Google Ads Audience Targeting for DTC Ecommerce: Demographics, Household Income, and Custom Segments

This guide covers how each audience type works, when to use custom segments versus in-market audiences, how Customer Match fits the DTC stack, and the specific settings that separate accounts that use these as real performance levers from accounts where they just collect data.

What Google Ads Audience Targeting Actually Does

Most DTC brands treat audience targeting in Google as a Display feature. It is not. Demographics, in-market audiences, custom segments, and Customer Match all work across Search, Shopping, and Performance Max. The mechanism is just different.

The distinction that matters most: "Observation" versus "Targeting" mode.

In Observation mode, Google serves your ads to everyone who matches your campaign's keywords and targeting settings. The audience layer sits on top as a data-collection and bid-adjustment tool. You can see how different segments perform and apply bid modifiers (up or down) without cutting off any traffic.

In Targeting mode, Google only serves your ads to people who are both in that audience and match your keyword or product targeting. Reach narrows significantly. For most Search and Shopping campaigns, this is too restrictive.

The default approach: start every new audience layer in Observation mode. Run it for 30-45 days to collect meaningful data. Apply bid adjustments based on what you see. Only use Targeting mode when you have a specific reason to limit reach, like a retargeting-only campaign or a brand campaign where you want to show only to known customers.

Demographic Targeting: Age, Gender, Household Income, Parental Status

Google exposes four demographic dimensions in campaign settings:

Age: 18-24, 25-34, 35-44, 45-54, 55-64, 65+, Unknown. The Unknown bucket is large on most accounts; mobile and signed-out users often do not have confirmed age data. Do not apply heavy bid decreases to Unknown without first checking whether your conversions index there.

Gender: Male, Female, Unknown. Same caveat applies on Unknown.

Parental status: Parent, Not a Parent, Unknown. Useful for baby and family products; less signal-rich for most DTC categories.

Household income (HHI): Top 10%, 11-20%, 21-30%, 31-40%, 41-50%, Lower 50%, Unknown. The most underused demographic layer for DTC brands.

How Household Income Targeting Works

HHI targeting pulls from IRS tax data and third-party providers correlated against Google account signals. It is available in the US, Canada, Australia, Japan, and a handful of other markets. In the UK and most of Europe, it is not available.

The tiers go from Top 10% (highest earners) down to Lower 50%. For a premium supplement brand or a $200+ product, the top two tiers are often where your best customers live. The question is whether the data is dense enough to bid meaningfully.

One thing to account for: a significant share of searches will fall into the "Unknown" bucket because the user is not signed into a Google account or does not have enough data history for Google to estimate income. Before applying large bid decreases to Unknown, check whether your conversion rate there is materially worse or if you are just looking at the noisier tail.

The move we see work for mid-to-premium DTC brands: add HHI in Observation mode across active Search campaigns, let it collect 4-6 weeks of conversion data, then apply a bid increase of 10-20% on top two tiers if they convert at better efficiency, and a modest decrease (-10 to -20%) on Lower 50% if it consistently underperforms. Avoid aggressive exclusions until you have statistically meaningful sample sizes.

The Four Audience Types in Google Ads

Affinity audiences are brand-awareness tools. They reach people who have an interest, not people who are shopping now. For DTC performance campaigns, their role is usually narrow. Broad awareness video campaigns are the right fit; Search and Shopping performance campaigns generally are not.

In-market and custom segments are where Search and Shopping campaigns get interesting.

Custom Segments vs In-Market Audiences for DTC Brands

In-market audiences are Google's pre-built groupings of people who its algorithm has identified as actively shopping a category. The categories are broad by design. "In-Market for Health & Wellness Products" includes a huge swath of behavior.

Custom segments let you build your own. There are three ways:

1. People who searched for these terms on Google: You enter specific search queries. Google targets people who searched those terms recently, across Search, YouTube, and Display. This is the most powerful option for DTC: you can build an audience of people who searched your competitors by name.

2. People who browse websites similar to: You enter URLs. Google targets people whose browsing history is similar to visitors of those sites. Useful for competitor brand targeting when you do not have that competitor's search intent data.

3. People who use apps similar to: Less commonly used for DTC, more relevant for mobile-first categories.

In practice, custom segments built from competitor brand search terms tend to outperform broad in-market audiences for mid-funnel buyers. Someone searching "Athletic Greens alternatives" or "AG1 vs competitors" is further along in the buying process than someone Google has categorized as in-market for supplements.

The move: build a custom segment with your top 5-10 competitor brand names as the seed terms. Layer it on your brand-agnostic Search campaigns in Observation mode. If it outperforms the baseline after a month, apply a bid increase and consider whether it deserves its own dedicated ad group or campaign.

Customer Match: Targeting Your Own Buyers and Building Lookalike Seeds

Customer Match uploads your customer email list to Google, which matches it against signed-in Google account holders. The matched list becomes an audience you can target, exclude, or use as a seed for similar-audience expansion.

Requirements: your Google Ads account needs a spend history (Google has made this threshold variable over time), and you need at least around 1,000 matched users for the list to be eligible for targeting.

For DTC brands, Customer Match earns its place in three scenarios:

Exclusions on prospecting campaigns. Upload your existing buyer list and exclude them from top-of-funnel Search and Shopping. This stops you from paying for clicks from people who already converted.

Retention and upsell campaigns. Run a separate Search or Shopping campaign targeting only Customer Match buyers. These people know your brand; you can bid more aggressively and tailor copy to retention messaging.

Lookalike signals for PMax. Even though Google deprecated traditional Similar Audiences in 2023, Customer Match lists fed into PMax as audience signals help the algorithm find users who look like your existing buyers. This is now the primary use case for Customer Match as a prospecting tool.

One thing to watch: Customer Match match rates vary significantly by email quality and list age. We have seen match rates run anywhere from 25% to 65% depending on how clean the email data is and whether it matches how users sign into Google. Upload multiple list variants (purchase email, subscription email, checkout email) if your data is fragmented across sources.

How Audience Targeting Works in Performance Max

PMax is different. You cannot apply Observation mode in the traditional sense. Instead, you provide "audience signals" (lists, demographics, interests, custom segments) and the algorithm uses these as starting hints. It will expand beyond them if it finds conversions outside your signals.

What this means in practice: your audience signals influence the early learning phase. A strong signal list (Customer Match buyers, custom segment from high-intent search terms) helps PMax find the right users faster. A vague signal list (broad affinity category) does not give the algorithm much to work with at launch.

The inputs that tend to matter most for PMax audience signals:

1. Customer Match of existing buyers: the strongest signal you can give PMax

2. Website visitors (remarketing): people who hit your site but did not convert

3. Custom segments built from your highest-converting search terms: essentially telling PMax to find more people who search like your best customers

4. In-market audiences relevant to your product: useful as a secondary layer when you do not have enough first-party data

What PMax does not let you do: exclude specific demographics from the asset group. You can apply account-level negative audiences (like excluding existing buyers from all campaigns), but you cannot tell PMax to stop serving to the Lower 50% HHI tier.

For more on PMax campaign structure and budget allocation, see Google Ads Performance Max Strategy for Ecommerce Brands in 2026.

Observation vs Targeting Mode: Which to Use and When

The common mistake: switching to Targeting mode because the audience looks like it performs well in reports. What Observation mode shows you is how people in that audience perform when they happen to see your ad. Switching to Targeting restricts your reach to only that audience, which may cut off converting traffic that happens to fall outside it. Before restricting, confirm that the excluded population genuinely underperforms, not just that the included segment over-indexes.

Frequently Asked Questions About Google Ads Audience Targeting