Thirteen ecommerce pricing strategies, and how each one fails

An ecommerce pricing strategy is a repeatable method for turning known inputs into a price. Thirteen are in common use, and no catalog runs on one. A branded item five competitors also sell is priced from the market. A private-label item nobody else stocks is priced from value or cost. A seasonal item on its way out is priced from a markdown calendar. The question is not which strategy is best. It is which one fits this category, what data it needs, and how it fails when that data goes stale.

What an ecommerce pricing strategy actually decides

Every price is the output of three decisions, and most teams argue only about the third. The floor: where a unit stops making money once landed cost, fees, shipping subsidy and returns are counted. The reference: whose price the buyer compares yours against. The position: where you sit against it, and why.

Strategy is assigned per category, not per company. A store selling resold power tools and its own private-label workwear needs competitive pricing on one and value-based pricing on the other. One method across both leaves half the catalog priced by a blind rule.

Thirteen ecommerce pricing strategies, one at a time

Each entry gives the same five things: what it is, when it is right, the mechanic, how it fails, and a worked example. Short of time, read the failure line first.

1. Cost-plus pricing

Cost-plus adds a fixed markup to landed unit cost. It fits private label and long-tail SKUs where costs are stable and no competitive reference exists. It is a floor-setting method wearing a price tag.

• The mechanic. Price equals landed cost times one plus the markup. Markup is not margin: 50 percent on $60 landed gives $90, a 33.3 percent margin. For a true 40 percent margin, divide: 60 / 0.60 = $100.
• How it fails. Blind to demand and to the market. The common version is a landed cost that excludes returns, fees and shipping subsidy, so the markup sits on a number too small. The ecommerce margin calculator closes that gap.
• Example. A tool bag costs $60 landed. A 50 percent markup sets $90.00, while three competitors sell the identical bag between $74.99 and $79.99. The rule ran correctly; the item sells four a month instead of forty.

2. Keystone pricing

Keystone is cost-plus with the markup fixed at 100 percent: double the wholesale cost. It is fast, and defensible on low-velocity merchandise where careful pricing costs more than the margin at stake.

• The mechanic. Retail equals wholesale times two, a 50 percent gross margin. In a store that absorbs rent and staff. Online it absorbs shipping, fees and returns, none of which scale with price.
• How it fails. Weight and return rate break it. A bulky item eats the doubling in freight; a light, expensive item is left underpriced. Velocity never enters it, so fast movers sell cheap and slow movers sit dear.
• Example. A $24.00 wholesale item keystones to $48.00. Shipping costs $9.40, 3 percent payment fees take $1.44, and a 6 percent return rate at $13.40 per return costs $0.80 per order. Contribution is $12.36, or 25.8 percent, not 50.

3. Competitive or market-based pricing

Competitive pricing sets your price from observed competitor prices: match the lowest, beat it by a set amount, or hold an index against the market average. It is the honest choice when others list the identical SKU.

• The mechanic. Index equals your price divided by the market reference, times 100, where 100 is at market. Competitors at $179.99, $199.00 and $209.95 average $196.31, so $189.00 is an index of 96.3. Price index and price positioning covers the reference set.
• How it fails. Four mechanical ways: matching a listing that is not the same product, matching one out of stock, reading a stale price from a source that quietly stopped reporting, and walking downward against another repricer. Repricing strategies covers the guardrail each rule needs.
• Example. A retailer matches the lowest price on a $249.00 monitor. The lowest is a marketplace seller clearing open-box units at $198.00. On new stock costing $186.00, contribution falls from $63.00 to $12.00.

4. Value-based pricing

Value-based pricing starts from what the buyer's next best alternative costs them, not from what the item cost you. It fits differentiated products, private label with a real specification advantage, and anything with a service component.

• The mechanic. Take the alternative's reference price, add the difference in money the buyer already spends, then hand some back so switching is worth doing.
• How it fails. Nobody does the math on a product page. If the value is not stated in the buyer's own units, value-based pricing collapses into wishful cost-plus and the item looks expensive.
• Example. A competing shop vacuum is $149.00 with $29.00 of filters a year, so $294.00 over five years. A washable-filter model at $219.00 costs $219.00 to own: a 47 percent shelf premium and $75.00 cheaper. That sentence has to appear on the page.

5. Penetration pricing

Penetration pricing launches deliberately low to buy volume, reviews and rank, with a planned move to the target price. It fits new listings with no review history and consumables with a genuine repeat cycle.

• The mechanic. It is an investment, so write the payback down. Foregone contribution equals the launch-to-target gap times launch-window units, repaid by repeat purchases or cheaper traffic from rank.
• How it fails. The launch price becomes the reference price. Customers acquired there churn at the target price, competitors' repricers copy you and make it the market, and a price never raised is a thin margin with a story attached.
• Example. Target $39.00 on a $22.00 cost is $17.00 contribution; launching at $29.00 is $7.00. Eight hundred launch units forgo $8,000. If they produce 250 repeat customers buying three refills each at $17.00, that is $12,750 back.

