Wallet Labeling and Crypto User Segmentation: A Practical Guide
Key takeaways
Wallet labels describe observable facts about an address; segments combine labels with user, wallet, campaign, and time-based properties.
Current DeFi behaviour labels cover lending, liquidity, yield, staking, derivatives, real-world assets, prediction markets, bridges, and aggregators.
Generic whale tiers are not a universal segmentation model. Start with the relevant behaviour and filter supplied, borrowed, deposited, staked, or position value.
Formo lifecycle stages are New, Returning, Power user, At Risk, Churned, and Resurrected. Each profile has one current stage.
Wallet clustering is probabilistic unless users explicitly link addresses. Preserve the underlying wallets and the confidence of every inferred link.
A wallet address is a pseudonymous account, not a complete user profile. Public chain data can show transactions, balances, contract interactions, and current positions. Product data can show how the same wallet reached and used an app. Wallet segmentation joins those sources so crypto teams can compare useful cohorts without treating every address as a known person.
This guide explains how wallet labeling and crypto user segmentation work, which DeFi labels and lifecycle stages Formo uses, which profile properties belong in a segment, and where wallet clustering and scoring require care.
What are wallet labels and crypto user segments?
A wallet label is a descriptive tag attached to a wallet profile. It can represent a lifecycle stage, a current DeFi position, an attestation, a campaign, an onchain registry, a Sybil list, or project-specific information added as a custom label.
A user segment is a saved cohort defined from one or more conditions. A segment can combine wallet labels with user activity, profile properties, campaign source, transaction count, volume, revenue, points, and a time window. For example: DeFi Lenders with at least $10,000 supplied who first arrived through a partner campaign and transacted within the last 30 days.
The distinction matters. Labels describe a profile. Segments answer a specific product, growth, or research question by filtering profiles.
Current wallet label categories in Formo
Formo groups wallet labels into Lifecycle, Attestations, DeFi behaviour, Sybil lists, Campaigns, Onchain registries, and Custom labels. The current rules and refresh behaviour are documented in the Wallet Labels reference.
DeFi behaviour labels
DeFi Lender — currently supplies at least $100; the label value is Supplied USD.
DeFi Borrower — currently owes at least $100; the label value is Borrowed USD.
Liquidity Provider — currently holds at least $100 of DEX liquidity; the value is Position USD.
Yield Farmer — currently holds at least $100 in vault or yield positions; the value is Deposit USD.
Staker — currently holds at least $100 in staking positions or tokens; the value is Staked USD.
Perp Trader, Options Trader, and Basis Trader — currently hold qualifying positions; the value is Position USD or Deposit USD, depending on the label.
RWA Holder and Prediction Market User — hold a qualifying current position or account; the value records the current position where available.
Bridge User and Aggregator User — have used a recognized bridge or aggregator; the label value records protocol names and has no USD floor.
Position-based labels reflect the latest wallet-profile snapshot. Closed positions are removed from that snapshot. USD-based categories use a $100 minimum, while interaction labels such as Bridge User and Aggregator User do not. Labels refresh daily and are removed when the current rule no longer applies.
Custom labels
Custom labels add information that chain data cannot determine. Teams can use them for project-specific classifications such as account tier, allowlist status, partner cohort, or internal research group. Keep custom-label definitions documented so a label has the same meaning across dashboards and campaigns.
How to use value in wallet segmentation
Fixed Retail, Professional, Dolphin, and Whale buckets are not a universal standard. A wallet’s estimated net worth may be useful context, but it does not show which activity creates value for a specific app.
Start with behaviour, then filter the amount attached to that behaviour. Lending teams can filter Supplied USD or Borrowed USD. Liquidity, staking, and yield products can filter Position USD, Staked USD, or Deposit USD. Trading and market apps can filter open-position value, deposit value, attributed volume, fees, or revenue. Net worth can remain a separate profile filter when the question genuinely requires it.
User lifecycle stages
Formo assigns each profile one current lifecycle stage from project activity. Projects can override the defaults. The current definitions are documented in Wallet Profiles: User lifecycle.
New — first seen within the last 30 days and still active.
Returning — an established active user who does not match another lifecycle stage.
Power user — first seen more than 30 days ago, still active, and active on at least five distinct days in the last 30 days.
At Risk — previously active, last seen at least 14 days ago, active on fewer than five days in the last 30, and active during the prior 30-to-60-day period without a 30-day inactivity gap.
Churned — last seen at least 30 days ago.
Resurrected — first seen more than 30 days ago and active again within the last 30 after an inactivity gap of at least 30 days.
The stages are mutually exclusive. Formo evaluates them in this order: Churned, New, Power user, Resurrected, At Risk, then Returning. That precedence prevents one profile from appearing in several lifecycle cohorts at once.
