TTD AE flashcards

33 cards. Read the term and the definition here. The deck at the top of this page is optional.

Advertiser account hierarchy

Top level holds advertiser settings, users, and billing; campaigns sit beneath it, and ad groups hold bids, targeting, and creatives. Data is scoped to the advertiser account.

Seat permissions

A seat grants user access to one or more advertiser accounts with role-based rights such as admin, campaign manager, or reporting-only. Least privilege is the default assignment.

Universal pixel

A single site-wide tag that fires on all pages and passes event data; it enables retargeting, conversion tracking, and audience building without per-page tags.

First-party data ownership

CRM, site, and offline conversion data uploaded by an advertiser stays within that advertiser account unless explicitly shared with a partner or other seat.

Naming convention impact

Consistent account, campaign, and ad group naming enables reliable reporting filters and automated QA. Inconsistent names break scheduled reports and make performance comparisons unreliable.

Campaign vs. ad group

Campaigns carry budget and flight dates; ad groups carry bid, targeting, creative, and inventory. Setting budget at both levels can cause double counting or unexpected pacing.

Bid type selection

CPM bids for reach, CPC for clicks, CPA or CPCV for conversions. The bid type defines what the bid value represents and how the optimizer paces spend.

Dayparting time zone

Dayparting schedules delivery windows using the ad group time zone. Confirm the time zone before scheduling or delivery runs in the wrong hours.

Frequency cap effect

Caps limit impressions per user per time window. Tight caps reduce available impressions and can cause underdelivery, especially on narrow audiences.

Daily vs. total budget

Daily budgets reset each day and can flex within platform rules; total budgets spread across the flight and stop when exhausted. Choose based on flight certainty.

Conversion tracking prerequisite

CPA and optimization bidding need a conversion pixel attached at the ad group level. Without a conversion signal the optimizer cannot adjust bids toward the goal.

CRM onboarding

Uploading hashed first-party identifiers such as email or phone to build audience segments. Match rate depends on hashing, identifier quality, and identity graph coverage.

Lookalike seed quality

Lookalike modeling expands a seed segment; seed size and quality determine reach and performance. The lookalike refreshes as the seed segment updates.

Exclusion precedence

Exclusion logic overrides inclusion targeting. Suppressing converters from prospecting or existing customers from acquisition prevents wasted impressions and distorted performance.

Identity graph match rate

Cross-device graphs link devices to a user or household using deterministic and probabilistic signals. Match rate determines how much cross-device reach a segment delivers.

Segment overlap reach

Layering multiple audience segments can shrink reach when intersection logic applies. Check estimated reach before launch to avoid underdelivery on narrow combinations.

Hashed identifier upload

Data onboarding requires hashed identifiers, typically SHA-256. Unhashed personally identifiable information should never be uploaded to the platform for audience building.

Private marketplace deal

An invitation-based deal with a deal ID granting access to publisher inventory, often at a floor price. The deal ID must be attached to the ad group to bid.

Programmatic guaranteed

A fixed-price deal with guaranteed volume, typically highest priority in the auction. It secures inventory but reduces flexibility compared with open exchange buying.

Supply path optimization

SPO shortens and deduplicates auction paths to the same impression, reducing take rates and improving win efficiency. Prefer direct or short paths to publishers.

Deal priority order

Auction priority generally runs guaranteed, then preferred, then private marketplace, then open exchange. Higher-priority deals can preempt lower-priority bids for the same impression.

First-price auction

Most modern exchanges run first-price auctions where the winning bid pays its bid amount. Bid shading or a bid strategy is needed to avoid overpaying.

Pre-bid brand safety

Pre-bid filters such as allow and block lists, viewability thresholds, and invalid traffic filtering shape available inventory. Set them before bidding to avoid wasted spend.

Pacing modes

Even pacing spreads budget across the flight, ASAP spends as fast as possible, and front-loaded spends early. Pacing controls how budget releases over time.

Effective CPM

eCPM equals total spend divided by impressions, multiplied by 1000. Comparing eCPM across ad groups reveals which inventory and targeting combinations are inefficient.

Target CPA bidding

The optimizer adjusts bids to hit a target cost per acquisition while spending budget. A target set too low causes underdelivery because few impressions qualify.

Budget pacing variance

Expected spend to date equals daily budget times elapsed days. Actual spend far below or above expected signals under-pacing or over-pacing that needs investigation.

Bid floor constraint

The minimum bid needed to win an impression. Bids below the floor never win, so a floor above the configured max bid produces zero delivery.

Conversion volume threshold

Optimization bidding needs sufficient conversion volume to learn. Low-volume ad groups should run CPM or CPC until data accumulates before switching to CPA.

Attribution models

Last touch, first touch, linear, time decay, and position-based models assign conversion credit differently. The model choice changes which touchpoints appear to drive performance.

View-through vs. click-through

View-through conversions credit impressions without a click; click-through conversions credit clicks. Report both separately to avoid double counting the same conversion.

Funnel metric formulas

CTR equals clicks divided by impressions; CVR equals conversions divided by clicks; CPA equals spend divided by conversions; ROAS equals revenue divided by spend.

Incrementality holdout

A control group withheld from advertising measures lift beyond baseline. Platform-reported conversions are correlational without a holdout, so incrementality tests establish causal impact.