6. Price skimming

Price skimming launches high, then steps down on a schedule, taking the least price-sensitive buyers first. It needs a genuinely new or scarce product, controlled distribution, and buyers separated by urgency rather than budget.

• The mechanic. A dated ladder where each step is triggered by demand decay, not by a competitor. The gain over one flat price sits entirely in the early units.
• How it fails. Early buyers feel punished when a step comes too fast, and say so in reviews. Unauthorized sellers arbitrage between steps. And if you enforce a minimum advertised price, the MAP must move with each step or your schedule creates violations.
• Example. A 1,200-unit run: 300 at $399.00, 400 at $369.00 and 500 at $329.00 returns $431,800. A flat $349.00 across all 1,200 returns $418,800. The ladder recovers $13,000 more, all of it in the first 300 units.

7. Psychological or charm pricing

Charm pricing chooses the ending digits, usually 9 or 99, to change how the number reads. It suits value-positioned and impulse price points. Round numbers do the opposite job on premium purchases.

• The mechanic. The left-digit effect. Moving $30.00 to $29.99 gives up 0.03 percent of the price and changes the digit the eye reads first. The cost is trivial, which is why it is worth testing.
• How it fails. On marketplaces sorted by price, a .99 ending sits above a competitor at .95, and a predictable ending is an easy undercut target. Charm endings also cheapen premium positioning.
• Example. Ten thousand sessions, $13.00 cost. At $30.00 and 2.40 percent conversion, 240 units earn $4,080. At $29.99 and 2.52 percent, 252 units earn $4,281. Two hundred dollars for one cent, if that conversion lift is real. That is the assumption to test.

8. Bundle pricing

Bundle pricing sells several items as one unit below the sum of the parts. It fits complementary products, slow movers riding fast movers, and listings no competitor sells in the same configuration.

• The mechanic. Compare blended bundle contribution against the components sold separately, then compute the extra volume needed. $46.00 against $69.00 means 69 / 46, or 1.5 times the units, to stand still.
• How it fails. If most buyers would have taken every component anyway, the discount is a transfer. Bundles also have no comparable competitor listing, so a category can drift out of position while its bundles look healthy. A bundle is not a way around MAP.
• Example. Drill $129.00 at $88.00 cost, bit set $34.00 at $17.00, case $19.00 at $8.00. Separately that is $69.00 of contribution; bundled at $159.00 it is $46.00. The bundle earns its place only on incremental units or dead stock.

9. Loss-leader pricing

A loss-leader is priced at or below cost to win the customer, with the money made on what follows. It needs an owned checkout, a predictable reorder, and a measured attach rate.

• The mechanic. Expected value equals the loss on the leader plus the attach rate times the attached item's contribution. Measure that rate on customers the leader acquired, not the store average.
• How it fails. Cherry-pickers buy the leader and nothing else, and resellers buy it in quantity, turning your acquisition spend into their inventory. Deep below-cost pricing also carries state-level legal considerations in the US. Educational, not legal advice.
• Example. A starter kit at $19.00 against $23.00 cost loses $4.00. If 62 percent buy a $34.00 refill carrying $23.00 contribution, expected value is -4.00 + 0.62 x 23.00 = $10.26. At a 30 percent attach rate it is $2.90, which no longer covers traffic.

10. Anchor and decoy pricing

Anchoring puts a higher-priced reference beside the option you intend to sell. A decoy makes that target obviously better value. Both require you to own the whole comparison set.

• The mechanic. Three options, with the intended seller in the middle or upper middle. The decoy is clearly worse on one dimension at nearly the same price.
• How it fails. It only works where the shopper does not open five tabs. On a marketplace the anchor is set by whoever is cheapest. Obvious decoys erode trust, and a top SKU that never sells still carries inventory cost.
• Example. Stockpots at 8-quart $89.00, 10-quart $119.00, 12-quart $129.00. The 10-quart is the decoy: two more quarts for $10.00 makes the 12-quart the easy pick. Remove it and $129.00 reads as $40.00 more than the entry model.

11. Tiered and volume pricing

Tiered pricing publishes a ladder where the per-unit price falls as quantity rises. It fits B2B accounts and consumables, where the customer's real alternative is buying a case somewhere else.

• The mechanic. Tie every breakpoint to a cost you actually save: a full case needing no repacking, a pallet that ships freight, one shipment instead of six. A ladder built on real savings can be defended to a customer and a reseller.
• How it fails. Breakpoints not tied to cost hand margin to buyers who would have purchased anyway. Ladders also leak: a case price dividing below your own retail feeds arbitrage sellers, and below MAP it is your own violation.
• Example. Single unit $12.50, six-pack $67.50 at $11.25 each, a case of 24 at $252.00 or $10.50 each, 96 units at $921.60 or $9.60 each, 23.2 percent off. If the real saving from 24 to 96 is $0.35 a unit, the $0.90 step gives away $0.55.

