When sales decline, the first response should not automatically be a discount, a larger advertising budget, or a product change. Start by checking whether the decline is real relative to an appropriate comparison period. Then break sales into order volume and average order value before examining customers, repeat purchases, products, and channels.

These metrics do not immediately prove the cause of the decline. Their purpose is to narrow the problem so the business can form better hypotheses and decide what evidence to examine next.

Quick Answer: What Should You Check First When Sales Decline?
Start with three business drivers:
‍Number of Customers × Purchase Frequency × Value per Purchase
‍Ask:
‍- Customers: Are there fewer buyers?
‍- Frequency: Are customers buying less often?
‍- Value: Are customers spending less each time they buy?
Then investigate when the change started, where it is concentrated, which customers are affected, and what evidence might explain it.
‍The key principle is:
‍First identify what changed. Then investigate why it changed.

Why Is “Sales Are Down” Not Enough Information?

Suppose a restaurant generated THB 500,000 last month and THB 420,000 this month.
What do we know?
‍FACT: Sales declined by THB 80,000.
‍But that number does not tell us whether there were fewer customers, lower repeat purchase, lower spending per transaction, stock problems, or weakness in one particular sales channel. More than one factor may also change at the same time.

So before asking: “What should we do?”
‍ask: “Which part of the numbers changed?”

Start With Three Sales Drivers

1. Number of Customers

Check whether the number of customers or actual buyers has declined, then distinguish new from existing customers where useful.
The right metric depends on the business model. E-commerce businesses may examine Visitors, Buyers, Orders, and Conversion Rate, while B2B businesses may focus on Leads, Opportunities, and Closed Deals.
A decline in customer numbers does not automatically prove that advertising is the problem.

FACT: Customer numbers declined.
HYPOTHESIS: Marketing may be one possible explanation.

2. Purchase Frequency

Sales can decline even when customer numbers remain stable if customers buy less often. Useful metrics may include Purchase Frequency, Repeat Purchase, Repeat Rate, and Days Between Purchases, depending on the business model.
If Frequency falls, investigate possible explanations—but remember that sales data can show that behavior changed without necessarily proving why it changed.

3. Value per Purchase

If Customers and Frequency remain relatively stable, examine how much customers spend each time. Useful metrics include Average Order Value, Items per Order, Revenue per Transaction, and Product Mix. For example, if AOV falls from THB 500 to THB 420, investigate whether customers are buying fewer items, choosing lower-priced products, responding to promotions, or encountering stock issues. Lower AOV narrows the problem. It does not by itself explain the cause.

Check “When” and “Where”

A single month-to-month comparison may be misleading when Seasonality, Promotions, operating days, or Stock Availability matter. Depending on the business, review Month-on-Month, the same period last year, and multi-month trends. Then break the result down by relevant dimensions:
‍- Customer: New vs Existing
- Product: Category / SKU
- Channel: Store / Website / Marketplace
- Geography: Branch / Sales Territory
Overall averages can hide where the problem is concentrated.

BEE Sales Diagnostic Framework

Use four questions to move from a symptom toward a more specific business problem.
WHAT - What changed?
Customer, Frequency, AOV, Conversion, Product Mix, or Channel?
WHEN - When did it start?
After a Campaign, Price Change, Promotion, or as part of a recurring seasonal pattern?
WHERE / WHO - Where is it happening, and among whom?
Which Product, Channel, Branch, Geography, or Customer Segment?
WHY - What might explain the pattern?
Once the problem is narrower, develop hypotheses and select appropriate evidence—such as Sales Data, Operational Data, Customer Feedback, Interviews, Surveys, Competitor Data, or Market Data.

Can Sales Data Tell You Why Sales Fell?

Not always. Suppose Repeat Purchase declines after a price increase. The data may support:
‍FACT: Repeat Purchase declined after the price change. But this is not the same as proving: “Repeat Purchase declined because prices increased.” Other factors may have changed at the same time.
A more appropriate statement is: HYPOTHESIS: The price change may be one factor associated with lower Repeat Purchase. Additional evidence is needed before drawing a causal conclusion.

Quick Checklist: What Data Should You Review?

You do not need to open every dashboard. Start with the metrics most relevant to the decision:
‍- Customer: Customer Count / New / Repeat
- Buying Behavior: Transactions / Frequency / Repeat Rate
- Transaction Value: AOV / Items per Order / Product Mix
- Funnel: Traffic / Leads / Conversion Rate
- Context: Price / Promotion / Stock / Channel / Seasonality / Competitor Changes

The better question is not: “What data do we have?”
‍It is: “Which data will help us make this decision better?”

FACT

A decline in total sales shows that the outcome changed, but the total alone does not identify whether Customers, Frequency, Value, or another driver caused the movement.

BEE INTERPRETATION

For SMEs, a practical starting point is to break the sales outcome into measurable drivers, then drill down by time, Product, Channel, and Customer Segment.

UNKNOWN

Even after identifying which driver changed, the business may still not know why it changed. Customer, operational, market, or research evidence may be needed to investigate the cause.

Limitations

Customers × Frequency × Value is a problem-decomposition framework, not a universal formula. Subscription businesses may focus on Subscribers, Churn, and ARPU, while B2B businesses may need Leads, Conversion Rate, Deal Size, and Sales Cycle. Metrics should also use compatible definitions and time periods.