Fictional worked example
Should we lower the price?
Sales access helps describe what happened. It does not, by itself, tell you what to change.
Every business, input, and connection status here is constructed. No account was accessed and no connector was tested. This example shows the output a review should produce, including the evidence still needed.
The question and the evidence
A category manager sees units fall from 120 in August to 90 in September for a $24 two-pack. A screenshot shows another offer at $15. The proposed action is to reduce the two-pack to $21.
The supplied fictional sales extract contains only those two monthly unit totals and the $24 price. The offer screenshot contains a price but no pack size, shipping, seller, availability, or observation date. There is no cost record or traffic history.
The evidence does not support a price cut yet.
Units fell 25%. That observation is checkable: (90 − 120) ÷ 120 = −25%. The cause is not established. A $15 single item would cost $30 per two units, but an actual two-pack at $15 would tell a different story.
A decision-to-source map
| Question | Required source | Status in the fictional review | What remains unproved |
|---|---|---|---|
| Did sales decline? | Internal sales extract | Two supplied monthly rows inspected; portal access described by owner. | Authentication, full catalog coverage, cancellations, and repeatable refresh are not tested. |
| Are competing offers comparable? | Current offer detail and product identity | One incomplete screenshot supplied. A possible marketplace connection is documented only. | Pack size, comparable specification, availability, delivered price, and permitted access. |
| Can we afford $21? | Dated product cost, channel fees, shipping, and returns | No extract supplied; accounting system named, access unverified. | Contribution at the current and proposed price. Revenue is not margin. |
| Is price the reason for fewer units? | Traffic, conversion, stock availability, promotions, and demand | No source supplied. Analytics capability is described, not tested. | Whether fewer visits, stockouts, seasonality, or conversion explain the decline. |
Coverage: one product, two months, one incomplete competing offer. No claim about the rest of the catalog is supported. “Described,” “documented,” “data inspected,” and “question checked” are different evidence states.
The next three moves
- Verify the comparison. Ask the category owner for dated offer evidence with pack size, specification, stock, seller, shipping, and delivered price. Acceptance check: independently reproduce the per-unit comparison for this product. A permitted manual sample can answer the first question before an integration is built.
- Establish the price floor. Ask finance for the effective cost and fees for this SKU and channel. Acceptance check: finance can reproduce contribution at $24 and $21 using the same units and cost date. If inputs are missing, withhold the margin conclusion.
- Separate demand from conversion. Ask the channel owner for visits, conversion, stockout days, and promotion dates for both months. Acceptance check: periods and SKU identity align, and coverage gaps remain visible. Then decide whether a bounded price experiment is warranted.
Priority: prevent a false comparison first, establish commercial consequences second, investigate the cause third. Effort, access, refresh frequency, and any vendor cost still need confirmation. No purchase or account permission is assumed.
What not to connect yet
Adding unrelated collaboration, CRM, or financial feeds would not resolve this decision. Do not request broad access because a connector exists. First prove a useful answer with the smallest relevant sample, then decide whether repeatable access earns its maintenance cost.
Handoff: category manager owns the decision; finance validates the cost basis; the channel owner validates offer and demand evidence. No price change is authorized by this review. The next review can change the recommendation when the missing evidence arrives.