Prepare for Agency Discovery Calls by Testing Buyer Hypotheses First
On this page
Bring a discovery agenda that separates verified account facts, a useful buyer-action comparison and questions the client must answer. Subconscious can compare controlled message or offer changes on simulated buyers, giving the agency a concrete alternative to discuss. Discovery should still explore the client's goals and allow the client to correct the agency's entire framing.
For River City Federal Credit Union, a study with 140 simulated respondents identified the weakest staff-proposed slogan. The same slogan later performed worst in live Facebook and Instagram ads. That agreement supports challenging an internal message preference before committing production effort. The record does not isolate the causal campaign lift attributable to the simulation or slogan, or establish agency close rates. River City case study
Prepare evidence without deciding the client's problem in advance
Research the account so the call can address consequential unknowns. Read the current offer, customer cases, reviews, brief and any authorized account records. Attach a source and date to each factual statement. A customer logo establishes a published customer claim; it does not establish the customer's current buying criteria or procurement authority.
Keep client statements, observed records and agency interpretations separate. Reviews can suggest questions about particular experiences without establishing the most common objection. Channel performance needs actual measurement; a modeled preference for an advertisement cannot establish channel return. Budget, approval rights and delivery constraints may be documented, supplied by accountable people or still unknown.
AgencyAnalytics' discovery guidance combines preparation with collaborative questions and an open mind. Tim Kilroy's checklist recommends identifying what the agency needs to learn and listening to the prospect. Neither source establishes that specific claims always produce better answers than open questions.
Choose the preparation that could change this engagement. If the offer or audience is still undefined, bring that gap into discovery. Testing is useful when a defined action remains unresolved, not a requirement before every call. Existing research may already answer the relevant question.
Compare an action that could change the brief
Consider an illustrative prospect selling a postcard pack. The agency's provisional question is whether the current product-name headline should become a use-led headline. This is an example of a proposed study and agenda, not a reported prospect or result.
Assume the seller confirms a complete offer: six blank, uncoated 300gsm postcards, each 148 × 105 mm, with six compatible C6 envelopes. USD12 is the total price, including applicable tax and delivery within the contiguous United States. Dispatch occurs within three business days, with arrival within ten business days of purchase. The seller offers a full refund within thirty days of delivery and pays return postage. No subscription or postage stamps are included. The seller must verify the product and fulfillment terms before fielding.
Current headline A is “Six blank postcards.” Proposed headline B is “Send six handwritten notes.” Both versions show the same product image, complete facts, price, terms and layout. The outside alternative is identical: skip this pack and continue the person's current correspondence practice. The intended audience is US adults in the delivery area considering sending a handwritten card within thirty days. Inspect which eligibility fields the actual modeled population supports; do not invent access, counts or coverage.
Randomly assign modeled respondent units to A versus the outside alternative or B versus the same outside alternative. Estimate B minus A in the probability of choosing the pack. Retain actual arm counts, exact stimuli, the estimator and uncertainty that accounts for repeated responses within each unit. Repeated answers are not additional independent respondents. The Subconscious methodology explains controlled choice comparisons and interpretation.
Before results, agree the minimum useful gain against actual creative and implementation costs with the client's marketing owner. Advance B when the estimated contrast and uncertainty satisfy that rule, retain A when evidence supports retention, or mark the choice unresolved. No result, threshold or observed purchase forecast is supplied here. The output concerns the headline's effect on modeled choice, not customer intent, comprehension or the prospect's account history.
Turn the evidence into a discovery agenda
The agenda should make a change of direction possible. For the illustrative postcard prospect, bring the following record and replace assumptions with actual evidence:
| Agenda item | Evidence brought into the call | Open question and consequence |
|---|---|---|
| Establish the business problem | Dated offer sources and seller-confirmed terms; unknowns remain labeled | What prompted the agency search, and what has already been tried? A fulfillment problem could change the scope from messaging to operations. |
| Discuss one buyer-facing action | Exact A/B versions, population basis and actual comparison if available; otherwise a proposed design | Does this comparison address the intended customer and current priority? Revise the brief if the client identifies a different audience, offer or problem. |
| Agree authority and the next commitment | Verified approval information, actual capacity and costs; no inferred decision rights | Which roles approve the purchase and the work? Name the accepting owner, evidence request and next action before proposing an engagement. |
Preserve a useful unanswered prompt: “Our working hypothesis is that the VP of Operations participates in this buying decision. Which roles actually approve the purchase, and what account evidence supports that?” A simulated choice result cannot establish approval rights. The client may also identify a different problem or decision-maker entirely.
Present a modeled finding with its exact scope: the compared alternatives, supported population, contrast, uncertainty and provisional implication. Ask what would make the finding irrelevant to this account. Do not label a planned comparison as completed or present generated commentary as customer testimony.
Leave with a corrected brief and an accountable next action
After the call, update the agenda with facts confirmed, assumptions corrected and questions still open. The agency lead should record whether the engagement addresses a message choice, a delivery issue, an account fact or another problem. The client owner accepts the scope and implementation commitment; the research lead accepts the analysis and permitted inference.
Reuse adequate evidence when the audience, alternatives, endpoint and context fit. Request additional evidence for a specific gap. Customer interviews can explore language and unmet needs, records can establish actual behavior, and a controlled comparison can resolve a defined action choice. These evidence jobs do not require a fixed sequence. How Subconscious works
Assess the preparation through actual call notes: what the client corrected, which questions remained unanswered and whether the evidence changed the proposed work. Qualitative feedback from a few calls does not show a causal improvement in close rate or deal size.
Book a decision review with the proposed discovery agenda, existing account evidence and the action that could change the brief. Agree whether existing evidence suffices or which controlled comparison would make the client conversation more useful.