Why does a prospect say one thing and then do another?

The prospect says your solution is exactly what the organisation needs. The conversation goes well. They request a proposal, describe the benefits and promise to discuss it internally. Yet no decision follows. Weeks later, the incumbent supplier remains in place. Was the prospect dishonest? That conclusion comes too quickly. What someone says in a sales conversation is not the same as a firm purchase intention, let alone an observed buying decision. A statement is made in one conversation. A decision may occur later, in a different setting, with different people, risks and practical constraints. Research on intention-behaviour gaps, social desirability and organisational buying helps distinguish these signals. Association research can generate additional, testable hypotheses, but it is neither a lie detector nor a way of predicting a signed contract from reaction times. [S01] [S02] [S03] [S08]

An account manager asks, 'Would this help your team?' The prospect replies, 'Absolutely.' That may be an entirely sincere evaluation. The solution might solve a technical problem and the presentation may clarify its potential. Yet the same person may also wonder whether their team can absorb the work, who will defend the spend to finance or what happens if implementation fails.

The error is jumping from appreciation to action. 'Is this interesting?' does not measure contract signing. Even 'I expect us to proceed' is a statement of intent with conditions attached. Morwitz's review explains that stated purchase intentions relate to subsequent behaviour but predict it imperfectly. In a meta-analysis of experiments, Webb and Sheeran found that increasing intentions did not produce an equal increase in actual behaviour. Those results cover contexts beyond B2B selling and should not be treated as a B2B conversion benchmark. [S01] [S02]

A seller who records the conversation as confirmed commitment risks taking the wrong next action. A better question is: what still needs to become true after this meeting before the organisation can buy?

Between a favourable remark and a signature lie several different kinds of information. They may reinforce one another, but each deserves its own interpretation. Record them separately in interviews, pipeline notes and research.

Layer	What you can observe	What remains unknown	Useful follow-up
Statement	'This looks good'	Politeness, appraisal or a specific intention?	'What stands out to you?'
Intention	'We would like to proceed'	How strong is it, and on what conditions?	'What needs to happen first?'
Association	'Implementation feels risky'	How context-dependent is that meaning?	'Which scenario comes to mind?'
Ability to act	Budget, authority, timing and capacity	Can this contact put the decision into effect?	'Who else needs to agree?'
Behaviour	Demo booked, documents shared, pilot launched or contract signed	Why did it happen, and will it continue?	'What concrete step actually occurred?'

Ajzen's theory of planned behaviour distinguishes attitudes, perceived social norms, behavioural control, intentions and behaviour. This table is a practical sales diagnostic, not a scientifically validated five-factor psychometric scale. Its purpose is to discipline interpretation, not to label an individual buyer. [S04]

Prospects can often describe what appeals to them. But the explanation may be partial. Price is easy to mention and easy to record. At the same time, uncertainty about colleagues' reactions, previous failed projects or a supplier's reputation may matter. These are competing explanations to investigate, not secret motives a seller is entitled to assume.

Research on self-reports shows that answers can be shaped by the questions asked and by which information is accessible at that moment. Feldman and Lynch describe how earlier questions can influence later answers. In a face-choice experiment, Johansson and colleagues found that some participants supplied reasons for an outcome that did not match their initial selection. The study does not establish that business buyers routinely fabricate their reasons. It cautions against treating every retrospective explanation as a complete replay of a complex decision. [S06] [S11]

That is why a chronological reconstruction can outperform a generic 'Why did you not buy?' question. At what point did the decision shift? Who entered the process? Which new requirement appeared? What action was blocked?

A prospect may want to preserve a relationship. They may appreciate the effort behind a demo or prefer not to challenge a presenter in front of colleagues. Friendly feedback does not automatically indicate buying intent. Equally, rejecting a proposal does not prove the earlier praise was a deliberate lie.

Social desirability is a known possible source of self-report distortion. Fisher's marketing experiments examined indirect questioning for socially sensitive items and found differences from direct questioning in some conditions. This does not mean that indirect techniques are universally better. It shows that the social context of a question can form part of the answer. [S05]

Professional roles add another layer. A departmental leader can genuinely see the value of a solution while only later having to account for cost, risk and implementation in front of management. 'This is interesting' can be true within that person's remit while the wider team reaches another conclusion. Study the decision process and responsibilities rather than only the enthusiasm of the person in the room. [S08]

During a demo, a platform may evoke 'faster', 'simpler' and 'more possibilities'. When an internal business case is being written, other meanings may become salient: 'dependency', 'migration risk' and 'who is accountable if it goes wrong?' The product has not changed. The decision question has.

This is a useful starting point for association research. An association is not a permanent label that is activated equally in every setting. Alternatives, task requirements, time pressure and social circumstances affect what people say and choose. Research on indirect measures also warns that their meaning depends on the procedure and context. [S09]

Do not measure only general brand favourability. Where appropriate, reconstruct realistic contexts: exploratory product evaluation, budget review, information security approval and final supplier selection. Study whether particular value and risk associations change across these settings. The findings can support hypotheses. They do not automatically establish why one individual opportunity was won or lost.

