What cross-device targeting is, how it works, and why it's getting harder

A relay race only makes sense when you can see all four legs. Watch just the anchor sprint and you'd credit one runner with a team effort, ignoring the three teammates who built the lead that made the finish possible. Cross-device ad targeting works the same way. The same consumer might see your CTV ad on a smart TV Tuesday night, search your brand name on a laptop Thursday morning, and convert on their phone during a lunch break Friday. Three specific devices, one person, one sale, and a journey spread across multiple devices that standard attribution tools often only see the end of.

Getting cross-device advertising right matters more than ever, but understanding how cross-device targeting works—and what's degrading its accuracy—is equally important. Consumers now move fluidly across multiple devices throughout a single customer journey, jumping across multiple devices in a single day. Brands have already been using tools to help them stitch together this experience to reach the same user with targeted ads across multiple devices. But while the identity solutions that help advertisers connect the dots across connected devices have been useful, they're becoming less reliable, and the brands that don't account for that will find their measurement quietly eroding underneath them.

Key takeaways

• Cross-device targeting is the practice of recognizing the same user across various devices—smartphones, laptops, tablets, connected TVs—and delivering a coordinated ad sequence rather than treating each device as a separate audience.
• Advertisers use cross-device identity graphs (XDIDs) to link scattered data points to a single person. The two main approaches are deterministic tracking (verified login data) and probabilistic matching (machine learning inference from signals like IP address and user behavior).
• Cross-device advertising supports three core marketing goals: delivering personalized messages across screens, managing frequency across all a user's devices, and attributing credit more accurately along the customer journey.
• Platform-reported cross-device attribution still tends to favor the last click or last device, which means awareness-driving channels routinely don't get the credit they deserve.
• Privacy changes—including Apple's iOS updates, third-party cookie deprecation, and growing regulatory pressure—are making device tracking less reliable over time, not more.
• Marketing mix modeling (MMM) is a future-proof alternative because it doesn't rely on device-level tracking at all; it uses statistical relationships between spend and revenue to measure impact.
• Prescient AI's MMM operates at the campaign level with daily updates and measures halo effects across channels—including branded search, organic, Amazon, and retail—regardless of what device a customer was on when they converted.

What is cross-device targeting?

Cross-device targeting is the digital advertising strategy of recognizing the same user or household across multiple screens to deliver a seamless, coordinated ad sequence rather than treating each device as an isolated audience. In digital advertising, this means your campaigns can follow a single user from a streaming device in the living room to a laptop at the office to a phone on the commute, serving a consistent message at each stage rather than starting from scratch.

Without cross-device ad targeting, those three interactions look like three different people to your ad platforms (three separate user IDs when it's really one user). That one user might never know how fragmented their data appears to your ad platforms—scattered across multiple devices, multiple sessions, and multiple environments—but marketers know and feel the strain of creating a cohesive experience for these users.

As cross-device marketing has matured, so has the complexity of measuring it accurately. Attribution is just the tip of the iceberg. This complexity affects frequency management, personalized messages, audience segmentation, and just about every other lever in a cross-device campaign.

How cross-device targeting works

The foundation of effective cross-device targeting is the identity graph, a data structure that builds a cohesive identity graph for the same person by linking multiple device IDs, user IDs, and behavioral signals to build a cohesive identity graph for a single person or household. Cross-device graphs generally use two approaches for identifying users and matching device IDs to a single person.

• Deterministic tracking uses verified login data. When a user signs into their Google account, a social platform, or an e-commerce site across different devices, those platforms link all their devices to one verified identity. Deterministic cross-device tracking is the more accurate of the two methods, but it only works for logged-in users, which limits its reach across your full target audience.
• Probabilistic matching uses machine learning algorithms to infer that specific devices belong to the same user. It analyzes signals like IP address, device type, online behavior patterns, location data, and timing to calculate the likelihood that a set of device IDs belong to one person. This approach extends reach well beyond what verified login data alone can cover, but it's making educated inferences instead of reading verified facts.

Most cross-device ad targeting at scale uses a combination of both. As households add more connected devices, device graphs need to keep pace with the expanding set of screens a single user might use. Cross-device ad targeting is only as accurate as the identity graph powering it. Key ad platforms have built substantial infrastructure around these identity solutions, and third-party device graphs from data providers fill the gaps:

Universal IDs are particularly valuable in environments where third-party cookies aren't available; they give advertisers a way to maintain continuity across devices and channels without relying on browser-based signals. The shift toward universal IDs reflects the broader pivot to more durable identity solutions, and many device graphs are being rebuilt around them as third-party signal degrades. Without stable device graphs, the infrastructure that connects user IDs across sessions starts to break down.

