Skip to content
Faisal Mahmud TusherDigital marketing, SEO & analytics
with Faisal Tusher

Programmatic Advertising › Display & Video 360

Audiences in DV360

How audiences work in DV360: first-party lists and the Floodlight link, Google audiences, third-party segments and fees, combined audiences and consent.
Guideintermediate9 min readUpdated

Audience sources

SourceExamplesWhere it comes fromCostTypical use
First-partyFloodlight activity lists, Customer Match uploads, Google Analytics audiences, YouTube channel lists, exposure-based listsThe advertiser's own tags, uploads and linked accountsMedia onlyRemarketing, suppression, loyalty, seed lists
Google audiencesAffinity, in-market, detailed demographics, life events, custom audiences from keywords, URLs and appsGoogle's own signals across its propertiesMedia onlyProspecting at scale, reach planning
Third-partySegments from data providers listed in DV360's audience marketplaceExternal data companiesMedia plus a per-segment data feeNiche prospecting, business or purchase-based targeting
CombinedAny of the above joined with and, or and not logicBuilt in DV360 from existing listsWhatever the components costPrecise inclusion and clean exclusion

Each row behaves differently in reach, cost and durability. The sections below take them in turn. For the display targeting options inside Google Ads, which overlap heavily with the Google audiences row, the display targeting options guide covers the same families from that platform's side.

First-party audiences and Floodlight

The most valuable lists are the advertiser's own, and in DV360 they arrive through the Floodlight configuration of the linked Campaign Manager 360 advertiser. A Floodlight activity tag fires when someone views a product page, adds to basket or completes a purchase; DV360 can build an audience from any of those activities, with rules on the activity's custom variables and a membership duration up to the documented maximum. The tag setup itself, naming, custom variables and the testing routine belong to the Floodlight tags guide; this page assumes the tags fire and covers what DV360 does with them.

Beyond Floodlight, DV360 accepts Customer Match lists uploaded as hashed customer identifiers, subject to the eligibility and policy requirements the Help Centre sets out; audiences shared from linked Google Analytics properties; lists of people who have interacted with a linked YouTube channel; and audiences built from exposure to the advertiser's own DV360 campaigns, such as everyone who saw a particular line item. All of these are created at the advertiser level and are available to every line item under it, and the Help Centre describes the conditions under which a list can be shared with other advertisers under the same partner.

This is also the layer where DV360 does the job a data management platform (DMP) used to do: collecting, segmenting and activating the advertiser's own audience data in one place. The practical work is designing the lists before the campaign needs them, because a list only begins collecting members when it is created, and a Remarketing line item launched the same day as its list has nobody to reach.

Google audiences

Google audiences are built from Google's own signals and are available in DV360 without a data fee. Affinity audiences describe long-run interests; in-market audiences describe people Google judges to be actively researching a category; detailed demographics cover attributes such as parental status and education where available; life events cover moments such as moving house. Custom audiences let an advertiser describe an audience by the keywords people search, the URLs they visit or the apps they use, and Google builds the list from those descriptions. Availability by country and by line item type varies, and the Help Centre lists which families work for display, video and YouTube inventory.

Similar audiences, Google's lookalike product (Lookalike Audience), were deprecated for display and video line items; Google directs advertisers to optimised targeting instead, a line item setting that lets DV360 serve beyond the selected audiences when its conversion signals suggest a good match. The feature has been renamed since it launched, so the current name and conditions should be checked on the Help Centre's audience expansion page rather than assumed from an older screenshot.

In planning terms, Google audiences cover the See and Think stages of the see-think-do-care framing, where the advertiser has no first-party relationship yet, and first-party lists cover Do and Care. Most plans use both, with Google audiences for prospecting line items and first-party lists for remarketing and suppression.

Third-party segments and data fees

DV360 lists segments from third-party data providers in its audience marketplace. Each segment shows a description, its provider and a price expressed as a cost per thousand impressions (CPM) that is charged on top of media whenever the segment is used on a served impression. The fee is set by the provider, and DV360 reports it separately from media cost, so the data fee metric in reports is the place to see what a segment actually added to the bill. The Help Centre describes how fees are applied when several priced segments are combined on one line item.

Two cautions. The first is coverage: third-party segments are built on identifiers that the provider could collect, and the same consent and cookie limits that shrink first-party lists shrink these, often more, because the provider's data was collected on other sites. The second is accountability: Google does not vouch for the quality of a provider's segment definitions, and a segment called "luxury intenders" means whatever the provider decided it means. Test a priced segment in its own line item against a Google audience of similar intent before scaling it, and read the data fee alongside the media cost when judging it. Where a line's price per thousand has come in above the plan's, the diagnosis of a programmatic CPM higher than planned shows how to tell a priced segment's data fee from the other causes.

Combined audiences and exclusions

A combined audience is a saved expression built from existing lists: this first-party list or that one, and this in-market audience, and not existing customers. It is created once at the advertiser level and assigned to line items like any other list, which keeps complex targeting readable and reusable. Combined audiences are the right place for inclusion logic that would otherwise be built differently on every line item.

