextual indicators that shape attention. Because OOH isn’t consumed in a vacuum; it’s consumed while people are late, relaxed, stressed, social, curious, hungry, thirsty, or bored — and those states change the value of exposure. This matters for cost-per-reach, media mix analysis, and proving OOH’s contribution to business outcomes.

Growing the Industry Through Credible Currency

The next era of OOH won’t be won by the fastest raw data feed. It will be won by the most credible currency: calibrated, explainable, people-based, and ready for the way the real world works. Our methodology enables modern reach and frequency modelling and deduplication across all OOH environments — not just volume delivery. We are ensuring data reflects real human behavior. By combining global expertise with scientific rigor, we are building the future of measurement where we value the person, not just the signal.

Credible currencies don’t just help individual advertisers make confident decisions; they reduce wasted spend across the entire industry and unlock OOH’s true growth potential. When measurement can be trusted, defended, and integrated with other media channels, the whole category wins.

Ipsos is building that future now.

Note and reference

– Sánchez‑Matilla et al. (2020). Benchmark for anonymous video analytics. EURASIP Journal on Image and Video Processing (Springer Nature).