From Data to Insight in a New Era

Sales have grown, but has promotion spending also increased? Has greater exposure improved click performance? Have more clicks translated into more sales? As the business expands, how is the performance of each Amoeba organization changing?

Omnichannel operations connect these questions more closely. Channel types, stores, product lines, products and listings offer different perspectives, while sales, expense and promotion data capture different parts of the business.

DONGJIAN combines these dimensions and metrics through 12 chart formats and cross-dimensional drill-down. Teams can spot differences, trace their sources and use business evidence to develop insights they can test.

1. Start with the business decision the team needs to make

“How have sales been recently?” is usually just the first question.

Business leaders may ask which channels drive growth and whether it is sustainable. Operations teams need to identify stores, products or listings worth pursuing. Promotion teams want to understand the impressions, clicks and attributed sales generated by additional spending. Amoeba organization leaders need to see the contribution these activities ultimately deliver.

Different questions call for different analysis.

To track sales progress, examine trends and comparison baselines. To assess promotion efficiency, connect spending, impressions, clicks and attributed sales. To understand operating performance, also consider internal transfer prices, refunds and other expenses included in the analysis.

Before organizing the data, clarify which decision the analysis will support, which objects matter, what to compare them against and who will follow up on the findings.

These questions define the scope and provide a sound basis for selecting metrics and drill-down paths.

2. Put sales, expenses, traffic and conversion in the same business context

A set of metrics usually describes only part of the business. Examining different metrics together, while respecting their definitions, helps explain changes step by step.

Sales outcomes: how scale changes and who contributes to growth

Sales revenue, units sold, order counts, customer counts and new versus returning customer metrics help explain business scale and customer mix.

Refund amounts, refunded units and net units add another perspective. Considering refunds alongside sales growth helps reveal the actual operating situation behind that growth.

Break down these results by channel type, store, product line, product and listing to identify the main contributors.

Expenses: how many resources went into these results

Promotion spending and advertising expenses record paid traffic acquisition. Together with commission-based promotion and other included operating expenses, they help teams examine the scale, composition and timing of spending.

Platform promotion expense ratios and channel expense ratios relate promotion expenses to sales within the corresponding scope.

Expense analysis requires clear inclusion and allocation rules. Spending can only be meaningfully compared with sales when both use a consistent reporting scope.

Traffic acquisition: the reach and clicks generated by spending

Promotion impressions indicate exposure, clicks indicate further visits or interest, and click-through rate describes the step from impression to click.

Average cost per click and cost per thousand impressions add a cost perspective to traffic acquisition.

For example, when clicks increase, compare promotion spending and average cost per click to understand how click growth relates to spending scale and acquisition cost.

Conversion: how sales outcomes change after traffic changes

To investigate conversion, examine promotion-attributed revenue, orders and customers alongside overall sales performance.

Return on ad spend connects promotion-attributed revenue with promotion spending, showing the relationship between input and output within the attribution scope.

Distinguish click-through rate, sales conversion rate and return on ad spend: each describes a different stage. Calculating conversion rate also requires matched click or visitor data and attributed orders or paying customer counts.

Keep promotion-attributed sales and total order results distinct, then compare them with their respective scopes clearly defined.

Operating contribution: the Amoeba performance behind growth

Building on sales and promotion analysis, use the company’s internal transfer prices, connected promotion spending and other operating expenses to assess each Amoeba organization’s operating contribution.

Return on ad spend explains the relationship between attributed revenue and promotion spending. Internal settlement costs, refunds and other operating expenses add to the assessment of operating performance.

Teams can then discuss business scale, spending efficiency and operating contribution together.

3. Five comparison relationships give analysis a clear direction

The same dataset can support different topics. The choice of comparison shapes what the team notices first.

Composition: understand the structure of resources and outcomes.

Which channels contribute most sales? Which stores account for more promotion spending? How does a product line’s sales share compare with its share of expenses?

These comparisons reveal business priorities, resource distribution and concentration.

Category comparison: identify differences between objects.

Compare stores, product lines or products using sales scale, incremental sales, promotion spending, average cost per click or return on ad spend.

Each ranking addresses a different question. Large scale, strong incremental growth and low acquisition cost provide different business clues.

Time series: understand when changes happen.

When did promotion spending start increasing? When did impressions and clicks change? Did attributed revenue and sales change around the same time?

The sequence helps narrow the investigation. Campaign and supply conditions can then be used to test possible explanations.

Distribution: examine differences within the whole.

Which price bands account for transactions? Are sales concentrated in a few products or listings? Is spending concentrated on a small number of business objects?

Averages summarize the overall level; distributions and concentration reveal the internal structure.

Correlation: observe how multiple metrics move together.

As expenses increase, what happens to clicks and attributed sales? Does click-through rate remain stable as exposure expands? How does the channel expense ratio change as sales grow?

These relationships suggest where to investigate. Explanations still need to be tested against business evidence.

4. Twelve chart formats for different business questions

Once the topic and comparison are clear, chart selection has a purpose.

DONGJIAN’s 12 chart formats can be applied flexibly to sales, expenses, traffic and conversion. Each metric is analyzed within the data coverage and business dimensions it supports.

1. Time-series trends

Track sales, promotion spending, impressions or click-through rate over time to find turning points and fluctuations. Cumulative views also show how scale and gaps build over a period.

