What price skimming is
Price skimming is launching a new item at a deliberately high price, taking the margin the earliest buyers will pay, then walking the price down in steps as the item ages and competition arrives. Penetration pricing is the mirror image: launch low, buy trial and doors, then raise the price once the item has velocity and the buyer has a reason to keep it. When I owned the new-item launch model at a natural-products brand, we picked between the two in a room with no data in it, and then lived with the consequence in the Sprouts POS file for the next two years.
Both are launch decisions, and both are usually described as marketing theory. In CPG they are not theory. They are a price series in your scan data, and they leave different fingerprints there.
The skim-then-decay trajectory
A skim launch does not hold one price. It holds a price until the early demand thins, then steps down, usually at the point where the brand needs either a second retailer or a promotional slot it cannot afford at the original price. Read week by week, it looks like a staircase, not a slope.
Take Copperfield Cold Brew, an illustrative 11 oz RTD launched into 380 Sprouts Farmers Market doors, and run the same SKU down both paths for a year.
| Period | Skim price | Skim units/store/wk | Penetration price | Pen. units/store/wk |
|---|---|---|---|---|
| Weeks 1-4 | $4.99 | 6.0 | $2.99 | 22.0 |
| Weeks 5-12 | $4.49 | 8.5 | $2.99 | 20.0 |
| Weeks 13-26 | $3.99 | 12.0 | $3.29 | 15.5 |
| Weeks 27-52 | $3.79 | 13.5 | $3.49 | 13.0 |
The two paths converge on almost the same place. By week 52 the skim SKU sells 13.5 units per store per week and the penetration SKU sells 13.0. What differs is everything that happened on the way there.
What each path is worth, per store, in year one
Hold the retailer at a 35% margin, so the brand's wholesale price is 65% of shelf, and put COGS at $1.60 a unit. Then the year reconciles like this, per store:
| Line | Skim | Penetration |
|---|---|---|
| Units sold | 611 | 803 |
| Retail dollars | $2,426 | $2,635 |
| Brand contribution | $597 | $428 |
Run the arithmetic on the skim column: 24 units at $4.99, 68 at $4.49, 168 at $3.99 and 351 at $3.79 is 611 units and $2,426 of retail. Wholesale on the first block is $3.24 against $1.60 of COGS, so contribution starts at $1.64 a unit and ends at $0.86. The penetration column never gets above $0.67 and opens at $0.34.
So penetration moves 31% more units and returns 28% less contribution. Across all 380 doors that is $226,860 for the skim against $162,640 for penetration, a $64,220 gap in the first year on one SKU. Penetration buys something real in exchange: 192 more units per store is a velocity number a buyer can see, and velocity is what earns the second retailer. The skim buys margin you can spend on the next launch.
Neither answer is right in the abstract. The question is whether your constraint this year is cash or distribution.
The baseline trap: never set a base off an intro-priced week
This is where launch pricing stops being a pricing question and becomes a measurement question, and it is the mistake I have watched cost the most money.
Take the penetration launch above. Weeks 1 to 12 run at $2.99 and average 20.7 units per store per week. That is the introductory price, and it is never coming back. The settled base at $3.29 is 15.5.
Now run a TPR in week 20 at $2.79 and read the lift:
| Baseline used | Base units | Promoted units | Incremental | Lift |
|---|---|---|---|---|
| Weeks 1-12 (intro-priced) | 20.7 | 26.0 | 5.3 | 26% |
| Weeks 13-19 (settled base) | 15.5 | 26.0 | 10.5 | 68% |
The same promotion, on the same units, reads as a 26% lift or a 68% lift depending on which weeks you called normal. Funded at $0.50 a unit across 380 stores for two weeks, the deal costs $9,880 and delivers 7,980 incremental units against the settled base, which is $1.24 per incremental unit. Measured against the intro-contaminated base it delivers 4,028, and the same deal now reads at $2.45. One of those numbers gets the promotion renewed and the other gets it killed.
The rule that survives contact: a launch has no baseline until the price has been stable for at least eight weeks at the price you intend to keep. Until then, every lift number is a comparison against a period that no longer exists. The same discipline is why incremental sales is a separate metric from total sales in the first place.
When each one is right
Skim first when the item is genuinely differentiated, the category has no close substitute at shelf, and your constraint is cash rather than doors. A skim also protects the price architecture: a SKU that launches at $4.99 can step to $3.99 and still read as a value, while a SKU that launches at $2.99 and needs $3.49 has to ask shoppers to accept an increase, which is a much harder conversation with a buyer.
Penetrate when the category is crowded, the shopper is comparing on shelf, and the item needs velocity fast to survive its first line review. Just budget for the step up, and take it in one move rather than three, because each price change resets the comparison the shopper is making. The step-up week is where that path lives or dies, and penetration pricing works through what it costs in units and in deal weeks afterwards.
The one path that reliably fails is a skim that never decays. A high launch price with no planned step down turns into a slow-velocity SKU with good margins and a shrinking store count, and the store count is what a buyer sees first. Retailers who run everyday low price are least tolerant of it, because there is no promotional week to hide the price gap in.
Where Scout fits
The work here is not the pricing decision, it is holding the price series and the unit series together well enough to know which weeks are a baseline. Scout connects your retailer and syndicated data and reads price, units and promoted weeks per SKU, so a launch curve shows up as a staircase with dates on it rather than a blended average. It measures the price you sold at and what it did to units. It does not set the retailer's shelf price, and it does not decide your launch strategy for you.
The short version
- Price skimming launches high and steps the price down; penetration pricing launches low and steps up. Both are visible in POS as a price staircase.
- On one illustrative SKU, penetration moved 31% more units in year one and returned $169 less contribution per store. Cash or distribution, pick one.
- Introductory weeks are not a baseline. Setting a base off intro-priced weeks turned a real 68% promo lift into a reported 26% and doubled the cost per incremental unit.
- A skim with no planned decay becomes a high-margin SKU with a falling store count, which is the number a buyer reads first.