Sales teams like to compare their offer with two other suppliers. Yet the decisive alternative may be retaining the current arrangement. Even when the benefits of switching are clear, change demands time, internal alignment, training and someone willing to take responsibility. That can make the familiar option appealing.

Samuelson and Zeckhauser studied status quo bias through decision experiments and applied choice examples. Their work suggests a mechanism worth examining when a customer delays or retains the incumbent. It is not proof that every stalled deal reflects a cognitive bias. Sometimes staying put is objectively the better decision. [S07]

If a prospect says, 'This would help us,' ask what it would cost to do nothing, and then what it would cost to change. Those are separate calculations. Repeating the benefits of a product without examining implementation demands may leave the real decision untouched.

The enthusiastic contact does not necessarily own the budget, sign off on security or carry the operational impact. A procurement professional may define risk differently from an end user. Finance may hear 'efficiency' and ask about uncertainty around returns. IT may admire the interface yet remain concerned about integrations and maintenance.

Webster and Wind's model treats organisational purchasing as a decision process shaped by multiple influences. It provides a reason to examine roles, procedures and organisational context. It does not establish that every company has one fixed buying committee. [S08]

Map who experiences the problem, sponsors the project, can stop it and must operate the solution. Ask each role what a good decision would require. If the project sponsor associates a tool with growth while operations associates it with additional work, those are different meanings attached to the same offer. That can reflect legitimate responsibilities, not a contradiction between honest and dishonest people.

Some meanings are difficult to articulate directly. When studying trust, risk or resistance to change, indirect tasks, including some reaction-time methods, may complement interviews and surveys. The interpretation must remain tied to what the task actually measured.

Method	Appropriate use	Unsupported conclusion
Open interview	Meanings, examples and reported barriers	A complete objective account of behaviour
Questionnaire	Explicit evaluation, intention and perceived risk	A high intention guarantees a purchase
Indirect association task	Task-specific and often relative association patterns	The person's secret individual decision is known
Choice or behaviour task	Choices observed in a defined setting	One association has been proven to cause real sales
CRM and process data	Actual actions, timings and outcomes	Every lost deal has a psychological cause

Fazio and Olson review limitations in interpreting implicit measures. Greenwald and colleagues' meta-analysis reports that implicit and explicit measures both carried predictive information across the domains investigated, with variation across circumstances. That is not a validated B2B sales prediction tool. A group-level association difference does not automatically support individual classification or a numerical 'likelihood to buy' score. [S09] [S10]

Start with a bounded research question. Not 'Why do prospects contradict themselves?' but 'At which stage between an enthusiastic discovery conversation and internal approval does willingness to proceed decline, and which alternative explanations fit the evidence?'

Track a meaningful part of the customer journey. Record the first statements without immediately coding them as commitment. Ask which conditions, stakeholders and milestones follow. Later, record what actually occurred: a meeting, a security questionnaire, a trial, a purchase or no action. Compare promises with observable steps instead of relying solely on the seller's judgement.

Use open questions before introducing suggested objections. If relevant, construct short, comparable decision scenarios with explicit alternatives. Consider indirect tests only when the research question, instrument and reliability justify them. This approach tests competing explanations instead of presuming to expose a hidden motive. [S04] [S06] [S09]

'What is holding you back?' can sound as though the prospect is concealing an objection. Sometimes they genuinely do not know yet, or the barrier has not occurred. A curious question is often more useful than a challenge.

Try moving from abstract approval to decision requirements: 'What needs to be demonstrated internally before this can progress?' 'Who will live with the change each day?' 'What would make a pilot successful for you?' 'What happened during your last supplier transition?' 'What might a colleague reasonably object to at this moment?'

Separate seller inference from buyer evidence. 'I think they fear change' is an interpretation. 'They referred to two previous migration problems and asked three times about rollback' is observable information. A sound diagnostic should be clear enough for another colleague to challenge it, rather than simply reinforce the salesperson's preferred story.

Fictional illustration - no Neurofactor client, original study or measured outcome. A software vendor presents a workflow platform to an organisation. The team manager is enthusiastic: 'This could remove many manual steps.' They request a proposal and say the business case looks persuasive.

At a later meeting, IT asks about integrations. Operations wants to know how many people need training. Finance asks what happens if the pilot fails. The decision is postponed. The salesperson concludes, 'They say yes but mean no.' Another explanation may fit better: the team manager valued the product but had not yet secured an executable plan across other responsibilities.

A careful analysis compares what the manager liked during discovery, what other stakeholders valued during decision-making and which process steps actually happened. Association research might examine candidate meanings such as 'productive - disruptive' or 'control - dependency', but the exact stimuli must first be tested. An intervention might be a migration plan, a reference conversation or a scoped pilot. These are proposals for testing, not proven results.

A spoken statement tells you what someone says they believe or want. An indirect task can provide another task-dependent signal. Behavioural and process records show what actions took place. The signals may align or differ, and third factors may explain both. Use a structured comparison instead of a single 'true buying intention' index.

