Target Group Discovery: Framework, Models, and Case Study

Analysis Background

In industry operation scenarios, we need to segment from the perspective of “people‑goods‑place‑store” to locate advantageous, problematic, and weak sub‑groups. By understanding the driving factors that affect goal achievement within these groups, we can uncover operational opportunities and formulate targeted strategies.

Identifying target consumer groups helps with market positioning and finding potential customers. Identifying target product groups uncovers opportunity categories and new markets. Identifying target merchant groups reveals business opportunities and improves economic efficiency.

Basic Concepts

Target Group

A target group is the set of entities (consumers, products, stores, or their combinations) that need to be acted upon under a specific operational goal.

Examples:

Target merchant group: Fast‑growing low‑sales merchants such as emerging beauty and skincare brands.

Target product group: Low‑price items in categories like daily goods or small appliances that have high click‑through and ROI.

Segmentation

Segmentation is the process of dividing the analysis object into sub‑groups. Target‑group discovery relies on repeated segmentation, which can be rule‑based (dimension or behavior metrics) or model‑based (scoring, prediction, clustering).

Metrics

Metrics measure quantitative characteristics of the analysis object. They are organized into three tiers:

Tier‑1 Index: Core goal‑directing metric (e.g., GMV).

Tier‑2 Index: Decomposition of Tier‑1 (e.g., GMV = UV × purchase frequency × average order value).

Tier‑3 Index: Further decomposition of Tier‑2 (dimension classification or related metrics).

Solution

Target‑group discovery follows a “goal definition → group discovery → group definition” workflow. For example, if the goal is to increase merchant sales, the group object is merchants, and we need to find merchant types and driving factors that can boost sales.

Solution Steps

Step

Problem Solved

Goal definition

Describe business status and preliminarily segment group types based on the goal.

Group discovery

Identify which groups can achieve the goal and through which driving metrics.

Group definition

Characterize the features of the target groups.

Main Application Scenarios

Business status description

Analyze product categories with low browsing but low conversion to optimize operational strategies.

Business problem diagnosis

For a low‑share but high‑growth women’s apparel sub‑category, determine whether low GMV is due to low price, low purchase frequency, or low buyer count.

Main Models

Benchmarking (BMK)

Compare key metrics of the target against a benchmark (another period or another group) to find differences and drive actions.

Benchmark Type

Comparison Method

Another period

Difference, change rate

Another group

Difference, difference rate, TGI, proportion

Note: TGI = (proportion of a feature in the target group / proportion of the same feature in the overall population) × 100.

Quadrant Analysis Model

The quadrant (matrix) analysis uses two key attributes as axes to classify objects. It can be used for categorization (e.g., BCG matrix) or diagnosis (locating problems). Bubbles (or scatter points) represent groups, with optional size and color dimensions.

Process Flow

Target‑group discovery typically follows these steps:

Data preparation – design metrics and dimensions, define analysis objects and time windows, and build base tables.

Goal definition – set target metrics, locate business combinations (e.g., high‑share low‑growth categories) using models such as growth‑share.

Group discovery – perform basic analysis, benchmark analysis, and quadrant bubble‑chart discovery; filter by axis ranges and select groups for deeper analysis.

Scoring model (optional) – score individual entities on driving metrics to rank groups.

Group definition – conduct feature comparison with benchmarks, perform clustering, and describe each segment.

Case Reference

A fast‑moving consumer goods (FMCG) e‑commerce platform wants to identify categories with high market share but stagnant growth in Q2 and diagnose the causes.

Data preparation: Category sales amount, buyer count, purchase frequency, average purchase amount for 2021‑04‑01~2021‑06‑30 vs. 2020‑04‑01~2020‑06‑30.

Goal Definition

Use benchmark analysis to compute sales growth rate per category (current / previous period).

Apply a “sales growth rate – sales amount” bubble chart (x = sales amount, y = growth rate).

Set reference lines (x = mean or a fixed value, y = 1) to divide quadrants.

Identify categories in the lower‑right quadrant as “high‑share, low‑growth” groups.

Group Discovery

For the identified “high‑share, low‑growth” group, compute buyer count growth, purchase frequency growth, and average purchase amount growth. Plot a bubble chart (x = buyer count growth, y = purchase frequency growth, bubble size = average purchase amount growth) and divide quadrants with reference lines at 1.

Group Definition

Quadrant 1 – buyer count ↑, purchase frequency ↑ (filter out categories where average purchase amount ↓). Quadrant 2 – buyer count ↓, purchase frequency ↑ (potential “buyer‑count‑decline” group). Quadrant 3 – buyer count ↓, purchase frequency ↓ (dual‑decline group). Quadrant 4 – buyer count ↑, purchase frequency ↓ (average‑purchase‑amount‑decline group). Each group can be targeted with specific strategies such as acquisition, repurchase incentives, or product upgrades.

About Us

Alibaba Mama SDS (Strategic Data Solutions) team uses data to make merchant and platform growth strategies more scientific and effective. We provide marketing insights, strategy, value quantification, and effect attribution services for Alibaba Mama’s advertising clients.