How to Build and Analyze Consumer Asset Models for Precise Marketing

Analysis Background

As the internet matures, traffic dividends shrink and competition intensifies. Merchants therefore shift from broad traffic acquisition to fine‑grained consumer operations. The foundation of fine‑grained operations is consumer analysis: segmenting users, creating consumer assets, measuring asset size, tracking asset flow, and evaluating marketing contributions to asset growth.

Basic Concepts

Consumer Asset

Consumer asset (customer equity) is the aggregate lifetime value of all customers. It is built through relationship management and can be visualized with models such as AIPL (Awareness‑Interest‑Purchase‑Loyalty) or BrandDynamics.

User Segmentation

User segmentation groups users based on characteristics and behaviors, enabling targeted product and operation strategies.

Typical Consumer Asset Models

Different business goals lead to various segmentation schemes, e.g., potential‑new‑old user models for stores or AIPL/DEEPLINK models for brands.

AIPL Example

Awareness (A) : Users reached by ads or category searches.

Interest (I) : Users who click ads, browse pages, interact, add to cart, subscribe, etc.

Purchase (P) : Users who have bought the brand’s products.

Loyalty (L) : Users who repurchase, comment, or share.

Solution Overview

The solution provides custom user segmentation, asset state‑change analysis, and marketing contribution evaluation. It includes multiple segmentation models, state‑transition models, and attribution models.

1. Behavioral Segmentation Model

Users are classified based on observed behaviors and their contribution strength.

Segmentation Steps

Define business objectives.

Determine segmentation scope (e.g., specific store or brand).

Define segmentation behaviors: baseline actions, behavior cycles, and depth.

Rank behaviors by conversion strength and map them to asset layers (e.g., exposure → potential, purchase → loyal).

2. Value Segmentation Model (RFM)

RFM scores users on Recency, Frequency, and Monetary value to create value tiers.

RFM Components

Recency (R) : Days since last purchase; smaller R indicates higher activity.

Frequency (F) : Number of purchases within a period.

Monetary (M) : Total spend in the period (or average order value).

Scoring Process

Segment R, F, M using equal‑width, equal‑frequency, or custom thresholds and assign scores (R_score, F_score, M_score).

Combine scores by weighted sum: RFM_score = R_score * w_R + F_score * w_F + M_score * w_M.

RFM Case Study

A skincare brand applied RFM to a year’s worth of member data, ranking users into high (20%), medium (30%), and low (50%) value groups. High‑value users contributed 67% of total revenue despite representing only 20% of the base.

3. User Flow Model

Analyzes how users transition between segmentation states over time (e.g., A → P) to measure flow volume, conversion rates, and deepening ratios.

Key Metrics

Initial/Final counts per layer.

Flow rate: proportion moving from one layer to another.

Overall deepening rate: proportion of users moving to deeper layers.

Layer‑specific deepening rate.

Flow Model Case Study

Comparing July 1 and July 7 AIPL states for a brand showed an overall deepening rate of 24.47% with detailed flow numbers per layer.

4. Flow Attribution Model

Attributes user state transitions to specific marketing touchpoints (e.g., exposure, click) to evaluate channel contribution.

Attribution Steps

Define the attribution model with flow conversion data and touchpoint data.

Assign each transition to the last touchpoint that influenced it.

Output per‑channel metrics: reach, conversion count, conversion rate.

Attribution Case Study

Using the DEEPLINK model, a brand measured how different channels moved users from the discovery layer (D) to deeper layers, identifying the most efficient channels.