Penetration pricing in ecommerce: does it still work?

Penetration pricing was designed for a world where getting noticed was hard and matching a price was slow. A low launch price bought attention that was otherwise unaffordable, and incumbents took weeks to respond because changing a price meant reprinting something.

Ecommerce inverted both conditions. So the fair question is whether the strategy survived the move online.

Mostly yes - but the assumptions underneath it changed enough that running the classic playbook unmodified is how you lose money.

What broke

Response time collapsed. Your competitor's repricer sees your launch price within hours and matches it automatically. The window where undercutting bought you anything used to be months. On a marketplace it can be an afternoon.

Price comparison is free. The strategy assumed shoppers noticed your low price and stayed loyal once they had tried you. Now the same shopper who found you on price finds the next entrant on price too. You rented the customer.

Acquisition costs eat the margin twice. Classic penetration pricing traded margin for cheap organic attention. Online, you are frequently paying for the traffic and discounting the product. That is two subsidies stacked on one transaction, and the payback assumptions rarely survive contact with real CAC.

Price memory is permanent. Screenshots, price-tracker extensions, cached listings. The €19 launch price does not fade from view the way an old shelf tag did, which makes the climb to €26 much harder than the theory allows.

What still works

Marketplace launch velocity. On Amazon-style platforms, early sales volume drives ranking, and ranking drives everything after. Buying that initial velocity with price is a real mechanic, not a hope - and it is time-boxed, which is exactly what penetration pricing needs.

Review accumulation. A new listing with no reviews converts badly whatever the price. Discounting to build a review base is legitimate and self-limiting.

Subscription and replenishment models. Where switching cost genuinely accrues, the original logic holds intact. First-order discounting works because the second order is easier than the first.

Categories with real scale economics. If your unit cost genuinely falls with volume, the bet still pays.

Where it does not work: undifferentiated products in transparent categories against well-capitalised incumbents. There you are not penetrating a market, you are funding a price war you will lose.

Running it properly online

Time-box it explicitly. A launch price with no end date is not penetration pricing, it is your new price. Decide the exit before the entry - the exit problem kills more launches than the launch price does.

Hold position, not a number. Competitors move. A fixed €19 becomes meaningless the moment the market shifts around it. Express the intent instead - a fixed distance below the category leader, bounded by a floor - and let dynamic pricing track it.

Set a hard margin floor. Non-negotiable. Penetration pricing without one is how a launch plan becomes a write-off, and it needs to be enforced as a rule that no recommendation can override, not as a note in a plan.

Watch the response. Live competitor monitoring tells you within hours whether incumbents are matching. If they are, the strategy has already failed and the correct move is to stop early rather than to commit harder.

Climb in small steps. Repeated 3-5% increases get absorbed. One 35% jump gets noticed.

Penetration pricing still works in ecommerce. It just needs a floor, a clock, and a way to see what the market does back.