A price increase can leave you with fewer orders and more contribution. It can also reduce demand enough to make the business worse. Unit sales alone cannot tell you which happened.
In Alfredo’s pricing breakdowns, the starting point is a product review: what does this item earn, how stable is demand, and what role does it play in the account? That review comes before a price recommendation. Use the process below to run a small, documented test with a clear decision at the end.
Start with the product’s job
Write down what you need from the product now. Are you protecting contribution, clearing excess stock, supporting a new customer offer, or trying to stabilize sales? A price decision needs to fit that goal.
For example, raising the price while the PPC team is spending more to recover unit volume may work against the plan. Lowering it while increasing bids can squeeze contribution from both directions. Neither combination is automatically wrong, but someone must approve the tradeoff and the amount the business can afford.
Check your product-level advertising allowance before the test. A price change can alter fees, conversion and the amount available to pay for clicks. The old ACOS target may need another look.
Choose a product with a readable baseline
Look for steady units, revenue and conversion, enough stock to support the test, and enough sales history to distinguish a change from normal variation. Avoid starting with a product whose performance is already deteriorating for an unknown reason.
Record the current offer price and actual average selling price. Discounts and the mix of orders can make those different. Review recent weeks alongside a longer period and the relevant seasonal context. A strong promotional week is a poor standalone baseline for a normal week.
| Signal | What to establish before testing |
|---|---|
| Price and costs | Realized revenue per unit and the variable costs included in contribution. |
| Demand | Units, traffic and conversion across comparable periods. |
| Inventory | Available stock and replenishment timing. |
| Account activity | Promotions, listing edits and planned PPC changes. |
If the product lacks useful history, label the uncertainty. You can still make a commercial decision, but it should not be presented as a confident reading of price elasticity.
Estimate the volume you can afford to lose
Price elasticity describes how demand responds to price. For an operating decision, translate it into a simpler question: how many units would you need at the new contribution per unit to match the old total?
Illustrative example, not an Enflet client result. Suppose a product leaves $8 per unit after the variable costs included in your model. At 100 units, that is $800. After a proposed price increase and recalculated costs, the estimate is $9 per unit. Matching $800 would require about 89 units: $800 ÷ $9 = 88.9.
This is a planning threshold, not a forecast. If advertising spend is excluded from the per-unit figure, subtract total spend separately from both periods. Do not subtract it twice. Changes in returns, fees or order mix also require an updated calculation.
Write one test with a review date
Select one SKU and a modest price move appropriate to its economics. Avoid changing a whole product family at once when you need to understand how each item responds.
- Record the original price, proposed price and reason.
- Name the person who approves and makes the change.
- Set a first review date and the maximum acceptable loss or deterioration.
- Note concurrent promotions, traffic changes and inventory events.
- Define what would justify keeping, reversing or extending the test.
Alfredo’s pricing-agent walkthrough uses a seven-day follow-up. Treat that as a checkpoint, not proof that every SKU produces a reliable result in seven days. Low-volume or seasonal products may need more evidence. Serious deterioration or an operational issue may justify stopping sooner.
Read contribution and demand together
At the review, compare units, realized selling price, conversion, revenue and total contribution. Check advertising spend and inventory movement too. A higher contribution per order does not rescue a test if the loss of orders leaves less total contribution.
Investigate what else changed. A competitor promotion, an offer problem or a shift in traffic can distort a before-and-after comparison. Log those events and avoid attributing the entire result to your price change. Read our attribution guide before treating attributed ad sales as proof of incremental demand.
Keep a successful price when the evidence supports the goal and the result remains acceptable across follow-up reviews. Revert when it breaches the limit you set. Extend only when the evidence is inconclusive and the cost of learning remains affordable.
Keep the recommendation separate from approval
Software can collect trends and suggest candidates. Require it to show the inputs, comparison dates and reason for the recommendation. Then have someone check the current PPC plan, stock position and product role before acting.
When a core price increase cannot work, review whether a multipack, bundle or different offer could improve order economics. Those options introduce their own costs and demand assumptions; they need a separate evaluation. The useful habit is a repeatable pricing review with recorded decisions, rather than a blanket instruction to raise every price.
From Enflet’s video library
These guides adapt the ideas from the original videos into a practical reading format.