Exclusions are the other half. DV360 lets a line item exclude audiences as well as include them, and the most common exclusion is the advertiser's own converters, so that prospecting budget is not spent on people who have already bought. Suppression lists built from Customer Match uploads serve the same purpose for offline customers. Set exclusions deliberately on every prospecting line item rather than relying on frequency caps to limit waste, and note that an exclusion only works where the person could be matched, which brings the discussion back to consent.

The equivalent mechanics in Meta are described in the custom audiences guide; the concepts map closely, but the matching rates and the identifier each platform relies on differ, so the same customer file produces lists of different sizes in each.

Audience insights and sizing

Each list in DV360 shows a size, broken down by the environments where it can be matched. The Help Centre documents how sizes are estimated and the delay before a new list shows a figure, so a list that reads as empty on the day it is created is not necessarily broken. Sizes are modelled estimates of matchable identifiers rather than counts of people, and they move as membership durations expire members and as consent and browser changes alter what can be matched.

DV360 also offers audience composition reporting, which shows which Google audiences over-index among the members of a first-party list. It is useful for choosing prospecting audiences that resemble converters and for sanity-checking a third-party segment before paying for it. Treat it as a planning input, not as a targeting change in itself.

Audience lists are only as large as the identifiers DV360 can see, and that set has been shrinking. Browsers that block third-party cookies remove their users from most display lists; app traffic without an advertising identifier cannot be matched; and users in regions covered by consent requirements only enter a list if consent for advertising use was recorded. Google's EU user consent policy and consent mode govern how DV360 and Floodlight treat these users, and the privacy and consent guide explains how consent signals reach Google tags in the first place.

Google documents routes that reduce the dependence on cookies, including publisher-advertiser identity reconciliation for matching first-party data with participating publishers, and the modelled reach and frequency methods that reports use where identifiers are missing. None of them restore the list sizes of the cookie era. The working rule is to read every audience size as a lower bound on the people the advertiser cares about and an upper bound on the people DV360 can reach, and to plan reach with that gap in mind.

Illustrative example — not a real result

Sizing a remarketing plan for a subscription meal kit

A meal-kit brand links its Campaign Manager 360 advertiser to DV360 and builds four Floodlight lists: visitors to the menu pages with a 30-day membership, people who started the sign-up with a 14-day membership, new subscribers with a 180-day membership used only as an exclusion, and lapsed subscribers from a Customer Match upload of 42,000 hashed emails. After a week the menu-page list reads 61,000 matchable identifiers against roughly 240,000 consented sessions, and the Customer Match list matches 23,500 of the 42,000 records. The trader runs the sign-up abandoners on a dedicated line item with a weekly cap of 6 impressions, excludes the subscriber list from every prospecting line item, and tests a third-party "meal delivery intenders" segment at a £1.20 CPM data fee in its own line item for two weeks before deciding whether it beats Google's in-market food delivery audience on cost per sign-up.

When the plan is judged on online orders, the recommended DV360 setup for online sales puts product-viewer and basket lists on their own remarketing line item and excludes recent purchasers from prospecting.

Common mistakes

  • Creating lists on launch day. A Floodlight list starts empty; build it as soon as the tag is live, weeks before the campaign that needs it.
  • Reading list size as a count of people. It is a modelled estimate of matchable identifiers and changes with consent, browsers and membership duration.
  • Forgetting the data fee. A third-party segment adds its own cost per thousand to every impression; judge it on media plus data.
  • Prospecting without exclusions. Existing customers see acquisition creative and the budget pays for it.
  • Assuming similar audiences still run. They were deprecated for display and video; optimised targeting is the documented replacement, under its current name on the Help Centre.
  • Combining logic on each line item by hand. Build a combined audience once and assign it, so the logic is auditable.
  • Treating consented traffic as all traffic. Reach plans built on total sessions overstate what any list can deliver.

The DV360 hub places audiences among the other line item decisions; inventory, bidding and creatives each have their own guide there.

Frequently asked questions

Why is my first-party audience in DV360 much smaller than the site traffic suggests?

A list only counts people whose Floodlight hit could be matched to an identifier DV360 can bid on, and that excludes browsers that block third-party cookies, users who declined consent, and app traffic without an advertising ID. Membership duration also trims the list continuously. Compare the list size against consented, cookie-capable traffic rather than total sessions.

Do similar audiences still exist in DV360?

Google deprecated similar audiences for display and video line items and points advertisers to optimised targeting on the line item instead, which expands beyond the selected audiences using the line item's conversion signals. The current status is on the Help Centre page for audience expansion, which is the page to check because naming has changed more than once.

Are third-party data fees included in the media cost shown in reports?

No. Each third-party segment carries its own cost per thousand impressions, set by the data provider, and DV360 reports it as a separate data fee metric alongside media cost. A line item's true delivered cost is media plus data plus the platform fee, which is why automated strategies with a cost constraint behave differently when a priced segment is added.

Terms used on this page

Frameworks

Where this page is used

Part of these learning paths

Also related

Sources

  1. Display & Video 360 Help — Google

Related