×
GlobalIndiaThailandUAEUSA
Back
Back
Head of Marketing - Earned Media
Advertising | DV360
DV360 Audience Analysis is gone, and that signals a bigger...
By Narender Singh
Feb 16, 2026 | 5 Minutes | 
| 
Google pulled the plug on Display & Video 360 Audience Analysis feature in mid 2024 and plenty of advertisers are still figuring out what to do about it.
The tool wasn't perfect, but it was familiar. It gave you a quick snapshot of who you were reaching, how your audiences overlapped and where you might find new pockets of potential customers. Now it gone and the platform wants you to think about targeting in a completely different way.
Whether you loved Audience Analysis or barely touched it, this change matters. It part of a bigger shift happening across programmatic advertising, one that forcing everyone to rethink how they find and reach their audiences.
If you used DV360 regularly, you probably spent time poking around in Audience Analysis. The feature showed you demographic breakdowns, interest categories, device preferences, all the usual stuff. But the real value was in the overlap analysis.
You could see how your selected audiences intersected with other available segments on the platform. That made it easier to spot similar groups worth testing without just guessing or adding random audience lists to your line items. For prospecting campaigns especially, that visibility helped you expand reach while maintaining some confidence you weren't burning budget on completely irrelevant impressions.
The tool worked for any line item that had served at least 1,000 impressions. Pretty low bar. You could check performance, spot trends, make adjustments. Simple stuff, but useful when you're managing dozens of campaigns and don't have time to dig through three different reporting interfaces.
Here the thing about third party cookies: they're dying. Everyone knows it. Google knows it better than anyone since they control Chrome and they've been telegraphing this shift for years now, even if the timeline keeps changing.
Audience Analysis relied heavily on cookie based data. As that data source becomes less reliable and less available, tools built on top of it stop making sense. Google decided to streamline rather than prop up features that won't work properly in another year or two.
The official sunset date was July 1, 2024. The feature became completely inaccessible shortly after that. Google gave advance notice, sure, but "advance notice" and "enough time to completely rework your audience strategy" aren't always the same thing.
Some advertisers saw this coming and started adjusting their workflows months earlier. Others got caught flat footed and spent July scrambling to figure out alternative reporting methods.
DV360 doesn't treat audiences the same way it used to. The platform has moved toward using audiences as signals rather than rigid targeting parameters.
What the difference?
When you add an audience as a signal, you're not saying "only show my ads to these people." You're saying "these people represent what good performance looks like, now go find more people like them." The machine learning algorithms take those signals and look for patterns, behaviors, characteristics that indicate someone might convert even if they're not on your original list.
This works particularly well when you enable Optimized Targeting. The feature actively expands beyond your selected audiences to find additional users who look like they'll perform well based on your campaign goals. You're giving up some control in exchange for better performance and broader reach.
Not everyone loves this approach. Some advertisers prefer knowing exactly who sees their ads. But that level of control was always somewhat illusory anyway and it becoming less viable as privacy restrictions tighten.
So Audience Analysis is gone. What now?
Your own data just became exponentially more valuable. You should already be collecting it aggressively through every channel available.
Google Analytics 4 connects directly to DV360. Set up proper tracking on your site, build audiences based on user behavior and import those segments into your campaigns. High engagement visitors, people who viewed specific product categories, users who abandoned carts. All of it can feed into your targeting strategy.
Customer Match lets you upload hashed email addresses and phone numbers to target existing customers or exclude them from acquisition efforts. If you have a CRM system, you should be using this. The match rates aren't always great, but the users you do reach are already warm leads.
Floodlight tags give you even more granular control. You can build remarketing lists based on specific actions, tag product categories, track micro conversions. The setup takes some technical work, but it pays off when you're trying to optimize mid funnel performance.
Remember when everyone thought contextual targeting was old fashioned? Turns out it works pretty well when you don't have reliable user level data.
Instead of following people around based on their browsing history, you're serving ads based on the content they're actively consuming right now. Someone reading an article about home renovation sees ads for power tools. Someone comparing mortgage rates sees ads for real estate services.
The relevance is obvious and you don't need to track anyone personal information to make it work. Privacy regulations don't scare you as much when your targeting method doesn't depend on user tracking in the first place.
Plus, contextual targeting has gotten smarter. Modern systems analyze sentiment, context, brand safety, all kinds of factors beyond simple keyword matching. It not your 2010 version of contextual anymore.
DV360 still offers plenty of audience segments that function without third party cookies. Affinity audiences, in market audiences, demographic targeting based on life facts like homeownership or household income.