Profile properties to use in segments
User and acquisition properties
Custom user properties such as user ID, display name, account tier, or other fields sent by the app.
First seen, last seen, sessions, country, device, browser, and operating system.
First and last referrer, referral code, and UTM source, medium, campaign, term, and content.
Product and business outcomes such as transaction results, volume, revenue, and points.
Wallet and portfolio properties
Wallet address, ENS or other available names, wallet age, and last onchain activity.
Total transaction count and transaction frequency.
Estimated net worth and portfolio allocation by chain.
Token balances and DeFi positions, including app, protocol, chain, position USD, and portfolio share.
Linked wallet addresses and the method used to establish each link.
These are profile properties, not wallet labels. Keeping the distinction clear makes segments easier to explain and prevents campaign or transaction fields from being presented as classifications.
Methods of wallet clustering
Wallet clustering attempts to identify addresses that may belong to the same user or coordinated group. The method determines how strong the conclusion can be.
User-linked wallets — a user explicitly links several addresses to one profile. This is the strongest relationship available in the product context.
Signed control — a wallet signature verifies control of one address at that moment.
Funding relationships — shared funding sources can indicate a relationship but also occur for unrelated users of the same exchange or bridge.
Timing and transaction patterns — synchronized actions, repeated amounts, and similar routes can indicate coordination but are not proof of common ownership.
Positions, balances, network activity, wallet age, names, and attestations — useful features for a model when their coverage and observation window are explicit.
Store the underlying addresses, the evidence used, and a confidence level for inferred clusters. Do not present probabilistic clusters as verified people.
Wallet scores
A wallet score is an application-specific ranking, not a universal credit score. Define the outcome first—for example, first transaction, repeat usage, retained volume, or revenue—then use only properties available before that outcome. Document the inputs and weights, validate the score on later cohorts, and monitor drift.
Wallet segmentation use cases
Product analysis
Compare onboarding completion, transaction failures, feature adoption, and retention across lifecycle, behaviour, and acquisition cohorts. Use the result to test a specific product hypothesis rather than assuming every profile with the same label needs the same experience.
Growth and attribution
Compare acquisition source, activation, volume, revenue, and retention for the same wallet segments. Use those results to move budget toward sources that bring activated and retained users, then measure each campaign against the outcome it was designed to produce.
Incentive design and abuse review
Define eligibility from relevant behaviour and time windows, check linked-wallet and coordination signals, and measure activity after distribution. Shared funding or synchronized activity should be treated as evidence to combine with other signals, not a single-rule verdict.
How to implement wallet segmentation
Capture in-app actions, product events, acquisition source, and confirmed onchain activity with a consistent wallet identifier.
Add the wallet labels and profile properties needed for the question. Avoid collecting fields that do not change an analysis or decision.
Define every segment with explicit conditions, time windows, networks, and value thresholds.
Validate inferred identity links and scores on later data, and preserve confidence levels.
Compare volume, revenue, retention, completion, and failure rates by segment; revise rules when they stop separating meaningful cohorts.
Wallet intelligence and segmentation in Formo
Formo combines product analytics, wallet intelligence, and onchain attribution. Teams can inspect wallet profiles, use current wallet and lifecycle labels, add custom labels and user properties, and build cohorts from product, campaign, wallet, and transaction data in one platform.
Explore Formo Analytics or review the wallet intelligence documentation.
Frequently asked questions
What is wallet labeling?
Wallet labeling attaches descriptive categories to a wallet profile based on lifecycle, current DeFi behaviour, attestations, campaigns, registries, Sybil lists, or custom project information.
What is the difference between a wallet label and a segment?
A label describes one profile. A segment is a saved group of profiles that match one or more label, property, activity, campaign, value, or time-based conditions.
Do net-worth buckets make sense for wallet segmentation?
Not as a universal primary model. Behaviour-specific amounts usually answer a clearer question: supplied USD for lenders, borrowed USD for borrowers, position or deposit USD for liquidity and trading profiles, and volume or revenue for app outcomes. Net worth can be an additional profile filter.
What lifecycle stages does Formo use?
The default stages are New, Returning, Power user, At Risk, Churned, and Resurrected. Each profile has one current stage, and projects can override the defaults.
Which profile properties can be used for segmentation?
Examples include campaign source, referrer, referral code, UTMs, sessions, first and last seen, custom user properties, transaction count and frequency, volume, revenue, points, wallet age, net worth, chains, token balances, DeFi positions, and linked addresses.
Can wallet clustering prove that several addresses belong to one person?
Only an explicit user link or equivalent verification establishes a strong relationship. Funding, timing, transaction patterns, balances, and other public signals produce an inference that should retain its evidence and confidence level.
