# How do you choose the next growth question when customer observations point in different directions? Compare the circumstances behind customer observations before choosing a growth direction. Similar requests can describe different work, an adequate approach or your own delivery problem. Keep episodes, sources and competing explanations separate. Choose a learning question that could change a decision, including whether to improve the current business or pause. You do not need a selected segment, offer or campaign to begin. The immediate output is a bounded investigation, not a declaration of demand. ## Start with episodes, not the most repeated phrase An episode describes what someone was trying to do, the circumstances, the approach they used and what happened. A request for better reporting leaves those details unresolved. Ask what led to the request before treating it as a need for a dashboard. A fashionable topic or widely repeated phrase is not an episode. Ask what, if anything, changed in the person's work before treating attention as a recurring need. Retain the permitted source reference, actual capture date, who described or observed the work, and limits of access. Separate recollection, workflow observation and a colleague's retelling. Repeating one account still describes one episode. Hearsay can suggest a question; it cannot establish that the event happened as described. Compare relevant circumstances before grouping records by industry. The same decision responsibility, input timing or workaround may occur across industries. Similar company descriptions do not establish similar problems. ## A completed episode sheet **Entirely hypothetical exercise:** every account, episode, source record and behaviour below is invented. No actual interviews, observations, quotations, dates or customer findings are reported. Better reporting is illustrative wording, not a customer quote. The letters identify fictional organisations. Imagine a team hearing reporting requests before choosing an offer or market: | Record and fictional source | Episode and current approach | What the record would support within this exercise | What remains interpretation or unknown | | --- | --- | --- | --- | | E1, account A: participant's recollection, with an existing report described | A project decision waited despite accurate information being available. Participants sought confirmation of who could decide. | A described decision delay and uncertainty about responsibility in this episode. | Unclear ownership may matter; competing priorities or unavailable participants may also explain the wait. No causal diagnosis. | | E2, account B: recollection and an input checklist described | Report preparation involved reconciling late, inconsistent inputs. A spreadsheet and email were used to resolve them. | Input reconciliation work in this episode. | A dashboard might display the same uncertain information. Its usefulness and the reason inputs differed are unknown. | | E3, account C: a described routine review and shared report | A named owner used the existing report to make the required decision. The participant described no reason to change the process. | An account of an adequate existing approach for this situation. | It does not establish adequacy for every task, but contradicts assuming all reporting work needs replacement. | | E4, account D: an illustrative delivery note from our team | Our onboarding stalled because our configuration required a customer-specific field mapping we had not prepared. | A condition of our own delivery was unmet in this case. | It does not establish a wider buyer problem or demand for a new reporting service. | | R1: a sales colleague repeats E1 | The account A delay is retold as another reporting request. No separate episode is described. | Another route by which E1 reached the team. | It adds no independent instance. The sales interpretation needs checking against E1. | E1 and E2 describe different unresolved work. E3 challenges the general dashboard idea. E4 may call for internal improvement. R1 must stay connected to E1 rather than inflate repetition. ## Compare explanations and the observations that could weaken them These records are not a vote. Even independent episodes cannot establish market prevalence. Compare the interpretations: | Possible explanation | Supporting clue and contrary case | Observation that could weaken it | Learning decision now | | --- | --- | --- | --- | | Decisions wait because relevant information is unavailable | E2 describes input problems. E1 describes accurate information already available; E3 describes an adequate report. | With permission, reconstruct when required information became usable relative to a delayed decision. | Separate input preparation from decision responsibility; do not select a general visibility offer. | | Decision responsibility is unclear in some project circumstances | E1 describes seeking confirmation. It could instead concern availability or competing priorities. | Reconstruct who was authorised, what they knew and what prevented action. Clear, understood