iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Lookalike segments start with your first-party seed audience and find new people who resemble it. Optimized targeting starts with audience signals but allows Google’s models to search beyond them for people it predicts are likely to convert. They are different controls, not automatically competing ones: Google says they can work together. The important 2026 caveat is that Demand Gen Lookalikes are moving to a suggestion model, so a Lookalike may no longer be a hard similarity boundary in every account.
Use the seed-based option when the quality and relevance of your customer, visitor, app, or YouTube-engagement list should shape prospecting. Use Optimized targeting when conversion volume or cost per action matters more than staying close to a manually defined audience. Check your account’s rollout and constraint status before promising a specific reach limit.
Lookalike segments and Optimized targeting at a glance
The two features answer different planning questions. A Lookalike asks, “Which new users resemble this valuable first-party audience?” Optimized targeting asks, “Who else, beyond these audience signals, is most likely to produce the campaign’s conversion?” Google’s Demand Gen FAQ describes the distinction as similarity-based discovery versus performance-oriented delivery: “Lookalike segments help you drive similarity based audiences and optimized targeting helps you drive performance by showing ads to people who are most likely to convert.” Google Ads Help
| Dimension | Lookalike segments | Optimized targeting |
|---|---|---|
| Starting point | A first-party seed, such as a customer list, website or app activity, or YouTube engagement. | Optional audience segments selected by the advertiser as signals. |
| Main job | Discover people similar to the seed audience. | Find people the model predicts are likely to convert, including users outside the selected signals. |
| Advertiser control | Historically included Narrow, Balanced, and Broad similarity choices; in the 2026 suggestion rollout those choices guide prioritization rather than necessarily limiting delivery. | Audience selections are starting signals, not guaranteed boundaries. |
| Best fit | You have a relevant, high-intent first-party list and want its characteristics to influence prospecting. | The campaign is judged primarily on conversions, conversion volume, or cost per action and you accept broader model exploration. |
| Relationship | Can be used with Optimized targeting when the campaign setup and objective support both. | Can use a Lookalike as one of the signals while continuing to search beyond it. |
Neither label proves who received an impression, and neither feature guarantees incremental lift. Judge the setup with your conversion reporting, audience reporting, and business objective.
#1 Best Overall
What a Demand Gen Lookalike segment does
It begins with a first-party seed
A Lookalike segment is built from an audience you already own or can lawfully use in Google Ads. Typical seeds include recent purchasers, qualified leads, website visitors, app users, or people who engaged with your YouTube content. The closer the seed reflects the action you want, the more useful its similarity signal is likely to be. Google’s rollout guidance recommends high-intent lists that closely match the campaign goal. Use Lookalike segments to grow your audience
For example, a seed of recent paid subscribers is a more direct prospecting signal for a subscription campaign than a large list containing every site visitor, including people who bounced immediately. A lead-generation campaign can use qualified or sales-accepted leads rather than every form submission if the latter includes substantial low-intent traffic.
Similarity was historically expressed as reach levels
Under the older constraint behavior, Demand Gen offered three reach choices. Narrow aimed at approximately 2.5% of users in the target location, Balanced at 5%, and Broad at 10%. Those figures described the intended share of users within the location, not a promise that every impression would meet a measurable demographic or behavioral profile. Google’s constraint documentation records these levels for the legacy behavior. Use Lookalike segments in Demand Gen campaigns as a Targeting Constraint
Free tools Windows power users keep installed
One-click scans. No signup required.
These percentages must not be presented as current hard limits by default. During the 2026 transition, the selected level can act as a model signal rather than a fence around delivery.
