Why Marketplace Sales Fell: A Root-Cause Workflow

A sales decline is visible immediately, but the cause is not. Cutting price, increasing bids, or ordering inventory before identifying the mechanism can make the problem worse.

Diagnose sales as a sequence: availability, discoverability, traffic, conversion, and economics. Act where the sequence breaks.

1. Confirm the comparison is fair

Choose comparable periods and note promotions, holidays, stock events, catalog changes, and reporting delays. Compare both a short window and a longer baseline. A weak week inside a strong quarter needs a different response from a sustained structural decline.

Check data freshness before diagnosis. Missing recent orders or advertising data can manufacture a decline.

2. Decompose the revenue movement

Revenue changes through a small number of mechanisms: products start or stop selling, units change, price changes, and availability removes potential demand. Build a bridge from previous-period revenue to current revenue so the contribution of each mechanism is visible.

Start at portfolio level, then drill into the few products or categories explaining most of the movement. Avoid reviewing every SKU equally.

3. Follow the product funnel

For a declining product, inspect gates in order:

1. Live and available: is the offer active with stock?
2. Competitive: is the Buy Box or equivalent position stable?
3. Discoverable: did impressions or organic visibility fall?
4. Engaging: did clicks or product-page visits decline?
5. Converting: did the conversion rate fall?
6. Profitable: would recovering volume create contribution?

Stopping at the first meaningful break prevents random action. No-stock products do not need more advertising. Stable traffic with weak conversion does not first need more traffic.

4. Separate common root causes

Use these as hypotheses, not automatic conclusions. Validate each with the relevant evidence.

5. Prioritize by explained impact

Rank root causes by euros of movement, confidence, recoverability, and time to act. A high-confidence stockout on a major product deserves immediate ownership. A possible ranking change on a small product may wait for more evidence.

Assign one action to each material cause. “Monitor” is valid only with a date and threshold that triggers escalation.

6. Avoid response collisions

Teams often change price, bids, content, and stock strategy together. That makes recovery hard to attribute. Separate urgent containment from the controlled improvement test.

For example, restore stock first, then observe visibility and conversion before raising bids. If the Buy Box is lost, fix competitiveness before paying for more impressions.

7. Close the loop

Record the diagnosis, action, owner, expected mechanism, and review date. After the observation window, compare actual recovery with the expectation. Wrong hypotheses are valuable when retained; they improve the next diagnosis.

Look for repeated causes across products. Many individual declines caused by stock timing indicate a procurement process issue, not a set of isolated SKU problems.

8. Worked example: the decline advertising could not fix

A product’s weekly revenue falls 35%. The campaign dashboard also shows fewer clicks, so the first instinct is to raise bids. The portfolio bridge shows most lost revenue comes from fewer units, not price. Product diagnosis reveals stock remained available, but the offer lost the Buy Box shortly before impressions and clicks declined.

The causal order matters. Lower visibility is downstream of competitiveness. Raising bids may buy some impressions while the weaker price or delivery promise continues to suppress sales and margin. The first action is to inspect offer competitiveness and determine whether regaining the position remains profitable.

After a delivery setting is corrected, the selling position returns. The team waits for traffic to recover before changing bids. Sales improve without additional advertising cost. The recorded diagnosis helps operators recognize the same sequence on other products.

Contrast this with a product whose impressions remain stable while conversion falls after negative reviews. The revenue bridge may look similar, but the repair belongs in product quality, listing communication, or assortment—not price or traffic. A fixed funnel prevents both problems from receiving the same response.

Diagnosis note template

• Observed movement and comparison periods.
• Products explaining most of the euro change.
• First broken funnel gate.
• Evidence supporting and contradicting the hypothesis.
• Containment action, improvement action, and owner.
• Expected recovery signal and review date.
• Result and learning after review.

9. Practical checklist

• Date windows are comparable and data is fresh.
• Portfolio movement is decomposed before product drill-down.
• Products are ranked by explained revenue impact.
• Funnel gates are checked in a fixed order.
• Root causes remain hypotheses until evidence supports them.
• Profitability is checked before recovering volume.
• One owner and review date exist per material action.
• Recovery is measured against the diagnosed mechanism.
• Repeated causes become process improvements.

The purpose of diagnosis is not to produce more charts. It is to reduce the number of plausible explanations until the next action becomes obvious and measurable.

Turn repeated diagnoses into operating improvements

Tag completed diagnoses by root cause and review the pattern monthly. If stockouts explain repeated declines, improve purchasing and lead-time assumptions. If Buy Box losses recur after price updates, review repricing governance. If conversion declines cluster around incomplete attributes, improve listing-quality controls before launch.

Track how much of each revenue movement the team can explain with evidence. The goal is not artificial 100% precision; it is to reduce unexplained material movement and shorten time from detection to responsible action. Record false diagnoses as well as successful ones, because they reveal misleading signals and missing context.

Use a common language in commercial, advertising, finance, and supply meetings. When everyone describes decline through the same funnel and movement bridge, ownership becomes clearer and response collisions decrease. Diagnosis then improves the operating system, not only the individual product.