Level 3 · Recommend: where to push first?

From numbers to a recommendation

Angela has ten minutes, so the paragraph has to stand on its own. The structure that works is what, how much, so what:

• What is going on? The return rate is steady, about one line in thirty, so the queue grew because sales grew. Two categories stand out, for different reasons.
• How much supports it? 3.3% of lines overall; Textiles 4.7%, mostly changed minds; Furniture 4.4% and the largest refunds in dollars, mostly damage; Lighting close behind on damage.
• So what should happen? Two fixes, each tied to a reason: packaging for the breakable, expensive items, better product information for Textiles.

Without the numbers it is an opinion. Without the fixes Angela has nothing to take to the review.

Answer the worry first

Angela’s request started from a feeling: returns feel high. The first sentence of the paragraph should meet that feeling with a number. Here the rate is 3.6% in 2023, 3.1% in 2024 and 3.2% in 2025, so the honest opening is that returns are not rising as a share of sales. That is useful news in its own right: the extra work is a staffing question as much as a product one.

Caveats belong in the paragraph

• Reasons are self-reported. The reason is what the customer picked on the return form. A customer who received a scratched table may tick “changed my mind” to avoid an argument. The codes point to the likely cause; a sample of inspected returns would confirm it.
• Refunded orders are separate. 777 returned lines come from fully refunded orders. They are a different process (often a whole order sent back before or on delivery) and deserve their own look.
• A fix needs a measurement plan. Comparing the months after a packaging change with the months before mixes in the season. Compare the same months a year earlier too, or roll the change out to some products first.

Caveats are not hedging. They tell the reader how far the result can be stretched.

The traps in the other options

• Counting instead of rating. Decor and Textiles return the most items partly because they sell the most. The rate is what tells you whether something is wrong.
• The wrong denominator. Dividing every return by every line mixes in refunded orders, which return all their lines by design, and inflates the rate to 5.5%.
• Cause from a code. “Changed mind” does not say why. Promising that new photos will bring Textiles down to 2.0% is a guess presented as a forecast.

The report structure behind this paragraph

In a written report this paragraph is the executive summary. Below it come the objective (the SMART question from the first level), the data and its limits (the status check, the excluded lines, the self-reported reasons), the method, the findings with the chart and the reason table, the conclusions, and the recommendations with an owner and a date for each. Most readers stop after the summary, which is why it carries the numbers and the caveats.