What Optimized targeting does
It treats audience selections as signals
Optimized targeting uses the audiences or segments you select to inform Google’s models, then looks beyond them for users it predicts are more likely to convert. The selected audiences are therefore a starting point, not necessarily the complete eligible population. Google describes the feature as a way to pursue performance by finding people most likely to convert. Demand Gen Frequently Asked Questions
If you add a custom segment, customer list, or interest-based segment and enable Optimized targeting, the system can use patterns in conversion data and user context to explore users who were not in that manually selected group. This is useful when the campaign has a clear conversion event and enough reliable feedback for the model to learn from; it is less suitable when you must restrict delivery to a narrowly defined membership list.
It is not the same as audience expansion in every campaign
Google documents audience expansion as a related feature for specified Video campaign goals. Do not substitute that term for Demand Gen Optimized targeting: the controls, campaign types, and documented behavior are not interchangeable. About audience expansion
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →The practical differences that affect campaign decisions
Seed quality versus conversion prediction
A Lookalike’s strongest input is the seed. If the seed contains the people most representative of your desired outcome, similarity-based discovery has a clear direction. If it is stale, tiny, mixed with unqualified contacts, or collected under a different business objective, the resulting signal can be weak or misleading.
Rank #2
Optimized targeting’s strongest input is the conversion objective and the data available to evaluate it. It can move away from the selected audiences when the model estimates that another user is more likely to complete the conversion. That can expand opportunity, but it also means the delivered audience may look unlike the manually chosen segments.
Audience control versus model exploration
Choose a Lookalike when you need the seed’s characteristics to remain central to prospecting. Choose Optimized targeting when exploration is acceptable and the campaign’s success metric rewards additional conversions or an efficient cost per action. Neither choice removes the need to define location, language, exclusions, consent practices, conversion actions, and budget correctly.
Interpretability and reporting
A Lookalike gives a clearer explanation for why a user was eligible: the user was modeled as similar to the seed. Optimized targeting gives a clearer explanation for the optimization goal: the system estimated a higher likelihood of conversion. These are explanatory frames, not user-level guarantees. In the 2026 suggestion mode, Google says the audience reporting table marks Lookalikes as a “signal,” which is a warning not to read the label as proof of a strict similarity-only impression.
Campaign objective
When the business question is “Can we reach more people like our best customers?” a Lookalike is the natural starting test. When it is “Can we obtain more conversions at an acceptable CPA?” Optimized targeting is the more direct fit. If both questions matter, test them in a structure that keeps messages and reporting understandable instead of changing several controls in one undifferentiated ad group.
Can Lookalike segments and Optimized targeting work together?
Yes. Google’s FAQ says the features can work well together. A Lookalike can supply a strong first-party signal while Optimized targeting searches beyond that signal for additional likely converters. Demand Gen Frequently Asked Questions
Combining them is most defensible when:
- The seed represents the conversion you actually value.
- The campaign is optimized to a measurable conversion rather than a weak proxy.
- You are comfortable with delivery outside the seed’s former similarity band.
- Your budget and reporting can distinguish the combined setup from a seed-only or signal-only setup.
If the audiences need different creative, bids, exclusions, or landing-page messages, put them in separate ad groups. Google’s Demand Gen audience guidance recommends separate ad groups when distinct audiences require different messages or targeting strategies, which keeps management and performance reporting clearer. Demand Gen audiences overview
Do not treat two ad groups as a laboratory guarantee. Auction eligibility, budget allocation, conversion lag, and other campaign settings can affect the comparison. Use the structure to make decisions and reporting clearer, not to claim a result before the campaign has accumulated comparable data.
The 2026 Demand Gen Lookalike transition
What suggestion mode changes
Google is phasing Demand Gen Lookalikes into suggestion mode during 2026. The rollout became available to advertisers starting in March 2026, with opt-out controls expected later in the year. Because the rollout is phased, two accounts can show different behavior at the same time. Use Lookalike segments to grow your audience
In suggestion mode, your seed list remains an input to Google’s models. The system may reach qualified users outside the former similarity thresholds to pursue goals such as conversions or CPA. Google’s guidance says the seed acts as a signal, not necessarily as a delivery constraint. The selected Narrow, Balanced, or Broad setting likewise guides prioritization; it no longer guarantees that impressions stay inside the corresponding 2.5%, 5%, or 10% band.