2. Dual-axis column and line charts

Compare two metrics in one view, such as promotion spending and attributed revenue, or impressions and click-through rate, to examine spending scale alongside another business indicator.

3. Stacked trends

See each object’s actual contribution within the overall trend, such as sales or promotion expenses by store, to understand where growth and spending originate.

4. Stacked composition

Track shifts in sales or expense shares to see changes in resource allocation and operating contribution. Interpret proportions alongside actual scale.

5. Doughnut composition

Identify the main contributors to sales or expenses in the selected scope, then break down the objects of interest to understand their composition.

6. Bubble analysis

Compare several metrics at once. For example, examine objects by promotion spending, attributed revenue and clicks to identify differences in scale, input and output.

7. Radar charts

Compare metric profiles on a consistent basis, such as sales, impressions, clicks and promotion efficiency, then use the underlying values to assess the differences.

8. Comparison with the mean

Add a reference within the group. Alongside spending scale, compare average cost per click with the group mean to find objects that warrant investigation.

9. Funnel charts

Arrange objects by sales or expense size to show concentration in the head and long tail, helping explain category scale and business dependence.

10. Ranked bars

Identify priorities using incremental sales, promotion spending or channel expense ratios to distinguish growth contribution, spending distribution and efficiency.

11. Price distribution

Examine transaction price bands and changes in their contribution to understand how changes in realized unit price relate to product and listing sales mix.

12. Heatmap matrices

Cross objects with dimensions such as time to scan for concentrated changes in sales, expenses, impressions or clicks, then use details to narrow the investigation.

An analysis might start with trends to locate the period of change, use rankings and composition to identify priority objects, then add bubble, radar or mean comparisons for further context.

5. Cross-dimensional drill-down traces changes to specific business objects

Channel types, stores, product lines, products and listings provide connected perspectives on operations.

Starting with channel type, identify channels with significant sales or expense changes, then examine individual stores. Within a priority store, break down product lines and products to see where results are concentrated.

Starting with a product line, compare its scale and growth across channel types and stores, then investigate related spending and promotion performance.

Starting with a product, examine its sales listings to distinguish each listing’s actual contribution to orders and revenue.

Different data sources may support different levels of detail. Compare expense and promotion metrics within their available scope; trace sales and orders through established relationships. Clear boundaries help teams understand how analyses relate to one another.

This cross-analysis tests a local finding within the broader business context.

When a store’s sales rise, examine which products grew. When spending on a product increases, compare its performance in other stores to formulate more specific investigation questions.

6. Connect expenses, traffic, sales and operating performance in one analysis

Suppose the team notices that sales in a channel type have grown recently while promotion spending has also increased significantly.

Step 1: establish the sources of sales growth.

Break down channel types, stores and product lines. Use current scale, incremental growth and growth rates to identify major contributors, then examine related products and listings.

Step 2: locate the increase in spending.

Compare the timing and composition of promotion spending to determine whether increases are widespread or concentrated in a few stores or products.

Step 3: understand changes in traffic acquisition.

Use impressions, clicks and click-through rate to assess the reach and engagement generated by increased spending. Check average cost per click and cost per thousand impressions to understand acquisition cost changes.

Step 4: investigate sales outcomes.

Compare promotion-attributed revenue and return on ad spend using a consistent attribution basis. Then examine orders, customers and sales to understand changes in outcomes and the mix of new and returning customers.

Where matched visitor, paying customer or attributed order data exists, analyze conversion rates for the relevant stage as well.

Step 5: return to operating contribution.

Use internal transfer prices, refunds and other connected operating expenses to assess the Amoeba organization performance associated with growth.

This path produces more specific questions: which objects need lower acquisition costs? Which need investigation of the path from click to purchase? Which products combine sales expansion with operating contribution?

Business teams can check these questions against campaigns, supply and operational actions, connecting analysis with subsequent business plans.

7. An insight needs evidence and a clear next step

A useful business insight explains what happened, what supports the interpretation and what should happen next.

“Promotion expenses increased” is a fact. Adding which stores account for the increase and how impressions, clicks and attributed revenue changed identifies the objects that deserve attention.

Verified campaign plans, product changes and supply conditions help teams build explanations. Unconfirmed explanations become follow-up investigation tasks.

Different metrics also require consistent aggregation methods.

For example, aggregate click-through rate by dividing total clicks by total impressions. Aggregate average cost per click by dividing total promotion spending by total clicks. This accounts for each object’s actual scale and keeps comparisons consistent.

Action recommendations should name the objects, owners and observation period: which stores or products need follow-up, what will change, which metrics will be monitored and how the next review will assess results.

Continued testing refines business insights and turns them into the team’s accumulated experience.

8. From data to insight in a new era: keep pursuing business questions

Omnichannel operations continuously generate new data and increasingly detailed decision needs.

Different charts help teams identify differences quickly. Cross-dimensional drill-down takes findings further, while charts paired with details provide values that can be checked.

Sales describe business scale; expenses record resource inputs; impressions and clicks describe traffic acquisition; attributed sales and orders show sales outcomes; internal settlement costs and operating expenses help explain Amoeba organization performance.

Together, these elements form the basis of business analysis.

Business experience connects data findings with reality. Teams need to understand metric definitions, investigate the context of changes, arrange actions and observe subsequent results.

DONGJIAN aims to make this process easier: organize data around questions, choose meaningful comparisons, trace sources across business dimensions and bring findings into daily operations, then test the decisions in the next review.