Why cross-device targeting matters for your marketing strategies

Effective cross-device targeting unlocks three things that single-device approaches can't deliver well on their own.

Personalized journeys across screens. Your brand gets to tell a continuous story across various devices rather than starting over on each screen. A user might see a brand awareness ad on a connected TV, click a search link on their laptop, and convert via targeted ads on mobile because your ad platforms know it's the same consumer across multiple devices.

Frequency management across all their devices. Without cross-device visibility, frequency capping operates at the device level. That means the same user can see the same ad far more times than you intended, just spread across different devices. Managing frequency across all their devices protects your brand from ad fatigue and keeps your budget from being wasted on over-exposing one user who just happens to have three screens.

More accurate attribution. Cross-device data gives attribution tools a better view of the customer journey across multiple touchpoints than single-device tracking does. The catch is that "more accurate" here means the identity graph is better at linking touchpoints to a single person; the attribution credit still flows through each platform's own reporting, which is designed to reflect well on that platform. Better visibility into the journey and unbiased measurement of that journey are two different things, and cross-device targeting only addresses the first one.

The attribution problem that cross-device targeting exposes

Even with identity graphs in place, most platform attribution still favors the conversion event: the last click, on the last device, before the sale. This is the core attribution challenge that cross-device advertising surfaces: the infrastructure shows you more of the journey, but the way credit gets assigned hasn't caught up. A journey that started with a CTV ad, moved through organic search, and ended on mobile will most likely credit the mobile touchpoint. The CTV ad that put the brand on the customer's radar gets little or nothing.

This matters because cross-device advertising has made the full customer journey more visible, but the attribution models built into most ad platforms aren't designed to distribute credit across that journey fairly. Instead, they're designed to report in ways that reflect well on their own channel. Each ad platform reports on its own performance using its own methodology, and none of them has a clear view of what happened across the others. When you add up the attributed revenue from all your platforms, it almost always exceeds your actual revenue.

Cross-device identity graphs improve the picture somewhat by linking touchpoints to a single user, but platform-reported numbers are still platform-reported, and different devices provide different slices of a larger story that no single platform can tell in full.

Why cross-device tracking is getting harder

The infrastructure that cross-device strategies and identity solutions depend on has been contracting for several years, and the trend isn't reversing. There are several forces at play driving the decline in cross-device tracking reliability:

• Apple's App Tracking Transparency (ATT) framework required apps to ask users for permission before tracking them across apps and websites. Most users opted out. Mobile ad IDs (MAID) became significantly less reliable almost overnight, and cross-device tracking stability took a direct hit. For brands running cross-device tracking across mobile apps, this was a break in the structure and not a temporary dip. Cross-device ad campaigns that depended on MAID-based targeting had to rethink their approach overnight.
• Third-party cookie deprecation has been a slow-moving story, but the direction is clear. Third-party cookies have been the connective tissue of probabilistic cross-device matching for years. As browsers phase them out, the third-party data that cross-device graphs use to stitch user IDs together is disappearing too. Brands whose first-party data infrastructure is strong will fare better, and those relying on third-party data providers will feel the impact most acutely.
• Ad blockers are increasingly common. Over a third of Americans reported using one in 2022, and that number has continued to rise. Every blocker is one more gap in the data cross-device graphs try to stitch together.
• Regulatory pressure—GDPR, CCPA, and similar legislation—continues to restrict the data collection practices that make probabilistic cross-device methods possible at scale. More individual states in the US are considering legislation, too, like the New York Privacy Act.

All of this means cross-device campaigns that worked well in 2019 are running on less complete data than they were then, and that gap will keep widening. Brands that don't plan for this will find their attribution is overstating what they can actually know. At this point, alternative methods to device-based tracking are a practical necessity for brands that want their measurement to hold up.