You can build custom segments using keywords, URLs and apps that align with your target customer profile. Life events targeting lets you reach people during major milestones: moving, getting married, graduating, having kids.
These aren't as precise as custom audiences built from your own data, but they're reliable and they scale. For broad awareness campaigns or when you're entering a new market, they get the job done.
Optimized Targeting in DV360 uses machine learning to automatically expand your reach beyond whatever audiences you've selected. It analyzes performance in real time and identifies new users who share characteristics with your converters.
This freaks out some advertisers who want complete control over their targeting. But here the reality: you were never going to manually identify every possible converting user anyway. The algorithm can spot patterns you'd miss and find users you'd never think to target.
The catch is you need to give it good signals to work with. Feed it quality first party audiences, set clear conversion goals and give the campaigns enough budget and time to learn. If you're too restrictive with your initial parameters or don't provide enough conversion data, the algorithm can't optimize effectively.
Losing Audience Analysis meant losing visibility into how different audience segments performed. That visibility gap is real, but it not insurmountable.
DV360 standard reporting dimensions still show plenty of useful data. The Optimized Targeting dimension specifically tells you how much of your performance comes from your seed audiences versus algorithmic expansion. If the algorithm is finding good users outside your original parameters, you'll see it there.
Focus on the metrics that actually matter: conversion rate, ROAS, CPA, incremental reach. Compare performance across different audience signal combinations and targeting strategies. The granular demographic breakdowns were interesting, but what you really need to know is whether your campaigns are hitting their goals.
Run continuous tests. Try different audience combinations, experiment with contextual layers, adjust your Optimized Targeting settings. Testing reveals what works for your specific campaigns and business goals better than any pre built analytics dashboard.
You might also want to consider third party measurement solutions if you need deeper audience insights. They cost money, obviously, but they can fill some of the gaps left by Audience Analysis removal.
The deprecation of Audience Analysis isn't an isolated incident. It part of a fundamental restructuring of how programmatic advertising works.
By the end of 2025, DV360 had already expanded its Audience List Type dimension from three categories to ten. Commerce audiences, lookalike audiences, demographics, affinity, in market, custom segments and more. The reporting got more granular even as some tools disappeared.
Google has introduced privacy focused solutions like PAIR (Publisher Advertised Identifier for Reach), which lets advertisers and publishers match first party data without cookies. Early testing showed PAIR audiences reaching 11% more potential customers compared to traditional cookie based methods. That not a massive improvement, but it movement in the right direction.
The platform is also integrating generative AI into audience creation. You can describe your ideal customer and let AI suggest relevant segment combinations. The suggestions aren't always perfect and you still need to vet them, but it speeds up the process considerably.
The shift away from Audience Analysis doesn't have to tank your campaign performance. Plenty of advertisers are seeing better results with the newer approaches once they get past the learning curve.
Start by auditing your first party data collection. Are you capturing enough user behavior data? Is your tracking implementation solid? Can you expand what you're collecting without creeping people out? Strengthening your data foundation should be priority number one.
Then experiment. Test contextual targeting alongside your first party audiences. Try different custom segments based on keywords and interests. Give Optimized Targeting room to work while monitoring how far it expands beyond your core audiences.
Document what you learn. Keep notes on what drives results and what doesn't. Test results from six months ago might inform your strategy two years from now when the next big platform change happens.
The advertisers who thrive aren't the ones resisting every change. They're the ones who adapt quickly, test aggressively and don't get too attached to any single tool or tactic.
Losing Audience Analysis feels like a setback if it was central to your workflow. But the programmatic advertising landscape has always been unstable. Platforms change, privacy laws evolve, user behavior shifts, new technologies emerge.
What worked two years ago might not work today. What works today will need adjustments tomorrow. That just the nature of digital advertising. You either accept that reality and build systems that can adapt, or you spend your career frustrated that things keep changing.
The removal of Audience Analysis is pushing advertisers toward more sustainable targeting methods. First party data, contextual signals, machine learning optimization. These approaches respect user privacy while still delivering solid performance and they're not going to break the next time a browser updates its tracking policies.
Google will keep evolving DV360. More features will get deprecated. New capabilities will launch. Privacy regulations will continue tightening. The advertisers who stay ahead of these changes instead of reacting to them after the fact will find themselves with a significant competitive advantage.
Your campaigns can perform just as well, probably better, without Audience Analysis. You just need to stop expecting the platform to work the way it did three years ago and start learning how it works now.