Pattern	Hypothesis to examine	Alternative to rule out
Positive stated intent, no next step	Execution or approval obstacle	No budget, authority or urgent need
Positive words, detailed risk questions	Benefit coexists with uncertainty	Genuine compliance requirements or prior problems
Different views from end user and procurement	Roles apply different success criteria	Actual deficiency in product, price or delivery
Indirect and explicit measures diverge	Task, context or social pressure may matter	Poor reliability, order effect or ambiguous stimuli
Pilot starts, full contract does not follow	New obstacle emerges after use	Product performance, procurement or implementation

Each row is a reason to investigate further, not an individual psychological diagnosis. Report the relevant population, measurement task, timing and uncertainty. The literature does not justify inferring a precise personal motive from the pattern alone. [S03] [S09] [S10]

A courteous conversation without a concrete intention. Treat interest as interest. Ask which problem the prospect wants to solve and what reasonable next step follows. More persuasion without a defined decision can create false pipeline progress.

Sincere intention without the ability to act. Explore authority, time, budget and interdependencies. Not every barrier is solved by stronger copy or another demonstration. The answer may involve another sponsor, different timing or a reduced scope.

An attractive solution that triggers risk associations around change. Identify which risks are real and which arise from expectations. Offer credible implementation evidence and a realistic fallback plan when possible. Do not promise what the operations team cannot deliver.

Positive approval from several people whose role priorities differ. Make each stakeholder's decision criteria visible. A story for the user is not automatically the right explanation for finance or information security. The aim is not to bypass objections; it is to make an informed decision possible. [S04] [S07] [S08]

1. Define the behaviour. Does progress mean a second meeting, shared data, pilot approval or a signed contract? Avoid vague pipeline sentiment.
2. State competing hypotheses. Include a contextual, organisational and associative explanation. Also allow for a genuine lack of fit.
3. Record the conversation. Capture exact statements, conditions, participants and dates. Do not rewrite them in light of the outcome.
4. Study comparable decision contexts. Use relevant alternatives and consistent conditions when contrasting responses across groups.
5. Include relevant roles. Map authority, execution and possible veto points without unnecessary personal profiling.
6. Link evidence to action. Record observed process steps and their timing. Beware selection bias when comparing only won and lost opportunities.
7. Test a specific change. For example, vary implementation evidence, stakeholder sequencing or the agreed next step. Measure a predefined behavioural outcome, not only presentation ratings.

Where appropriate, define an explicit if-then action: 'If security completes the documentation review, we will hold a joint risk meeting.' Gollwitzer's research on implementation intentions gives a theoretical rationale for such planning, but does not itself demonstrate a higher B2B close rate. [S12]

Suppose research suggests that prospects worry primarily about implementation feasibility. Another feature demonstration may not address that concern. A transparent migration roadmap, named responsibilities, a documented risk assessment or a conversation with the team that will carry out the change may be more appropriate.

When stakeholders interpret the decision differently, translate one truthful value proposition into relevant evidence for each role. End users need to see daily workflow effects. Sponsors need to assess organisational goals. Finance needs assumptions and uncertainty. IT needs technical boundaries. This is not manipulation when the information remains accurate, verifiable and sufficient for people to say no.

Choose one primary outcome for each proposed intervention: completion of a security review, a full business case, an approved pilot or the eventual purchase. Where feasible, compare against a suitable control or predefined baseline. Higher intention scores alone cannot establish that sales performance improved. [S02] [S03]

The scientific literature supports an imperfect relationship between intention and behaviour, social influences on self-report, possible status quo effects and multiple influences in organisational purchasing. The process described here is a practical synthesis, not a separately validated Neurofactor prediction model. [S01] [S02] [S05] [S07] [S08]

Indirect association tasks may reveal additional group-level patterns, but results depend on their design, stimuli, quality and context. EEG, eye tracking and response latencies are not direct windows into hidden motives and cannot predict an individual's signature without independent validation. The IAT meta-analysis spans numerous judgement and behaviour domains; it was not designed as a B2B funnel accuracy study. [S09] [S10]

Respect privacy and unequal power in business relationships. Be transparent about research purposes, minimise personal data and separate sales notes from research datasets where appropriate. Do not recruit a prospect into a psychological study without an appropriate basis and information. A difference between speech and action is not grounds for calling someone dishonest.

A prospect can be honestly enthusiastic yet never buy. They can be polite without intending to proceed. They can intend to switch and later face budget, authority or changed risk assessments. Sometimes no simply means no.

The point of association research is therefore not to catch prospects out. It is to distinguish what was said, what was intended, what was associated, what was feasible and what was actually done. That distinction supports better questions, more relevant evidence and more concrete next steps. Not by making the sales pitch sound more psychological, but by learning which decision the organisation really faces and what it would need to make that decision responsibly.

Related reading: types of resistance, target-group objections, measuring associations and correlation versus causation.