What constraint mode means
Google also documents a constraint route for advertisers who prefer the former behavior. Where the account supports it and the advertiser opts in or remains on that mode, the selected similarity level functions as a targeting constraint rather than merely a suggestion. The availability of an opt-out control is account- and rollout-dependent, so do not promise that every advertiser will see the same switch. Use Lookalike segments in Demand Gen campaigns as a Targeting Constraint
How to verify your account
- Open the relevant Demand Gen campaign and its audience or ad-group settings in Google Ads.
- Inspect the Lookalike audience entry and the accompanying reach, signal, or constraint wording.
- Check the audience reporting table for the status Google assigns to the Lookalike. Suggestion-mode reporting identifies it as a signal.
- Look for the documented constraint or opt-out control. If it is absent, do not assume that the account can enforce a legacy similarity boundary.
- Record the account state and date in your campaign notes so later performance comparisons use the correct interpretation.
These checks are necessary before describing a campaign as “Lookalike-only” or claiming that a slider limits delivery to a fixed percentage.
Quick wins for a faster PC:
Clear out junk files and repair common Windows errorsFree Scan →Scan for outdated or missing drivers - takes under a minuteDriver Scan →How to choose the right setup
Choose a Lookalike-led setup when
- You have a recent, consented first-party list tied closely to the desired outcome.
- Your prospecting message is intentionally designed for people resembling purchasers, qualified leads, active users, or engaged viewers.
- You want the seed audience to be the primary strategic explanation for expansion.
- You can tolerate the possibility that suggestion mode will go beyond the former similarity thresholds, or your account supports the constraint behavior you require.
Choose Optimized targeting when
- Conversion volume or CPA is the primary success criterion.
- You have a clear conversion action and enough dependable conversion feedback for automated optimization.
- You are willing to let Google search beyond manually selected audiences.
- The campaign can absorb broader discovery without violating geographic, regulatory, brand-safety, or customer-exclusion requirements.
Use both when
- The seed is a valuable source of direction but you also want performance-oriented exploration.
- You can explain to stakeholders that the Lookalike is a signal in suggestion mode, not a hard audience wall.
- You can separate materially different creative or audience strategies into ad groups.
Prefer a constraint or rethink the setup when
- Your policy, contractual, or operational requirement is to reach only a defined similarity band.
- Your conversion event is too sparse, delayed, or noisy to guide automated exploration responsibly.
- Your seed combines incompatible goals, such as free-trial users and high-value renewals, and cannot be cleaned before use.
A practical setup and testing workflow
1. Define the outcome before selecting an audience
Write down the conversion that determines success, the acceptable CPA or value target, and any users who must be excluded. This prevents a large but weak seed from becoming the default simply because it has more records.
2. Prepare the seed or signal
For a Lookalike, select first-party data that closely represents the desired outcome: recent purchasers for purchase acquisition, qualified leads for lead quality, or meaningful product users for activation. For Optimized targeting, select the audience segments that provide useful context, then make sure the campaign’s conversion action is the one you actually want the model to pursue.
3. Confirm Demand Gen eligibility and rollout state
In the campaign’s audience controls, note whether the Lookalike is presented as a signal or a constraint and whether Narrow, Balanced, or Broad is described as guidance or enforcement. Save a screenshot or campaign note if several teams will interpret the results later. The public rollout documentation warns that availability changes by account and phase. Use Lookalike segments to grow your audience
4. Decide whether to combine or separate
Combine the features when one message and one conversion goal apply to both. Create separate ad groups when the seed-led and model-led strategies need different creative, exclusions, landing pages, or reporting views. Google’s audience overview supports this separation for distinct messages or targeting strategies. Demand Gen audiences overview
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errors5. Launch with an interpretable change set
Avoid changing the seed, conversion action, budget, creative, and targeting mode simultaneously if you intend to learn which audience setup affected results. Keep location, language, attribution settings, and landing-page experience stable where possible, and document unavoidable changes.