What accurate cross-device attribution actually requires

Running a successful cross-device campaign is hard enough before you factor in that accurately measuring its impact is one of the harder problems in marketing measurement. Giving proper credit across a multi-device journey requires a few things that are genuinely hard to deliver together:

• A complete, accurate identity graph that links devices without over-inferring (a cross device graph built on strong first-party data, not just probabilistic inference)
• Data from every touchpoint, including channels that don't report through a pixel—CTV, out-of-home, podcast ads—and ensuring your marketing strategies aren't built on a foundation of shrinking signal
• An unbiased measurement methodology that isn't operated by any of the ad platforms being measured
• First-party data infrastructure strong enough to reduce dependence on third-party data as privacy restrictions tighten; first-party data is the most durable foundation for cross-device identity
• Consistency over time, even as privacy regulations and platform policies change

We know that's a high bar. Deterministic data gets you closer to it than probabilistic data, but it only covers logged-in users. Probabilistic inference extends coverage but introduces error. And both depend on an environment that's becoming more restrictive, not less. User-level tracking approaches are limited here: they can only measure what they can see, and what they can see is shrinking. Getting accurate, unbiased attribution for cross-device campaigns requires a measurement approach that doesn't depend on device tracking or platform-reported data at all. 

Where Prescient comes in

Marketing mix modeling (MMM) takes a fundamentally different approach. Instead of trying to track individual users across various devices, MMM uses statistical modeling to identify the relationships between your marketing spend, impressions, and revenue outcomes, no device tracking required. Privacy changes don't degrade the data the model runs on, and the model doesn't depend on reaching a logged-in target audience to function. Whatever happens with third-party cookies, mobile ad IDs, or the next platform policy update, your measurement stays intact.

If you're already using an identity resolution platform, that doesn't change the case for Prescient, it changes the pairing. Identity resolution tools are useful for planning the content pathway your potential customers move through: making sure the creative story is coherent across screens, managing frequency so no one sees the same ad too many times, and mapping out what the messaging looks like at each stage. What they can't give you is unbiased attribution. If you're pairing an identity solution with in-platform reporting today, you're getting better content sequencing but you're still relying on platforms to tell you what worked, and they have every incentive to overclaim. Prescient replaces that reporting layer. You keep the identity tool for planning and sequencing; you lean on Prescient for attribution, forecasting, and optimization that isn't beholden to any platform's methodology. See how Prescient can reveal previously hidden impacts of your campaigns when you book a demo.

FAQs

What's the difference between cross-device targeting and cross-channel marketing?

These terms are related but distinct. Cross-device targeting is specifically about reaching the same user across multiple screens; the focus is on identity and device linkage across a single channel or across several. Cross-channel marketing is about running digital advertising campaigns across different platforms or media types, like paid social, paid search, CTV, and display. A campaign can be both at the same time, but you can also run cross-channel campaigns without any cross-device identity resolution in place. The confusion matters for attribution: assuming your cross-channel data gives you cross-device insight when it doesn't can lead to serious over-crediting of last-touch conversions.

Does cross-device targeting work without third-party cookies?

It depends on the method. Deterministic cross-device tracking—based on verified logins rather than cookies—still works and is generally the more accurate approach. Probabilistic approaches, however, lean on browser-based signals that third-party cookies help facilitate, so cookie deprecation has made these methods less reliable. Ad platforms with large logged-in user bases (like Google and Meta) are less exposed to this shift. Providers that supply probabilistic signal are losing access to the data they've historically depended on. That said, even deterministic data is subject to opt-outs and platform policy changes, so no user-level tracking approach is fully insulated from the broader privacy trend.

How does frequency capping work across devices?

Frequency capping limits how many times a specific person sees a given ad, but without cross-device identity resolution, "person" effectively means "same device." Each device gets its own counter, and targeted ads keep showing up as if it's the first time. A single user switching between a smartphone, laptop, and smart TV can be served the same ad the maximum number of times on each specific device, meaning they've seen it far more often than intended. Effective cross-device frequency management requires a unified identity graph so the cap applies at the person level. The tradeoff is that probabilistic identity graphs introduce some error, so true person-level frequency capping is more of an approximation than an exact science at scale.

How do you measure the ROI of a cross-device campaign?

Measuring the ROI of cross-device campaigns accurately—and attributing it correctly across cross-device ad targeting touch points—is one of the harder problems in marketing measurement. Platform-level attribution will give you a number, but it's subject to the limitations covered above: siloed reporting, last-touch bias, and incomplete data accuracy across devices. A more reliable approach is to use a measurement methodology that doesn't depend on device tracking at all. Marketing mix modeling evaluates the statistical relationship between your marketing investment and your revenue outcomes across the full picture, including channels and touchpoints that never generate a trackable click.