At Unacast, we are always listening for what’s on adtech buyers’ minds. In this post, we’ll share the key trends we’ve noticed in meetings with the platforms, agencies, and data teams we work with every day.
Over the past several months, the same questions kept coming up. How do we work with location data more efficiently? How do we build better audiences faster? How do we know the devices we're reaching are real? And how do we stay compliant as privacy rules keep shifting?
Here are the five trends we see defining adtech in Q4 2026, and what they mean for teams planning for 2027.
1. Data flexibility is the new differentiator
Buyers want data that fits their stack, not the other way around. The most consistent theme we heard: teams want a partner whose data can be a reliable ingredient in a product they're building themselves.
That shows up in three ways:
- Blending sources. Adtech teams most often combine location data with their own first-party customer data and with transaction data.
- Pulling only what they need. APIs let teams query data on demand instead of storing and ingesting an entire country's worth of records.
- Choosing the delivery model. Some teams want on-demand API access. Others want bulk delivery through data-as-a-service. Many want both, for different use cases.
Our Visitations API was built for this shift. Beyond running analysis on custom polygons a client defines, it lets teams pull brand-level visitation insights in a format that fits their own platform. The feedback from recent evaluations has been consistent: on-demand access is simply easier for teams to integrate into existing workflows.
The pattern isn't limited to adtech. We see the same demand from commercial real estate firms and site selection tools that want location data as a building block, not a finished report.
2. AI is moving audience building in-house
The biggest AI use case we see right now isn't a chatbot. It's customers building their own large language model (LLM) tools on top of our data, so any user can type a request like "build me an audience for this brand" and get a usable segment.
That puts audience creation in the hands of non-technical users, too. A planner or account manager can describe the people they want to reach, and the tool does the querying.
The wider market is moving the same way:
- The Trade Desk and Dstillery are linking their AI agents so buyers can build custom audiences in minutes. The manual process typically took about four days (Adweek).
- LiveRamp has proposed Agentic Audiences, an open standard hosted by IAB Tech Lab for how advertising agents exchange identity and contextual signals.
The AI layer is only as good as the data underneath it. For an LLM or agent to build a strong audience, it needs data it can query on demand, at the right level of detail, in a structure it can work with. That's where flexible delivery (trend 1) and AI meet.
Not every team wants to build its own tools. For those that don't, off-the-shelf persona audiences are still the fastest path from idea to activation.
3. Buyers want proof they're reaching real people
Skepticism about adtech is running high. Buyers are asking a blunt question: how many of the devices we pay to reach are real?
The research gives them reason to ask:
- Only 43.3% of programmatic spend in Q1 2026 delivered fraud-free, measurable, viewable impressions on sites that weren't made for advertising, according to the ANA Programmatic Transparency Benchmark.
- In a 2025 test by CTV security firm CleanTap, counterfeit devices built from Raspberry Pi computers and emulators were accepted into live auctions 100% of the time. They were served ads from 54 brands (TV Tech).
At the same time, Apple is cutting off signals many teams relied on. Since September 14, 2026, iOS 27 has blocked Safari and other WebKit browsers from reaching domains tied to identity and data providers (PMG).
The practical effect: it's harder to connect what someone browses across websites with what they do in the real world. Strategies built on cross-site browsing behavior feel this most.
That's why identity resolution and hashed email (HEM) intelligence keep coming up in our conversations. An added identity layer gives context about a device: whether it ties to a real person with a real history, rather than a synthetic or never-before-seen identifier. The goal is simple: fewer synthetic impressions and more confidence in reach. (Learn more about identity data providers.)
4. Privacy compliance is now table stakes
More states passed privacy laws in 2026 than many people expected. Alabama, Louisiana, Oklahoma, and Vermont enacted comprehensive consumer privacy laws (MultiState).
The rules are also getting stricter. Connecticut, Virginia, and Oregon now prohibit the sale of certain sensitive data, including precise geolocation data. Maryland and New Jersey have broader bans on selling sensitive data (Regulatory Oversight).
California went the other way. Governor Newsom vetoed AB 1542, which would have banned selling or sharing sensitive personal information with third parties (Consumer Reports). He called a categorical ban "a step too far" (Regulatory Oversight).
For many of our customers, that veto was a relief. The lesson for 2027: compliance can't be a fire drill every time a new state law passes. At Unacast, privacy protections are built into how our data is processed before it ever reaches customers. That lets teams focus on their use cases instead of tracking every statehouse.
5. Growth is going global
More of our customers are evaluating markets outside the US, including teams whose business has been US-first until now.
When they expand, they want to bring the same data, delivery options, and workflows into each new market. Rebuilding a data stack country by country slows launches and makes results harder to compare across regions.
That’s where a global data provider helps. Unacast data supports international use cases across our product lines, so customers can test and grow in new markets without starting over with a new vendor.
What this means for your 2027 data strategy
These five trends point to one shift: buyers are judging data partners on flexibility, trust, and reach, not just volume. If you're evaluating location or identity data for next year, these are the questions worth asking.
Looking ahead to 2027
We expect these trends to speed up, not slow down. AI agents will take on more of the audience-building work. Browser-level identity signals will keep shrinking. And the state privacy map may keep filling in.
The teams that come out ahead will work with data that is flexible enough to plug into their own tools, verified enough to trust, and compliant by design. That's what we're building at Unacast.
Want to see how our location intelligence APIs fit into your workflow? Book a meeting with us.
