Chapter 9: Pricing

Marcus Lee has two spreadsheets open and a problem with both of them. The first tracks the Lumen Commute against a growing list of online e-bike brands selling for under $1,000, about half of the Commute's $1,899 price. Sales of the Commute have slowed, and two dealers have asked whether Lumen plans to cut the price. The second spreadsheet is the forecast for taking the $2,999 Lumen Cargo national. Dana Okafor, the VP of Marketing, wants to know whether $2,999 is the right price for families in cities where nobody has heard of Lumen, and whether the company should try the dynamic pricing tool a software vendor has been pitching for the website.

None of these questions has an answer that falls out of a formula. Price is the only part of the marketing mix that brings money in; product, place and promotion all cost money. It's also the part customers notice fastest and competitors can copy fastest. A company can spend a year designing a better bike and lose much of the benefit with one careless price decision.

This chapter covers how companies set and change prices: the three main approaches, the back-of-the-envelope arithmetic every marketer should know, the factors that push prices up or down, and the strategies companies use for new products, product lines and price adjustments. Then it turns to algorithmic and dynamic pricing, which AI has made common, and the questions of fairness and law that come with it.

9.1 Approaches to pricing

A price can't sensibly be lower than what it costs to make and sell the product, at least not for long. It also can't be higher than what customers believe the product is worth to them, or they won't buy. Between those two limits sits the range of prices a company could charge, and the prices of competitors and substitutes help decide where in that range a price lands.

Limit	What sets it	What happens if you cross it
Price ceiling	Customers' perceptions of the product's value	Customers don't buy, or buy from someone else
Reference points in between	Competitors' prices, substitutes, the company's own strategy	Customers compare and switch
Price floor	The company's costs	The company loses money on each sale or overall

The three main approaches to pricing each start from one of these points. In practice good pricing uses all three, but the starting point shapes the result.

9.1.1 Customer value-based pricing

In Chapter 1. Marketing in an AI Era you saw that marketing is about creating value for customers and capturing value in return. Price is where that capture happens. Customer value-based pricing starts with the customer rather than the factory.

The logic runs in the opposite direction from the usual way people think about prices. In cost-based thinking, a company designs a product, adds up what it costs, adds a profit and then tries to convince customers the price is fair. In value-based thinking, the company first asks what customers value and what they'd be willing to pay for it, sets a target price, and then designs a product and a cost structure that can deliver that value at that price.

The hard part is measuring value, because asking customers directly what they'd pay tends to produce unreliable answers. Marketers use methods you met in Chapter 6. Marketing Research and Analytics: choice-based surveys such as conjoint analysis, price tests, and studies of what customers currently spend to solve the same problem. For a family considering a Lumen Cargo, the relevant comparison might be the cost of a second car: purchase, insurance, fuel and parking. Measured against that, a $2,999 cargo bike can look inexpensive, even though it's expensive compared with an ordinary bicycle.

Two forms of value-based pricing are common.

IKEA keeps prices low by having customers transport and assemble their own furniture; Costco offers a narrow selection in bulk and charges a membership fee. Retailers that practice everyday low pricing (EDLP), such as Walmart, are a form of good-value pricing: they charge a consistently low price with few temporary discounts. The contrasting approach is high-low pricing, in which a retailer charges higher everyday prices but runs frequent sales on selected items.

Value-added pricing is the main defense against low-price rivals. Apple rarely competes on price; it adds design, integrated devices and services, and stores where customers get help, and charges more for the package. For Lumen, it would mean answering an $899 rival not by cutting the Commute's price but by showing what the extra money buys: a certified battery, dealer test rides and service, the app, Lumen Care. Whether that works depends on whether customers actually value those things, which is a research question.

9.1.2 Cost-based pricing

Cost-based pricing starts with the seller's costs. It's the most common approach among small businesses and resellers because it's simple and feels safe.

To use it, you need to know the two kinds of cost every business has.

Total cost is fixed costs plus variable costs at a given level of production. A company has to cover both, but in different ways: each sale has to cover its own variable cost, and the sales as a group have to cover the fixed costs.

9.1.2.1 Markup pricing

The simplest form of cost-based pricing is to add a standard markup to the cost of a product.

Markups can be stated two ways, and mixing them up is one of the most common errors in business arithmetic.

• Markup on cost = (price − cost) ÷ cost
• Markup on selling price = (price − cost) ÷ price

Retailers and dealers usually talk about markup on selling price, which is the same as their gross margin. Manufacturers often talk about markup on cost. Always ask which one someone means.

Worked example: a cost-plus price for the Fold. Suppose Lumen's full cost to make and deliver one Lumen Fold, including a share of fixed costs at the expected sales volume, is $700. (All Lumen cost figures in this chapter are illustrative.) If Lumen wanted a 50% markup on selling price, the price would be:

Price = cost ÷ (1 − markup on selling price) = $700 ÷ (1 − 0.50) = $700 ÷ 0.50 = $1,400

Lumen actually charges $1,399, a dollar less, for reasons you'll see under psychological pricing. At $1,399 the markup on selling price is ($1,399 − $700) ÷ $1,399 = $699 ÷ $1,399 = 0.4996, or about 50%. The markup on cost is $699 ÷ $700 = 0.9986, or almost 100%. Same price, same cost, two very different-sounding percentages.

Worked example: the dealer's markup. Lumen's independent dealers buy bikes from Lumen at a discount from the retail price and resell them at that retail price. Suppose dealers earn a 35% margin on the retail price of a bike. For a Lumen Cargo at $2,999, the dealer's margin is $2,999 × 0.35 = $1,049.65, so the dealer pays Lumen about $1,949 (rounded to the nearest dollar). The table below shows the same calculation for all three models.

Illustrative dealer pricing for Lumen bikes (amounts rounded to the nearest dollar)

Model	Retail price	Dealer margin (35% of retail)	Price the dealer pays Lumen	Dealer's markup on cost
Lumen Fold	$1,399	$490	$909	$490 ÷ $909 = 53.9%
Lumen Commute	$1,899	$665	$1,234	$665 ÷ $1,234 = 53.9%
Lumen Cargo	$2,999	$1,050	$1,949	$1,050 ÷ $1,949 = 53.9%

The dealer's 35% markup on selling price is the same as a 53.9% markup on cost. You can convert directly: markup on cost = markup on price ÷ (1 − markup on price), so 0.35 ÷ 0.65 = 0.538.

Markup pricing stays popular because sellers know their costs better than they know demand, and because similar markups across an industry keep price competition limited. Its weakness is serious, though: it ignores what customers will pay and what competitors charge. A cost-plus price can leave money on the table when customers value the product highly and price it out of the market when they don't. It's also circular: the "full cost" per unit includes a share of fixed costs, which depends on how many units you sell, which depends on the price.

9.1.2.2 Contribution margin and break-even analysis

A more useful cost-based tool asks how many units a company must sell at a given price to cover its costs.

Worked example: breaking even on the national Cargo launch. Here are illustrative numbers for a Lumen Cargo sold through the website.

Variable cost per Cargo sold online	Amount
Manufacturing, including battery and motor	$1,480
Shipping to the customer	$95
Payment processing (about 1.5% of price)	$45
Warranty reserve	$29
Total variable cost per unit	$1,649

The contribution margin per unit is:

Contribution margin = price − variable cost = $2,999 − $1,649 = $1,350

The contribution margin ratio is $1,350 ÷ $2,999 = 0.450, or about 45%. Out of every dollar of Cargo revenue, about 45 cents is left to cover fixed costs and profit.

Now suppose taking the Cargo national adds $1,350,000 of fixed costs in the first year: a launch campaign, extra customer-service staff and a demo-ride program in new cities. The break-even volume is:

Break-even units = fixed costs ÷ contribution margin per unit = $1,350,000 ÷ $1,350 = 1,000 Cargos

In dollars, that's 1,000 × $2,999 = $2,999,000 of revenue. You get the same answer by dividing fixed costs by the contribution margin ratio: $1,350,000 ÷ 0.45015 ≈ $2,999,000.

Break-even analysis is most useful when you compare prices. The table below shows the break-even volume for the national launch at three possible prices, keeping costs the same.

Illustrative break-even volumes for the national Cargo launch

Price	Contribution margin per unit	Break-even units ($1,350,000 ÷ margin)
$2,799	$2,799 − $1,649 = $1,150	1,173.9, so 1,174
$2,999	$2,999 − $1,649 = $1,350	1,000
$3,199	$3,199 − $1,649 = $1,550	870.97, so 871

Round break-even units up, because selling 1,173 units at $2,799 leaves you a little short ($1,150 × 1,173 = $1,348,950).

The table doesn't tell Lumen which price to pick. A $200 price cut raises the break-even volume by 174 units, about 17%. Whether that's a good trade depends on whether the lower price would bring in more than 17% more buyers, which is a demand question. Break-even analysis tells you what has to be true for a price to work. Market research tells you whether it's likely to be true.

You can extend the same logic to a profit target, an approach called target return pricing. If Lumen wanted the national launch to earn $500,000 above its fixed costs in the first year at $2,999, it would need to sell ($1,350,000 + $500,000) ÷ $1,350 = 1,370.4, so 1,371 Cargos.

9.1.3 Competition-based pricing

Competition-based pricing sets prices mainly in relation to competitors' prices and offers.

Where customers can see every price at a glance, as with gas stations on the same corner, competition-based pricing is almost unavoidable. Elsewhere, competitors' prices are one input among several, and the real question isn't "what do they charge?" but "how does our value compare with theirs?" If Lumen's offer is better in ways customers care about, it can price above competitors. If not, it has to price at or below them, or improve the offer.

List the substitutes as well as the direct competitors. A Seattle family choosing transportation for school drop-offs is comparing the Cargo not just with other cargo e-bikes but with a second car, ride-hailing, a cheaper e-bike plus a trailer, and a used cargo bike. Each sets a reference point.

9.1.4 Putting the approaches together

The three approaches aren't rivals so much as three questions you ask in order:

1. What is it worth to the customer? (value-based: sets the ceiling)
2. What do the alternatives cost, and how does our offer compare? (competition-based: places the price within the range)
3. Can we make money at that price, and at what volume? (cost-based: checks the floor)

9.2 Factors in pricing decisions

Customer value, costs and competition set the range. Many other factors, some inside the company and some outside it, push the price around within that range.

9.2.1 Internal factors

9.2.1.1 Overall strategy and objectives

Price should follow from the decisions you made in Chapter 7. Segmentation, Targeting and Positioning. A company positioned as the premium choice can't run a permanent discount without undermining its position. A company positioned on low price has to keep its costs lower than its rivals' or it will run out of margin.

Pricing objectives vary with the company's situation:

Common pricing objectives

Pricing objective	What it emphasizes	Example situation
Survival	Covering variable costs and some fixed costs to stay in business	Excess inventory, a price war or a sudden drop in demand
Profit maximization	The price that produces the highest current profit	A product with steady demand and few close substitutes
Market share	Low prices to win customers and volume	Entering a new market where scale lowers costs
Product-quality leadership	High prices that signal and pay for quality	A premium brand
Customer retention	Prices that keep existing customers loyal	Subscriptions and service plans
Partner support	Prices that leave enough margin for dealers and retailers	Products that need help from resellers to sell

A company often pursues more than one objective at once, and they can conflict. For Lumen, building share for the Cargo in new cities pushes toward a lower price, while protecting its quality position and its dealers' margins pushes the other way.

9.2.1.2 The rest of the marketing mix

Price can't be set apart from product, place and promotion. A product sold through dealers has to carry the dealer's margin in its price; a heavily advertised product has those costs to recover. Some companies set a target price first and design the product to hit it, an approach called target costing.

9.2.1.3 Costs

Costs set the floor, and they change with scale. As a company produces more, it gets better at production and buys in larger quantities, so unit cost falls. This is the experience curve. A company expecting costs to fall may price below today's cost to build volume faster, a bet that fails if the volume doesn't come.

9.2.2 External factors

9.2.2.1 The market and demand

Demand sets the ceiling, and market structure decides how much pricing freedom a seller has. Sellers of commodities such as wheat have almost none. In most consumer markets, including e-bikes, sellers differentiate their offers and so have some room to price differently. Where a few large sellers dominate, as with airlines, each watches and reacts to the others' prices closely.

The key measure of how demand responds to price is elasticity.

Worked example: elasticity. Suppose a company raises a price by 10% and the quantity sold falls by 5%. Price elasticity is −5% ÷ 10% = −0.5. Demand is inelastic: buyers are not very sensitive to price, and the increase raises revenue. If instead quantity fell by 20%, elasticity would be −20% ÷ 10% = −2.0. Demand is elastic, and the increase lowers revenue.

Demand tends to be less elastic when the product is unique or high in perceived quality, when substitutes are hard to find or compare, when the cost is small relative to income, and when someone else pays (an employer, an insurer). It tends to be more elastic when close substitutes are easy to find and compare, which is exactly what online shopping and AI shopping assistants make easier.

Worked example: what a price cut has to achieve. Elasticity helps explain why price cuts are riskier than they look. Suppose the Lumen Commute costs Lumen $1,050 per unit in variable costs when sold online (illustrative). At $1,899, each sale contributes $1,899 − $1,050 = $849. Dealers have suggested cutting the price to $1,699 to compete with low-cost brands. At $1,699, each sale contributes $1,699 − $1,050 = $649.

To earn the same total contribution, Lumen would have to sell $849 ÷ $649 = 1.308 times as many bikes, about 31% more. For every 100 Commutes it sells today (contributing 100 × $849 = $84,900), it would need to sell 131 at the lower price (131 × $649 = $85,019) just to stand still. A price cut of about 10.5% ($200 ÷ $1,899) that needs 31% more volume implies an elasticity of roughly −3, very price-sensitive demand. And at $1,699 the Commute would still cost nearly twice what an $899 rival does.

9.2.2.2 Customers' perceptions of price

Customers don't evaluate prices the way an accountant does. They compare each price with a reference price, and they often use price as a signal of quality when they can't judge quality directly, as with wine, a medical specialist or an e-bike battery whose safety they can't inspect. A price that's too low for the category can raise suspicion rather than demand.

9.2.2.3 Competitors

Every price change invites a response. Ask how competitors will react and whether they can afford to match a cut. A cut that competitors match immediately leaves the whole industry with lower margins and the same market shares, the classic outcome of a price war.

9.2.2.4 The economy, resellers, government and society

Economic conditions affect what customers can pay and how they judge value. In a downturn, buyers trade down to cheaper options or put off big purchases, which matters for a product like a $3,000 cargo bike. Resellers (dealers and retailers) need a margin worth their effort, which you'll see in Chapter 10. Channels, Retail and the Customer Journey. Governments regulate pricing practices, as you'll see later in this chapter, and tariffs on imported components can change costs quickly. Society's expectations about fairness also matter. Customers who feel a price is unfair, even if it's legal, can punish a brand.

9.3 Pricing strategies

With an approach chosen and the factors weighed, a company still has to set specific prices and decide how they change over time and across customers. This section covers three groups of strategies.

9.3.1 New-product pricing strategies

A new product's price sets the customer's reference price for everything that follows. Two broad strategies are available.

Game console and smartphone makers have often launched new models at high prices and reduced them, or added cheaper versions, as the product aged.

Penetration pricing is common for subscription services that need a large base of users and for products with network effects, where each new user makes the product more valuable to others. The risk is that customers anchor on the low price and resist increases later, as many streaming subscribers did when services raised prices.

Skimming compared with penetration

	Market skimming	Market penetration
Starting price	High	Low
Goal	Revenue per unit from less price-sensitive buyers	Volume and market share
Works best when	Quality and image justify the price; competitors can't enter quickly	Demand is price sensitive; costs fall with volume
Main risk	Invites competitors; may seem greedy	Hard to raise prices later; may signal low quality

Lumen's national expansion isn't a new product so much as a new market for an existing one. That still raises a launch-pricing question: should Lumen use an introductory offer in new cities, and if so, how does it avoid upsetting customers in Denver who paid full price? You'll come back to this in the Lumen section.

9.3.2 Product mix pricing strategies

When a product is part of a line or a larger mix, the company sets prices to maximize the profit of the whole mix, not of each product alone.

Lumen's three bikes, at $1,399, $1,899 and $2,999, form a product line. The gaps between them should reflect differences customers can see and care about. If the gap between the Commute and the Cargo is $1,100, customers should be able to explain what the extra $1,100 buys: a longer frame, more carrying capacity, a stronger motor, seating for children. If the gap seems arbitrary, customers either trade down or don't buy.

Several other product-mix strategies are worth knowing:

Product mix pricing strategies

Strategy	What it means	Example
Optional-product pricing	Pricing optional or accessory products sold with the main product	Car options packages; Lumen's locks, racks and child seats
Captive-product pricing	Pricing products that must be used with the main product	Printer ink, razor blades, game cartridges
Two-part pricing	A fixed fee plus a variable usage rate, the service version of captive pricing	A phone plan with a monthly fee plus data charges; an amusement park entry fee plus ride charges
By-product pricing	Setting a price for by-products to help offset the cost of the main product	A sawmill selling sawdust; a zoo selling animal manure as compost
Product bundle pricing	Combining several products and offering the bundle at a reduced price	Fast-food value meals, software suites, cable and internet packages

Captive-product pricing shows how the mix, not the single product, earns the profit: a printer maker can sell printers near cost and earn its margin on ink. It backfires when customers feel trapped, and ink prices have long drawn complaints. Lumen has something like a two-part structure: the bike, then an optional $149-a-year Lumen Care plan. A buyer who keeps Lumen Care for three years pays an extra $447, recurring revenue that may matter more to Lumen's long-term profit than a small difference in the bike's price.

Bundles work when customers value the combined package and the bundle price is clearly lower than the sum of the parts. Suppose Lumen sold child seats for $179 (illustrative). A "family starter" bundle of a Cargo, a child seat and the first year of Lumen Care would be worth $2,999 + $179 + $149 = $3,327 if bought separately. Offered at $3,199, the bundle saves the customer $128 and, if the seat and service plan have healthy margins, may earn Lumen more total contribution than a Cargo sold alone. It also gets more customers onto Lumen Care, which builds the long-term relationship.

9.3.3 Price adjustment strategies

Companies adjust their basic prices for different customers and situations. The table below summarizes the main adjustment strategies.

Price adjustment strategies

Strategy	Description
Discount and allowance pricing	Reducing prices to reward customer responses such as paying early, buying in volume or promoting the product
Segmented pricing	Adjusting prices for differences in customers, products or locations
Psychological pricing	Adjusting prices for psychological effects
Promotional pricing	Temporarily reducing prices to increase short-run sales
Geographical pricing	Adjusting prices for customers' geographic location
Dynamic and personalized pricing	Adjusting prices continually to meet the characteristics and needs of situations and individual customers
International pricing	Adjusting prices for international markets

9.3.3.1 Discounts and allowances

Discounts reduce price in return for something the seller wants: a cash discount for paying quickly ("2/10, net 30" means a 2% discount if the bill is paid within 10 days, with the full amount due in 30), a quantity discount for buying in volume, a functional (trade) discount for resellers who perform channel functions such as selling and service, and a seasonal discount for buying out of season. Lumen's dealer margin is a functional discount: dealers get a lower price because they provide test rides, assembly and service. Its fleet program could offer quantity discounts to universities and delivery businesses that buy many bikes at once, a topic you met in Chapter 5. Business Markets and Buyer Behavior.

Allowances are payments or price reductions for specific actions, such as a trade-in allowance for turning in an old product or a promotional allowance paid to dealers who advertise or display the product.

9.3.3.2 Segmented pricing

Movie theaters charge less for students and seniors, software is sold at an education discount, and airlines charge different prices for the same seat depending on when it's booked. Segmented pricing works when segments differ in demand, when low-price buyers can't easily resell to high-price buyers, and when the practice doesn't create resentment or break the law. It is the ancestor of the personalized pricing you'll meet in the next section.

9.3.3.3 Psychological pricing

Prices ending in 9, such as $1,399 instead of $1,400, are the most familiar form. Many shoppers process prices from left to right and weight the first digit heavily, so $1,399 can feel closer to $1,300 than to $1,400. Prices ending in round numbers can do the opposite job: luxury brands often use them to signal quality rather than value. Psychological pricing also uses reference prices: showing an original price next to a sale price, placing a product beside a more expensive one, or offering a high-priced "decoy" option that makes the middle option look reasonable.

9.3.3.4 Promotional pricing

Promotional pricing includes temporary sale prices, special-event pricing, cash rebates, low-interest financing and free extras. These tools create urgency and clear inventory. Used too often, they teach customers to wait for the next sale and erode the brand. Habits, once formed, are hard to change: when JCPenney replaced its frequent sales and coupons with simpler everyday prices in 2012, deal-seeking shoppers stayed away, and the company reversed course. You'll see promotions as a communications tool in Chapter 11. Integrated Marketing Communications.

9.3.3.5 Geographical and international pricing

When shipping costs differ by location, a company can charge each customer the actual shipping cost, charge everyone the same delivered price, set prices by zone, or absorb shipping to win distant markets. "Free shipping" builds the cost into the price or a membership fee. Large bikes are expensive to ship, so Lumen's national expansion raises this question directly. Across countries, costs, tariffs, exchange rates and incomes differ, so prices usually differ too, and large gaps can create gray markets in which goods bought cheaply in one country are resold in another.

9.3.4 Initiating and responding to price changes

Price changes send signals. A cut can suggest that a product is about to be replaced or has quality problems; an increase can suggest strong demand, or greed. When costs rise, companies can raise prices openly or recover the cost less visibly by removing features or shrinking the package, a practice critics call shrinkflation. Customers usually notice, and hidden changes damage trust.

When a competitor cuts its price, a company can hold its price and emphasize value, improve quality and raise its price, match the cut, or launch a lower-priced fighter brand to take on the rival without cutting the main brand's price. Each has risks. A fighter brand, for example, can steal sales from the main brand instead of from the competitor.

9.4 Algorithmic and dynamic pricing

Airlines and hotels have adjusted prices continually for decades, using revenue management (also called yield management) systems that change fares based on how many seats are left and how close the departure date is. What AI has changed is scale, speed and reach. Software can now watch competitors' prices, inventory, demand signals and individual shoppers' behavior, and change prices on thousands of products many times a day, as large online retailers, ride-hailing apps and ticket sellers do.

It helps to separate three things that are often lumped together:

Three forms of algorithmic pricing

Type	What changes the price	Example
Time- or demand-based dynamic pricing	Demand, time, remaining capacity	Airline fares, hotel rooms, surge pricing for rides
Competitor-responsive pricing	Competitors' prices	Online retailers that match or undercut rivals automatically
Personalized pricing	Data about the individual shopper	Different shoppers seeing different prices or discounts for the same product at the same time

9.4.1 What AI does well in pricing

Used carefully, AI-driven pricing has real benefits:

• Matching supply and demand. When capacity is fixed and perishable, such as an airline seat or a concert ticket, prices that rise and fall with demand can fill more capacity and earn more revenue. Peak prices can also draw more supply, as when ride-hailing drivers go out during surges.
• Speed and scale. No team can review 50,000 prices a day. Software can spot a competitor's price change or a jump in demand in minutes.
• Better estimates of demand. Machine-learning models can estimate how demand responds to price, season, weather and promotions more precisely than rules of thumb, especially when they learn from many small price experiments.
• Markdown optimization. For seasonal goods, AI can recommend when and how much to discount to clear inventory with the least loss of margin.