12. Subscription pricing

Subscription pricing charges on a schedule, usually below the one-time price, in exchange for predictable replenishment. It fits consumption cycles under about ninety days and low switching costs.

• The mechanic. The discount is bought back by lifetime, so compare contribution per customer rather than per order, and count subscription life in orders, not months. A skipped delivery extends the calendar without adding revenue.
• How it fails. The discount lands on customers who would have reordered anyway, which is margin leakage until incremental retention is measured. Subscriptions also trap you under a price customers watch, making cost inflation hard to pass through.
• Example. One-time $34.00 on $11.00 cost is $23.00 contribution, and the average one-time buyer reorders 1.6 times: $36.80. A 15 percent subscribe price of $28.90 gives $17.90 across 4.2 orders: $75.18. Worth roughly double, if 4.2 comes from your own cohorts.

13. Markdown and clearance laddering

A markdown ladder is a planned sequence of discounts that clears a set quantity by a set date. It applies to seasonal, end-of-life and overstock inventory. The strategy is not choosing a discount. It is choosing a sequence.

• The mechanic. Work backwards from the deadline and the unit count. At each step, ask what price clears the remaining units in the remaining weeks at the highest recovery. Earlier and shallower usually beats late and deep.
• How it fails. A ladder with no deadline is a slow price cut. A predictable one trains customers to wait. On a MAP-controlled brand each step needs the brand's agreement, or clearance becomes an enforcement problem.
• Example. Nine hundred units at $80.00, $46.00 cost, twelve weeks. Full price sells 120, then 20 percent off sells 260, 35 percent off sells 300, and 50 percent off clears 220: $50,640. A single 40 percent markdown on all 900 returns $43,200. Against $41,400 of cost, that is $9,240 of contribution against $1,800.

The decision table: which strategy fits which situation

How to choose a pricing strategy for a category

Run this in order and stop at the first answer that fits. Ten minutes per category, and it produces a decision you can defend.

1. 1

   Establish the floor first

   Compute contribution per unit at today's price, with landed cost, payment fees, shipping subsidy and the return rate included. Every strategy below operates above that line. Without it, no choice here is meaningful.

2. 2

   Ask whether the identical SKU is sold elsewhere

   If shoppers can find the same model number at three or more retailers, the market sets the reference and competitive pricing is your base method. If not, skip to step four. This question sorts most catalogs correctly.

3. 3

   Check the constraints that override your choice

   A minimum advertised price, a marketplace policy or a channel agreement can remove the decision entirely. Constraints are not a strategy, but they define the range one may operate in, and they are cheaper to find now.

4. 4

   For products only you sell, price from value or cost

   If you can name the buyer's alternative and quantify the difference in money they already spend, use value-based pricing and put that arithmetic on the product page. Otherwise use cost-plus at a target margin, as a hypothesis to test.

5. 5

   Handle lifecycle separately

   New with no reviews and a real repeat cycle points to penetration. New, scarce and distribution-controlled points to skimming. Seasonal or end-of-life with a clearance date points to a markdown ladder. Lifecycle overrides the steady-state method while it applies.

6. 6

   Layer presentation last, never first

   Charm endings, bundles, anchors, tiers and subscription discounts shape how a price is read or packaged. They are additions to a base price, not replacements. Choosing a price ending before you know your floor is the standard sequencing error.

How to tell whether a pricing change actually worked

The default measurement is a before-and-after comparison of the four weeks either side of the change, and it is wrong almost every time. Prices are rarely changed at random. They change because demand softened, stock piled up, a competitor moved, or a campaign launched. Each of those moves volume too, and a naive comparison credits all of it to price.

Set the success threshold before the change, not after. Break-even volume gives it directly: divide the old contribution per unit by the new one. A cut from $199.00 to $179.00 on a $130.00 cost takes contribution from $69.00 to $49.00, so 69 / 49 means 40.8 percent more units just to hold flat. Price elasticity, worked end to end has that arithmetic, and the dynamic pricing guide covers holdout design for prices that change continuously.

The data each strategy runs on

Look again at the third column of the decision table. Seven of the thirteen need nothing but your own cost and sales data. The other six need something about the outside world: what competitors charge, whether they are in stock, how that has moved. That is the real constraint on strategy choice, and the work of competitor price monitoring.

• Own data only. Cost-plus, keystone, bundle, anchor and decoy, subscription, tiered and volume, and markdown laddering.
• Needs outside data. Competitive pricing, value-based pricing against a named alternative, penetration and skimming ladders, loss-leaders exposed to reseller arbitrage, charm pricing on marketplaces.
• Needs it continuously. Any strategy running as an automated rule. A rule is only as current as its inputs.