Most Advertisers Are Optimizing the WrongSignal
Every major ad platform willtell you the same thing: give us your budget and we will find your audience.Meta has more than three billion active users.¹Google processes over eight billion searches every day.² The inventory is enormous. The promise of algorithmic targetingis that somewhere in all of that activity is the person who wants what yousell.
The problem is not the size ofthe pool. The problem is the signal the platforms use to decide who belongs init.
The dominant currency of digitaladvertising in 2026 is behavioral interest: the algorithm tracks who hasvisited pages like yours, who has clicked on ads in your category, and whobrowses content related to what you sell. It builds an inferred model of yourcustomer and serves your impressions to people who fit that model.
This works adequately when youhave large volumes of conversion data, a long optimization window, and apatient budget. It fails when your product is high-ticket, your sales cycle islong, or your audience is tightly defined. In those cases, the platform isguessing with your money while it learns.
Search intent is a fundamentallydifferent signal. And most advertisers are not using it at all.
The Difference Between Interest and Intent
Interest tells you what someonehas engaged with in the past. Intent tells you what someone is actively lookingfor right now.
A person who reads personalfinance articles fits the interest profile of an insurance buyer. But thattells you nothing about whether they are shopping for coverage today. They mayhave renewed last month. They may be years from the next decision.
A person who searched "besthome insurance quotes" four hours ago is expressing intent. The timing isactive. The need is live. The expected value of reaching that person with an adis dramatically higher than reaching the article reader, not because they aremore likely to be a customer eventually, but because they are more likely to bea customer now.
This timing gap is where most adbudgets disappear. You are paying platform CPMs to reach people who fit theprofile of a buyer, when the signal you actually want is people who arebehaving like a buyer at this specific moment.
Why Search Intent Beats Lookalike andBehavioral Audiences
Lookalike audiences find peoplewho resemble your existing customers. This is a useful tool when you have goodseed data and your product has broad appeal. But lookalike modeling has twostructural weaknesses that search intent sidesteps entirely.
First, lookalike modeling isretrospective. It identifies people who are statistically similar to people whohave already bought from you. It says nothing about whether those lookalikesare currently in a buying cycle.
Second, lookalike accuracydegrades as you scale. A one percent lookalike might be genuinely similar toyour customer base. A ten percent lookalike has been expanded so far from theseed that the statistical resemblance becomes thin.
Search intent solves the timingproblem because the signal is prospective, not retrospective. You are notmodeling who has bought. You are identifying who is currently shopping.
How Intent Data Is Captured and Structured
Intent data at the audiencelevel is captured by associating device identifiers with search query behavioracross the open web. When someone searches for a term within a category, theirdevice identifier is matched to that query and logged.
The IAB taxonomy is commonlyused to classify these behaviors into standardized categories spanningthousands of sub-segments. The resulting data is matched to known identityprofiles, producing either a device ID or a hashed email, and refreshed on a 24-hourcycle.
The lists are uploaded to adplatforms as first-party data. The platform matches your records to useraccounts and delivers exclusively to those users.
The Verticals Where Intent Targeting Has theHighest Impact
Intent-based targetingoutperforms interest and lookalike audiences most dramatically in categorieswhere:
• The purchase decision is high-consideration and notmade impulsively
• The buyer is actively comparing multiple providers
• The cost per acquisition in the category is alreadyhigh
• The search-to-purchase cycle is measured in days, notmonths
Insurance, legal services, realestate, automotive, financial products, higher education, and healthcare allmatch this profile.
Intent Decay: Why Freshness Is a Core Variable
Search intent is time-sensitivein a way that behavioral interest is not. A person who searched for autoinsurance quotes yesterday is in a different decision state than someone whosearched three weeks ago.
Research on consumer decisiontimelines in high-consideration categories suggests that in-market windows forinsurance, financial products, and similar verticals are typically 7 to 30 daysfrom initial search to purchase.³ Intentdata older than a week has materially lower expected value than data from thelast 24 to 48 hours.
Campaigns built on stale intentdata can look like intent targeting while producing results indistinguishablefrom behavioral interest targeting.
Frequently Asked Questions
Is intent targeting only for B2C advertisers?
No. B2B categories such assoftware, professional services, logistics, and manufacturing have significantintent signals available at the device and company level. The key difference isthat B2B intent tends to cluster around longer consideration periods, so thelookback window for intent capture should extend further than B2C.
Does this replace retargeting?
It complements retargetingrather than replacing it. Retargeting reaches people who have already visitedyour properties. Intent-based custom audiences reach people who have expressedcategory-level intent but may never have encountered your brand.
What match rates should I expect?
Platform match rates onwell-sourced intent lists range from 60 to 90 percent depending on the platformand the quality of the identity matching underlying the list. Meta and Googletend to achieve the highest match rates due to their large authenticated userbases.
Sources
1. Meta Investor Relations, Q42024 Earnings Report. investor.fb.com/investor-news/press-release-details/2025
2. Google, How Search Works.google.com/search/howsearchworks
3. Forrester Research, ConsumerDecision Timelines in High-Consideration Categories.forrester.com/report/consumer-decision-timelines-2024
4. Interactive Advertising Bureau, Audience Buying Guide 2026.iab.com/guidelines/audience-buying-guide
