Every time a smartphone user opens an app, a silent auction takes place. Within milliseconds, dozens of advertising exchanges receive a bid request containing the device's advertising ID, its precise GPS coordinates, the app being used, the device model, the operating system version, the user's language settings, and often their browsing history and behavioral profile. This auction happens billions of times per day, across every country where smartphones exist.
For the advertising industry, this data ecosystem exists to serve targeted ads. For law enforcement and intelligence agencies, it represents something else entirely: a global, commercially available surveillance infrastructure that most agencies have barely begun to exploit. This is advertising intelligence — sometimes called ADINT — and it is one of the most powerful and least discussed investigative tools available today.
How the Advertising Ecosystem Creates Intelligence
To understand the investigative value of advertising data, you need to understand how the programmatic advertising ecosystem works.
When a user opens an app or visits a website on their mobile device, the app's advertising SDK sends a bid request to one or more ad exchanges. This bid request is designed to help advertisers decide whether to show an ad to this particular user. To make that decision possible, the bid request contains remarkably detailed information:
- Device advertising ID — a persistent identifier (Apple's IDFA or Google's GAID) that tracks the device across apps and sessions
- GPS coordinates — often accurate to within a few meters, updated every time a bid request is generated
- Device fingerprint — model, OS version, screen resolution, carrier, language, and timezone settings that can uniquely identify a device even without the advertising ID
- App usage — which app generated the request, revealing what the user was doing at that moment
- Behavioral segments — categories assigned to the user by data brokers based on their browsing history, purchase behavior, and app usage patterns
- IP address — which reveals the network the device is connected to
This data is generated continuously, passively, and at global scale. A single smartphone can produce hundreds of bid requests per day. Across the global device population, the advertising ecosystem generates a location and behavioral dataset of unprecedented scope.
Intelligence Applications
The investigative applications of advertising data fall into several categories, each with distinct operational value.
Location Tracking Without Warrants for Cell Data
Traditionally, tracking a suspect's location required obtaining cell tower data from telecommunications providers, a process that involves legal requests, carrier cooperation, and delays. Advertising data provides an alternative path. Because bid request data is commercially available through data brokers and ad exchanges, agencies can acquire location histories for specific device IDs without the legal and procedural complexity of telecommunications data requests.
The location precision of advertising data often exceeds what cell tower triangulation provides. While cell data typically places a device within a few hundred meters, GPS coordinates from bid requests can be accurate to under ten meters. This precision is sufficient not just to establish that a device was in a neighborhood but to determine which building, which floor, and sometimes which room.
Device-Level Behavioral Profiling
The behavioral segments attached to advertising IDs reveal patterns that have direct investigative value. A device associated with cryptocurrency trading apps, encrypted messaging services, and VPN usage presents a different behavioral profile than one associated with mainstream social media and shopping apps. These profiles do not establish criminality, but they provide context that helps investigators prioritize targets and develop leads.
Advertising intelligence platforms can analyze these behavioral signals at scale, identifying devices whose app usage and behavioral patterns match profiles of interest. Combined with location data, this enables identification of devices that were present at a specific location and that exhibit specific behavioral characteristics.
Network Mapping Through Co-Location
One of the most powerful applications of advertising intelligence is network mapping through co-location analysis. By analyzing which devices repeatedly appear at the same locations at the same times, investigators can identify associations between individuals that may not be visible through any other intelligence source.
Two suspects who are careful never to call each other, never to interact on social media, and never to transact financially may still be meeting in person. If their devices consistently appear at the same GPS coordinates within the same time windows, advertising data reveals the association. This technique has proven particularly valuable for mapping criminal networks that practice disciplined operational security in their digital communications but cannot avoid physical proximity during planning meetings and handoffs.
Geofence Intelligence
Geofencing — identifying all devices present within a defined geographic area during a specific time period — is a standard advertising technique used for location-based marketing. For investigations, it provides the ability to answer a critical question: who was at a specific location when an event occurred?
A bombing, a shooting, a drug transaction, a meeting of persons of interest — any event with a known location and time can be investigated by querying advertising data for all devices present in the geofence during the relevant period. The resulting device list can then be enriched with behavioral profiles, cross-referenced against known devices of interest, and analyzed for patterns that distinguish witnesses from bystanders from participants.
Historical Pattern Analysis
Advertising data is archived by data brokers, often for months or years. This means investigators can conduct retrospective analysis: given a suspect's device ID, reconstruct their movements over extended periods. Where do they go regularly? What locations do they visit that deviate from their normal pattern? Which other devices consistently appear in proximity to theirs?
