Many small businesses begin with what they know how to make.
A founder has a good recipe, so production starts.
A factory has spare capacity, so more units are produced.
A trend becomes visible, so inventory is ordered.
A new idea emerges, so the Product is completed before Marketing begins.

That sequence is not always wrong. The risk appears when Production Commitment moves faster than Evidence of Demand. Once production has happened, some costs have already become sunk, and the organization faces pressure to “sell what we have” rather than reconsider what the market actually wants.

A Demand-driven approach changes part of that sequence. It does not require perfect information. Instead, before making a larger commitment, it asks who the customer is, whether the problem or Need is real, whether the proposed solution creates enough Value, whether customers will exchange time, information or money for it, and whether the economics remain workable.

Strategyzer recommends beginning Business Idea testing by identifying Critical Assumptions around Desirability, Feasibility and Viability and using Experiments to build Evidence before increasing commitment.

Demand-driven means letting Evidence lead investment instead of letting confidence lead production
A “produce first, find the market later” model creates risk because Product, Quantity and Investment decisions are made before the business understands what customers genuinely need, what they will pay and whether enough Demand exists.

The U.S. Small Business Administration recommends examining Demand, Market Size, Customer Characteristics and the Competitive Landscape when using Market Research to confirm and improve a Business Idea. Business.gov.uk similarly recommends testing Customer Needs, Competition, Route to Market and Willingness to Pay, including the use of a Minimum Viable Product before building the full business.

For an SME, the transition can begin with a simple sequence:
Demand Signal → Hypothesis → Small Test → Customer Behavior → Decision Rule → Scale

Demand-driven does not mean “only produce after receiving an order”

For SMEs, Demand-driven is better understood as a Decision Making principle than as one fixed Supply Chain model.
The core idea is:
Higher Demand uncertainty → Smaller commitment
Stronger Demand evidence → Larger commitment becomes more defensible
Some businesses can accept Pre-orders before production.
Others need Samples, minimum Stock or Capacity before customers can buy.
The objective is therefore not necessarily Zero Inventory.

It is: Do not commit substantially more than the current Evidence can support.

Business.gov.uk describes an MVP as a basic version of a business that can be taken to market to test Demand, gather Feedback and improve plans without first building everything at full scale.

Separate a Demand Signal from Demand Evidence

SMEs often see Signals before they see actual Demand.
For example:

• Search Volume increases
• Social Media conversation grows
• Competitors launch similar products
• Customers say an idea is “interesting”
• A post generates strong Engagement

These can be useful signals that say: “This may deserve investigation.”
They do not yet prove: “Customers will buy our Product, at our Price, through our Channel.”
Search Interest ≠ Sales
Social Mentions ≠ Market Demand
Liking ≠ Willingness to Pay
Purchase Intent ≠ Actual Purchase
Signals are useful for forming Hypotheses.
Evidence should progressively move closer to actual Customer Behavior.

Strategyzer distinguishes experiments by Strength of Evidence and notes that evidence closer to real-world buying behavior provides stronger support for questions about whether customers will pay.

Ask “What must be true before we produce more?”

Suppose an SME wants to launch a new Healthy Snack.
Instead of asking only: “Does this taste good?”

Break the Business Idea into assumptions:

• The Target Customer has a meaningful Need for Healthy Snacks
• The Product addresses that Need
• Taste reaches an acceptable level
• Customers will pay the Price required by the business
• Customers may buy repeatedly rather than only try once
• Margin remains workable after Production, Packaging, Delivery and Channel Cos

The business does not need to test everything simultaneously.

Strategyzer recommends prioritizing assumptions with high Business Impact and weak existing Evidence because disproving a Critical Assumption early can prevent investment in less important parts of the idea.

Replace the Big Launch with an Evidence Ladder

An SME does not need a large Research Project to begin validating Demand.
Evidence can be developed in stages.

1. Understand the Problem
   Talk with Target Customers to understand the Need, current alternatives and unresolved frustrations.
2. Test the Proposition
   Use a Concept, Sample, Prototype or Mock-up to see whether customers understand and value the offer.
3. Test Commitment
   Use a Waitlist, Request for Quote, Pre-order, Deposit, Booking or Paid Pilot where appropriate.
4. Test Real Usage
   Run a Small Batch or Pilot and observe Purchase, Usage, Complaints and Operational Reality.
5. Test Repeatability
   Examine Repeat Purchase, Retention, Margin and Cost-to-Serve before scaling.

The U.S. SBA identifies Surveys, Questionnaires, Focus Groups and In-depth Interviews as direct Research methods, while Business.gov.uk recommends MVPs as a lower-cost way to take an early offer into the market and learn before full development.

Example: Moving from 5,000 units before launch to production that increases with Evidence

Suppose an SME plans to produce 5,000 bottles of a new beverage to achieve a lower Unit Cost.
A Produce-first sequence might be: Produce 5,000 → Launch → Run Ads → See whether it sells

A Demand-driven sequence could be:
Round 1: Interviews and Concept Testing to understand Need, Occasion and Price Expectations
Round 2: Produce limited Samples and collect Feedback from Target Customers
Round 3: Open Pre-orders or sell a Small Batch through one or two Channels
Round 4: Examine Conversion, Actual Selling Price, Repeat Interest, Margin and Complaints
Round 5: Increase Production when Evidence clears a predefined Decision Rule
This is an illustrative workflow, not a universal number of stages.

BEE INTERPRETATION: The important change is dividing one large Investment Decision into smaller decisions so market Evidence can still change the Product, Price, Quantity or Channel before a major commitment is made.

Demand-driven requires Decision Rules, not simply more testing

Having more Data does not make a business Demand-driven if nobody has agreed what evidence would change the production decision.
Before a test, define:
Hypothesis
Metric
Threshold
Time Period
Action if Pass
Action if Fail
For example:

Hypothesis: Target Customers will buy the Product at 89 baht.
Test: Sell a Small Batch of 300 units in the Target Channel.
Go: Conversion, Margin and Customer Feedback clear the agreed criteria → increase the next production run.
Revise: Interest exists but Price or Packaging is a major barrier → adjust and retest.
Stop: Paid Demand remains below the level required by Business Economics → do not increase production under the current Proposition.

The Threshold should not come from: “The numbers look good.”
It should connect to:
Break-even
Minimum Margin
Capacity
Inventory Risk
Cash Flow
and the size of the next investment.

Existing SMEs can begin with the Data they already have

Moving toward Demand-driven decisions does not require AI Forecasting or a new Data Platform as the first step.
Useful basic fields include:

• Product / SKU
• Transaction Date
• Units Sold
• Selling Price
• Discount
• Customer or Channel
• Repeat Purchase
• Returns / Cancellation
• Gross Margin
• Stock-outs and Unsold Inventory

Then ask practical questions:
Which products sell without heavy discounting?
Which products sell well initially but generate little Repeat Purchase?
How does Demand vary by Channel or period?
Which products generate strong Sales but high Inventory Risk?
Which customers repeatedly purchase?

Market Research can then address questions Transaction Data cannot answer, including why customers do not buy, what alternatives they use, whether their Needs are changing and whether a new Proposition solves a meaningful problem.

Demand-driven is therefore not a choice between “Data” and “Research.”
It connects: Market Evidence + Customer Evidence + Behavioral Data + Business Economics

The takeaway: Do not replace “produce first” with “research forever”

Demand-driven does not eliminate Uncertainty.
No business can know future Demand perfectly.
The goal is to align the size of an investment with the Strength of Evidence supporting it.

The U.S. SBA positions Market Research as a way to confirm and improve a Business Idea while examining Demand and Market Conditions. Business.gov.uk recommends validating ideas with Potential Customers and using an MVP to learn before committing to a complete build.
For SMEs, the starting process can remain simple:

1. Select the Product or Idea requiring the next investment.
2. Write down the most important Demand Assumption.
3. Examine the Data already available.
4. Find the smallest useful way to collect missing Evidence.
5. Move from Stated Interest toward Actual Behavior where possible.
6. Define Go / Revise / Stop Criteria before seeing the result.
7. Increase Production when the Evidence supports the next commitment.

The key difference is: Produce-first asks, “What can we make, and how can we sell it?”
Demand-driven adds, “What do customers need, how strong is our Evidence, and how much should we commit right now?”

An SME does not need to know everything before starting.
But before spending more the next time, it should know more than it did the time before.