Skip to main content

Example segment taxonomies

Three ready-to-copy segment sets — performance e-commerce, lead-gen SaaS, and content/awareness.

Segments work best when the taxonomy is small (5–8 segments) and consistent across organizations in the same vertical. Three patterns we see most often:

Performance e-commerce

For shops where the goal is purchases and ROAS at scale.

SegmentCodeWhat goes inColour
AcquisitionACQNew-customer prospecting (Search non-brand, social cold audiences, lookalikes)Red
RemarketingREMVisitors / cart abandoners / past purchasers warming upBlue
RetentionRETExisting-customer LTV plays — VIP, win-back, cross-sellPurple
BrandBRANDBranded search, brand-awareness videoAmber
AwarenessAWATop-of-funnel reach campaignsPink

A Segments report on this taxonomy answers the classic CMO question: "Where did we spend, and what did each part of the funnel return?"

Lead-gen SaaS

For B2B / lead-gen funnels where Purchases is rare and Lead is the primary metric.

SegmentCodeWhat goes inColour
DemandDEMSearch intent / problem-aware audiencesRed
Account-basedABMTargeted account lists and custom audiencesCyan
ContentCONTTop-of-funnel ebook / webinar / whitepaper download campaignsTeal
BrandBRANDBranded search, retargeting on brand-aware visitorsAmber
Free trialTRIALLower-funnel campaigns optimising for trial signupGreen

Pair this with a Lead system metric mapped to your demo-request and trial-signup conversion definitions (conversion mapping) for a clean funnel view.

Content / awareness

For media properties or product-led brands where reach and engagement matter more than direct conversion.

SegmentCodeWhat goes inColour
ReachREACHPure CPM / impressions playsPink
EngagementENGVideo views, post engagement, social participationPurple
SubscribeSUBNewsletter / app-install / channel subscribe campaignsBlue
BrandBRANDBranded search, branded displayAmber

For this taxonomy, metrics like engagement rate and cost per subscriber matter more than ROAS — define them once on the Metrics page, and check the built-in formula presets before building your own.

Picking codes

The code does more than label the segment: during each sync, campaigns whose names contain a segment's code get that segment assigned automatically. So the code is also your naming convention's hook into Level. Rules of thumb that hold across all three patterns:

  • 3–5 letters, uppercase — readable in narrow columns and chips.
  • Distinctive — a code that appears as a fragment of ordinary words will auto-match campaign names you never intended. TRIAL is safe; a two-letter code rarely is.
  • Put the code in campaign namesACQ | Search - Non-brand, REM | DPA - Cart abandoners. New campaigns then tag themselves on the next sync, on every platform.
  • No platform suffixAcquisition, not Acquisition - Google. Segments are platform-agnostic; that's the point.
  • No date suffixBrand, not Brand 2026. Segments stick around for years; date filters belong on the report.

Migrating an existing taxonomy

If you already have a taxonomy defined in spreadsheets or a tagging tool:

  1. Pick Level codes that match the tokens already present in your campaign names as closely as possible.
  2. Create the segments on the organization.
  3. Let the next sync auto-assign every campaign whose name contains a code.
  4. For campaigns whose names don't carry a token, use bulk-assign on the Campaigns page to tag them in batches.
  5. Spot-check a few reports to confirm the rollup looks right — anything still untagged shows up under Unknown.