This historical dimension transforms advertising intelligence from a real-time tracking tool into a comprehensive behavioral analysis capability. It provides the kind of pattern-of-life intelligence that was previously available only through sustained physical surveillance.
Why Most Agencies Are Not Using ADINT
Despite its capabilities, advertising intelligence remains underutilized by most law enforcement and intelligence agencies. Several factors explain this gap.
Awareness. Many agencies simply do not know that advertising data is available or understand its investigative applications. ADINT is not taught in most law enforcement training programs and is absent from most agencies' intelligence collection doctrines.
Procurement complexity. Accessing advertising data requires relationships with data brokers and ad exchanges that operate in the commercial sector. Agencies accustomed to procuring intelligence through government channels or established law enforcement data providers may not have the procurement frameworks to engage with adtech companies.
Technical capability. Raw advertising data is voluminous and complex. A single day's bid request data for a modest geographic area can contain millions of records. Without specialized analytical platforms designed to process, correlate, and visualize advertising data, the information is overwhelming rather than actionable.
Legal uncertainty. The legal frameworks governing law enforcement use of commercially available advertising data vary significantly across jurisdictions and are evolving rapidly. Some agencies are cautious about deploying a capability whose legal foundation may shift. However, in many jurisdictions, commercially purchased data does not require the same legal process as data obtained through compulsory means.
Integrating ADINT Into Intelligence Fusion
Advertising intelligence delivers its greatest value not as a standalone capability but as one layer within a multi-source intelligence fusion environment. When ADINT is correlated with other intelligence streams, each source becomes more powerful.
ADINT + Financial Intelligence. A device ID associated with a suspicious financial transaction can be tracked through advertising data to reveal the device owner's movements, associates (through co-location), and behavioral patterns. The financial trigger provides the investigative focus; ADINT provides the operational intelligence.
ADINT + Communications Intelligence. When a target's phone number is known but their movements are not, advertising data for the device associated with that number can fill the gap. Conversely, when advertising data reveals an unknown device at a sensitive location, communications metadata can help identify its owner.
ADINT + OSINT. Open source intelligence can establish a person's online identity and public activities. Advertising data can connect that online identity to physical locations and real-world behavioral patterns, bridging the gap between digital and physical investigation.
ADINT + Video Intelligence. A suspect identified on CCTV at a specific time and place can be correlated with advertising data for the same geofence, potentially identifying their device and enabling ongoing tracking through the advertising ecosystem.
Operational Considerations
Agencies deploying advertising intelligence capabilities should consider several operational factors.
Data latency. Advertising data is not real-time in the way that communications intercepts are. Bid request data is typically available through brokers with a delay of hours to days. For time-sensitive operations, this latency must be factored into operational planning. Some platforms offer near-real-time access through direct exchange integrations, but this is not universal.
Device ID persistence. Apple and Google have increasingly restricted the persistence of advertising IDs in response to privacy concerns. Users can reset their advertising ID, and Apple's App Tracking Transparency framework requires apps to obtain consent before accessing the IDFA. These changes reduce but do not eliminate the utility of device-level tracking, as device fingerprinting and other persistent identifiers continue to function.
Proportionality and oversight. The breadth of advertising data raises legitimate civil liberties questions. Agencies should establish clear policies governing the use of ADINT, including scope limitations, oversight mechanisms, and documentation requirements. The fact that data is commercially available does not mean it should be used without controls.
Vendor evaluation. The quality and coverage of advertising data varies significantly between providers. Agencies should evaluate data freshness, geographic coverage, accuracy of GPS coordinates, and the breadth of behavioral segments before committing to a provider. A proof-of-concept with real operational data is the most reliable evaluation method.
The Untapped Advantage
Advertising intelligence represents a rare convergence: a globally available, commercially accessible, continuously generated intelligence source that most agencies are not yet exploiting. The data is there. The analytical techniques exist. The platforms capable of processing ADINT at investigative scale are operational.
For agencies facing sophisticated adversaries who use encrypted communications, disciplined operational security, and counter-surveillance techniques, ADINT offers a collection avenue that is extremely difficult to evade. A suspect who avoids using a phone can still be tracked through the advertising ecosystem if they carry any smartphone. A network that communicates exclusively through dead drops and face-to-face meetings still generates co-location data through the advertising IDs in their pockets.
The agencies that develop ADINT capabilities now will have a significant and growing advantage over those that do not. As the advertising ecosystem continues to expand, particularly in developing regions where smartphone penetration is growing rapidly, the intelligence value of this data will only increase.
The investigation tool no one is talking about may be the one that makes the difference.