6. Read delivery and conversion data with the rollout caveat
Review spend, impressions, conversions, conversion value, CPA, and conversion lag alongside the audience reporting label. A Lookalike signal receiving impressions outside the old band is not automatically a configuration error in suggestion mode. Conversely, a good CPA does not prove that the seed caused the improvement; it only shows that the campaign met its measured performance under that setup.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.Common mistakes and how to correct them
Calling a Lookalike a guaranteed audience boundary
Why it fails: Suggestion mode can expand beyond historical thresholds.
Correction: State whether the account is in suggestion or constraint behavior, and describe Narrow, Balanced, or Broad as guidance when that is how the account reports it.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Using a broad, unqualified seed
Why it fails: Similarity modeling learns from the people included. A list dominated by low-intent visitors can steer discovery away from purchasers or qualified leads.
Correction: Build the seed around the conversion and recency that matter, and remove records that represent a different objective.
Assuming Optimized targeting is restricted to selected audiences
Why it fails: Its purpose is to look beyond audience signals for likely converters.
Correction: Use exclusions and campaign settings for true restrictions, and explain to stakeholders that the selected segments are inputs rather than a complete delivery list.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchPC Slower Than It Used to Be?
A free scan shows the junk files, broken settings and background clutter dragging Windows down - then fixes them in one click.Free scan · Windows 10 & 11Comparing labels instead of outcomes
Why it fails: “Lookalike” and “Optimized” describe model inputs and behavior, not guaranteed incremental performance.
Correction: Compare the setups against the same conversion definition, attribution window, budget context, and observation period. The official documentation reviewed for this comparison does not provide a quantified head-to-head lift statistic.
Confusing Demand Gen with other expansion controls
Why it fails: Google’s audience expansion documentation covers specified Video campaign goals and should not be used as a synonym for Demand Gen Optimized targeting.
Correction: Use the control and documentation that match the actual campaign type. About audience expansion
What to monitor after launch
- Seed relevance: Are the people in the seed still representative of the customers or leads you want?
- Conversion quality: Are reported conversions the business outcomes that justify expansion, or merely inexpensive proxy events?
- CPA or value: Is performance acceptable under the campaign’s stated goal, accounting for conversion delay?
- Reach interpretation: Is the Lookalike a signal or a constraint in this account?
- Audience and creative fit: Do users reached through the setup receive a message and landing page appropriate to their likely intent?
- Reporting clarity: Can another person tell which ad group, seed, signal, and rollout state produced the result?
If performance changes after Google moves the account from constraint to suggestion behavior, treat the rollout state as a material change in the campaign, not as a minor label update.
Frequently asked questions
Frequently Asked Questions
Do the 2.5%, 5%, and 10% figures describe the percentage of my budget or impressions?
No. In Google’s legacy Lookalike guidance, Narrow, Balanced, and Broad referred to an intended share of users in the target location—approximately 2.5%, 5%, and 10%. They were not budget percentages, and in suggestion mode they are not strict delivery limits.
Will Google tell me exactly which users were reached because they matched my seed?
The audience label is a modeling explanation, not a user-level eligibility receipt. In suggestion mode Google identifies the Lookalike as a signal, so reporting should not be read as proof that every impression came from a fixed similarity band.
Can I keep a Lookalike seed but prevent all expansion?
Only if the account offers and uses the documented constraint behavior. Google is phasing accounts toward suggestion mode and says opt-out controls arrive later in 2026, so availability must be checked in the specific account.
Recommended Free Tools
Is a larger seed always better?
No. Relevance to the conversion objective matters more than raw record count. A smaller, recent list of purchasers or qualified leads can provide a more useful direction than a large list mixing low- and high-intent users.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