responsibility would weaken this explanation. | Investigate the E1 circumstances without presenting an ownership solution first. | | Reporting requests identify a repeatable new service opportunity | E1 and E2 raise questions, but E3 reports adequacy, E4 concerns our delivery, and R1 repeats E1. | Compare an independently described similar task where the current approach works, and ask why. | Leave the service hypothesis open. Neither shared wording nor these records establishes demand. | Handle E4 separately. Review our configuration requirement before interpreting the delay as a customer need. Improving preparation could be worthwhile without a new offer. It would not explain other accounts' reporting difficulties. ## Choose one next question for its decision value Account A's fictional contact has offered an exploratory conversation. Permission to inspect a record is unagreed. This offers a learning route, not buying intent. **Filled decision note:** investigate what prevented action in the delayed decision described in E1. Request a consented reconstruction using a non-sensitive or anonymised example, if the contact wishes to share one. Compare when information became usable, who could decide and what else competed for attention. Ask about a comparable decision that proceeded adequately. If responsibility was understood and another constraint explains the wait, revise the ownership hypothesis. If the existing approach worked in comparable circumstances, narrow the investigation or pause it. If a relevant uncertainty remains, examine an independent episode before proposing a service. If permission or access is unavailable, choose another feasible learning route; do not replace evidence with pressure to participate. The business lead reviews the question after the reconstruction, or sooner if access closes. The delivery team investigates E4 separately. No dashboard commitment, market shortlist or demand estimate follows. ## Use the same learning for different role decisions A **CEO** can limit investment to resolving the uncertainty that matters before funding a growth initiative. Better core delivery or a pause remains a valid outcome. A **marketer** can distinguish a topic people mention from the work they need help understanding. E1 and E2 should not automatically receive the same campaign. A clearer message cannot validate an unestablished need. A **business lead** can examine whether comparable work recurs and whether the team can deliver it within a credible scope. Our configuration shortfall in E4 is not evidence that customers need a repeatable new service. ## Keep a reusable synthesis record | Field | What to enter | | --- | --- | | Episode and provenance | Task, circumstances, consequence, current alternative; permitted source and actual capture date. Link retellings to their originating episode. | | Evidence and interpretation | What was described or observed; what you infer and what remains unknown. | | Competing explanations | Alternatives consistent with the account, including our delivery conditions. | | Contrary case | Comparable work that went adequately or conflicts with the preferred explanation. | | Next learning question | A feasible, permissioned investigation and an observation that could weaken the assumption. | | Decision and review | Improve internally, investigate, narrow or pause; owner and review point. | Update the record when evidence changes. Once a buying situation becomes specific enough to challenge, [evaluate its buyer value, alternatives and segment assumptions](https://buyervalues.ai/growth/en/evaluate-a-new-uk-segment-with-buyer-value.md). That guide uses a hypothetical UK case; buyer conclusions require fresh evidence in each context. ## Sources and context - [Stanford d.school: Design Thinking Bootleg](https://dschool.stanford.edu/tools/design-thinking-bootleg) — general methods for exploring and defining problems and testing ideas. - [U.S. Small Business Administration: Market research and competitive analysis](https://www.sba.gov/counseling/plan-your-business/#market-research) — general guidance on combining existing information with direct audience research and examining alternatives. Buyervalues.ai publishes this guide. The synthesis sheets are our editorial application of general research principles, not a validated diagnostic instrument or a claim that the platform automates this work. Neither source verifies the fictional records or establishes demand in any market. ## Related reading - [Where might your next UK growth opportunity come from?](https://buyervalues.ai/growth/en/find-your-next-uk-growth-opportunity.md) - [How should a Finnish B2B software company evaluate a new UK customer segment?](https://buyervalues.ai/growth/en/evaluate-a-new-uk-segment-with-buyer-value.md) --- Prepared and editorially reviewed with AI assistance. Publisher: [Buyervalues.ai](https://buyervalues.ai/?bv_article=%2Fgrowth%2Fen%2Fturn-customer-observations-into-a-growth-question.md). Updated: 2026-10-03 · Editorial review: 2026-10-03 · Revision: 1 · Locale: en-GB Canonical URL: