Use CDM’s five-constraint diagnostic to identify whether weak demand comes from awareness, affordability, availability, trust or product fit before spending more on promotion.
Documented fact: The UK government’s evaluation guidance distinguishes observed outcomes from the evidence needed to support causal explanations; a dashboard movement alone does not establish its cause.
CDM inference: The Five-Constraint Demand Diagnostic is CDM editorial synthesis. It is a decision aid, not a validated universal model or a claim that only five causes can affect demand.
Start with the symptom, not the favourite tactic
“Demand is weak” can mean several materially different things. Branded search may be flat, qualified pipeline may be declining, first purchases may have slowed, or category demand may have contracted. Each is an observation. None is yet an explanation.
Write the symptom with a unit, segment and time period. “Qualified opportunities from mid-market UK buyers fell across the last two completed quarters” is testable. “People have stopped caring about us” is not. Reconcile the observation with finance, sales and product records before building a narrative around it.
This discipline follows a basic evaluation principle: specify the outcome and the causal question before selecting a method. The UK government’s Magenta Book makes that separation explicit for public-policy evaluation. Marketing teams can use the same logic without pretending a campaign is a policy trial.
Then name at least three plausible explanations. The purpose is not to make the analysis slow. It is to prevent the easiest-to-change lever—usually media or creative—from becoming the default cause.
The Five-Constraint Demand Diagnostic
CDM’s model treats a purchase as passage through five gates. A customer can clear some gates and fail at another. The binding constraint is the gate that most limits suitable demand under current conditions.
| Constraint | Question | Useful evidence | Common false fix |
|---|---|---|---|
| Awareness | Do enough suitable people recognise the category, problem or brand at a relevant moment? | Prompted and unprompted research, branded search, qualified reach, direct interviews | Publishing more content without showing that reach is the constraint |
| Affordability | Can the buyer fund the purchase and justify its total cost now? | Price objections, finance decline, basket composition, competitor prices, willingness-to-pay research | Blanket discounting that protects volume but destroys margin or reference price |
| Availability | Can the buyer obtain the right product, location, variant, appointment or delivery window? | Stock, coverage, lead time, fulfilment, accessibility and abandoned-demand records | Buying reach while inventory or capacity remains constrained |
| Trust | Can the buyer verify identity, claims, safety, terms and post-purchase support? | Loss reasons, service contacts, review provenance, claim evidence, usability sessions | Adding generic testimonials or urgency |
| Product fit | Does the offer solve a sufficiently important job for the intended buyer better than available alternatives? | Activation, repeat use, retention, returns, outcome interviews and non-customer research | Changing the message around an offer that does not create enough value |
Awareness is often overdiagnosed because it is visible to the marketing team. Availability problems can hide behind successful reach: a national campaign drives interest to areas the company cannot serve. Affordability differs from willingness to pay; a buyer may value the offer but lack budget authority or cash flow. Trust differs from familiarity; a recognised brand can still leave a material claim unproven.
Product fit is deliberately last, not because it matters least, but because teams often use it as an unfalsifiable verdict. Define the job, segment and competing alternative. Fit for one situation says little about another.
Build a constraint evidence packet
Create one page per constraint with three columns: supporting evidence, contradicting evidence and missing evidence. Use behaviour, operational records and direct research together. Behaviour shows what happened; interviews and observation can illuminate why; operational evidence shows whether the promise could be fulfilled.
CDM asset: Constraint Evidence Card
- Observed symptom: metric, definition, segment and period.
- Proposed mechanism: how the constraint is expected to produce the symptom.
- Supporting signals: sources that agree, with dates and denominators.
- Disconfirming signals: evidence that should reduce confidence.
- Nearest alternative: the most plausible competing explanation.
- Decision threshold: what evidence would change the next action.
Do not treat interface friction as proof of weak trust, or survey intention as proof of future purchase. Check whether different sources converge. If conversion fell at the same time that stock availability deteriorated, repairing the form may improve a local rate without recovering completed orders.
Also inspect whether the design of the journey is shaping the evidence. The Competition and Markets Authority’s work on online choice architecture documents how digital design can affect decisions. A pressured completion rate can therefore be a poor proxy for durable demand.
Choose the smallest decisive test
Rank constraints by likely impact, confidence and cost of learning. Start with a reversible test that can separate the two strongest explanations. If buyers reach a product page but repeatedly ask whether migration is supported, create a precise migration evidence path and observe qualified progression, support demand and post-purchase outcomes. Do not call a colour test a trust experiment.
Predefine the expected sequence. If trust is the constraint, engagement with proof should change before completed purchases; cancellations or complaints should not worsen. If availability is the constraint, expanding serviceable inventory should change completed orders in newly covered segments, not merely clicks.
Set a stop rule. A constraint diagnosis should lose priority when its predicted leading indicators do not move, when a guardrail deteriorates, or when new evidence makes an alternative more plausible. The aim is to learn, not defend the initial theory.
Worked example: a regional service business
Consider an illustrative home-services platform. Enquiries have grown, but completed bookings have not. The media team proposes more brand investment because unprompted awareness is modest. The operations review shows that evening appointments—the most requested slot—are unavailable in three high-volume areas. Customer interviews show that many households abandon after seeing the next available date.
The binding constraint is provisionally availability, not awareness. The team tests additional evening capacity in one matched area, preserves prices and creative, and measures eligible demand, booking completion, fulfilment quality and repeat use. If bookings rise only where capacity changes, the evidence supports the availability mechanism. It does not prove that awareness is universally sufficient.
The next constraint may then emerge. More bookings could expose service-quality limitations. Diagnosis is a cycle, not a permanent label.
Limits and risks
Demand systems contain interactions. Price can signal quality; distribution can affect trust; product changes can alter affordability. Five categories make discussion manageable, but they do not remove complexity. Some causes—regulation, category contraction, macroeconomic shocks or competitor innovation—sit partly outside the model and the team’s control.
Small samples and weak instrumentation can create false certainty. Use ranges, preserve contradictory evidence and seek specialist review for regulated claims, pricing, credit or vulnerable audiences. If an unsafe or misleading practice is discovered, correct it immediately rather than waiting for a complete diagnosis.
CDM verdict
Weak demand is not a media brief. It is a diagnosis problem. Separate the observed symptom from its mechanism, test awareness, affordability, availability, trust and product fit, and find the constraint that currently limits suitable demand. The best next move is the one that produces decision-quality evidence while protecting customer and commercial outcomes.
Related guides
Frequently asked questions
Which demand constraint should a team test first?This article is editorial guidance. Apply the principles in proportion to your market, evidence, and responsibilities.
