Cross-device targeting is the capability to recognize and reach the same individual user across multiple devices — desktop computers, smartphones, tablets, smart TVs, and connected devices — as a unified entity rather than as separate anonymous users. Without cross-device identity resolution, a user who researches a product on their work laptop, reads reviews on their phone, and makes a purchase on their home desktop appears as three unrelated users to the advertising system. Cross-device targeting closes this gap, enabling coordinated campaign delivery across the full device ecosystem a consumer uses.
The average person now uses 3–4 internet-connected devices regularly. Any advertising strategy that fails to account for cross-device behavior is working with a fragmented, incomplete picture of the consumer journey — leading to wasted frequency, missed attribution, and incoherent campaign sequencing.
How cross-device identity is established
Deterministic identity matching links devices based on known, verified identifiers. When a user logs into the same account (Google, Facebook, Apple ID) on multiple devices, the platform can definitively connect those devices to a single identity. Deterministic matching is highly accurate but limited to logged-in environments.
Probabilistic identity matching infers cross-device connections from behavioral signals when deterministic data is unavailable. IP address correlation (devices on the same household network), browser and device fingerprinting signals, location pattern matching (devices that regularly appear in the same locations), and behavioral sequence similarity are all probabilistic signals that device graph vendors use to estimate cross-device relationships. Match accuracy varies — typically 70–90% confidence — and errors introduce noise into targeting and attribution.
Device graphs are proprietary data products maintained by identity resolution companies, major platforms, and telcos that map billions of cross-device relationships. Advertisers access device graphs through DSP partnerships and data providers to enrich their audience profiles with cross-device connections.
First-party login data is the most reliable source of cross-device identity for advertisers with their own authenticated platforms. An e-commerce brand that requires account login can link all of a customer's purchases and browsing sessions across devices through their account ID, building a complete cross-device behavioral profile from first-party data.
Why cross-device targeting matters
Journey continuity allows campaigns to follow users through a natural multi-device research and purchase process. A user who sees a video ad on their TV can be served a follow-up display ad on their phone, then a conversion-focused ad on their desktop — a sequenced narrative rather than three uncoordinated first exposures.
Accurate frequency management requires cross-device visibility. Frequency capping at the device level will cap a user's phone while their laptop sees unlimited impressions, effectively doubling or tripling intended frequency across a household. Cross-device frequency management enforces caps at the person level, not the device level.
Attribution accuracy depends on connecting touchpoints across devices. A user who clicks a mobile display ad, abandons the page, and later converts on desktop will appear in last-click reporting as a direct or organic desktop conversion — completely miscrediting the mobile ad that initiated the journey. Cross-device ad attribution assigns credit appropriately across the full multi-device funnel.
Audience deduplication prevents advertisers from inflating reach metrics by counting the same person on three devices as three separate users. True unique reach measurement requires cross-device identity resolution.
How AI improves cross-device targeting
Machine learning is central to probabilistic cross-device identity resolution. Neural network models process hundreds of behavioral, technical, and contextual signals to predict cross-device relationships with high confidence — a task that would be computationally intractable with rule-based systems given the billions of daily connections that must be evaluated.
Soku AI applies cross-device intelligence to campaign orchestration, automatically detecting where users are in the purchase journey across all their devices and selecting the optimal device and format for the next ad impression. A user who engaged deeply with a product video on mobile during their commute may receive a more direct conversion-focused ad unit on desktop during work hours — a device-and-context-aware sequencing strategy driven by AI.
AI also continuously updates cross-device probability scores as new behavioral data arrives, keeping identity graphs fresh and reducing the decay that makes static device matching unreliable over time.
Challenges and considerations
Identifier deprecation has significantly degraded cross-device targeting capabilities. Apple's ATT framework limits IDFA availability on iOS devices, and Chrome's third-party cookie deprecation removes a key signal used in probabilistic device graphs. The industry is in active transition toward new identity infrastructure, and cross-device accuracy will remain under pressure during this period.
Household vs. individual confusion affects shared-device households. A family laptop shared by two adults and a teenager may show behavioral patterns that appear to be one person but represent very different purchase intents. Probabilistic device graphs cannot always distinguish household-level device sharing from individual cross-device usage.
Privacy regulation constraints on device fingerprinting and cross-device tracking are tightening. GDPR's requirement for explicit consent before tracking, and growing regulatory scrutiny of device fingerprinting as a consent bypass, restrict how deterministic and probabilistic identity resolution can be conducted in regulated markets.
Data quality decay affects probabilistic device graphs over time. Users change phones, reset advertising IDs, move households, and change behavioral patterns. Identity relationships that were accurate six months ago may be stale today. Continuous graph refresh is essential but resource-intensive.
Walled garden fragmentation limits cross-device visibility. Meta's cross-device data stays within Meta's ecosystem. Google's cross-device data stays within Google. An advertiser running campaigns across Meta, Google, and programmatic cannot build a single unified cross-device view from platform-reported data alone, requiring independent identity resolution infrastructure.