9.4.2 Where AI pricing fails or causes harm

9.4.2.1 Fairness and customer trust

Customers accept some price variation, such as higher hotel prices on a holiday weekend, because they understand the reason and it applies to everyone. They react very differently when they feel exploited. When Wendy's mentioned in 2024 that it would test "dynamic pricing" on digital menu boards, many customers assumed it meant surge pricing on burgers at lunchtime. The backlash was quick, and the company said it had no plans to raise prices at peak times. Sharp jumps in demand-priced concert tickets have drawn similar anger.

Personalized pricing raises sharper concerns. An algorithm that learns which shoppers rarely compare prices, or which neighborhoods have fewer alternatives, can charge those people more. Even if no one intends it, data correlated with income, age, race or disability can produce prices that fall hardest on vulnerable groups. Regulators have started to study what they sometimes call "surveillance pricing," and at least one U.S. state, New York, has passed a law requiring companies to disclose when a price was set by an algorithm using a customer's personal data. These rules are changing fast, so check the current law wherever you operate. You'll look at privacy and data more broadly in Chapter 13. Ethics, Responsible AI and Sustainability.

9.4.2.2 Collusion

The clearest algorithmic cases involve people using software to carry out an agreement they made. In 2015 the U.S. Department of Justice brought a criminal case against online poster sellers on Amazon Marketplace who had agreed to fix prices and used pricing algorithms to do it. Harder questions arise when competitors never talk to each other but use the same pricing software, which pools their data and recommends prices to each. In 2024 the Department of Justice sued RealPage, which sells rent-pricing software to landlords, alleging that the software let competing landlords share confidential information and align rents. Cases like this test how old competition law applies when the "agreement" may run through a shared algorithm. Economists also debate whether independent algorithms, each maximizing its own company's profit, could learn to keep prices high with no human agreement at all.

9.4.2.3 Other legal limits on pricing

Algorithmic or not, pricing must stay within the law. The main U.S. limits an introductory marketer should know are:

Legal limits on pricing (U.S.)

Practice	What it is	Why it's a problem
Price fixing	Competitors agreeing on prices	Illegal under the Sherman Act
Predatory pricing	Selling below cost to drive competitors out, then raising prices	Can violate antitrust law, though it is hard to prove
Price discrimination between business buyers	Charging competing business customers different prices for the same goods when it harms competition	Restricted by the Robinson-Patman Act, with defenses such as cost differences or meeting a competitor's price
Deceptive pricing	Misleading reference prices, hidden fees, bait-and-switch offers	Enforced by the Federal Trade Commission and state laws
Resale price maintenance	A manufacturer requiring resellers to charge a set price	Legal treatment varies; many manufacturers use advertised-price policies instead

Personalized pricing to consumers is not generally illegal in the United States as long as it doesn't discriminate on a protected characteristic, deceive customers or break specific state rules. That doesn't make it a good idea, as you'll see.

9.4.2.4 Errors at machine speed

Algorithms can also simply go wrong. In a well-known 2011 case, two sellers' repricing programs on Amazon kept pricing off each other until a biology textbook was listed for more than $23 million. That example is harmless. A bug that prices a popular product far below cost, or an algorithm that raises prices on essentials during an emergency, is not. Many U.S. states have price-gouging laws for declared emergencies, and an algorithm doesn't know a hurricane has been declared unless someone tells it.

9.4.3 What stays a human judgment

AI can estimate demand, recommend prices and change them faster than any person. It can't decide what a brand stands for, what customers will feel is fair or how much legal and reputational risk a company should accept. A sensible approach to algorithmic pricing includes:

• Guardrails. Floors and ceilings on every price, limits on how fast and how far prices can move, and rules that freeze prices on essentials during emergencies.
• Clear rules about inputs. Decide which data the algorithm may use. Many companies choose not to let personal characteristics, or close proxies for them, affect the price an individual sees.
• No shared data with competitors. Know where a pricing vendor gets its data and whether competitors' confidential data is involved.
• Transparency. Be ready to explain to a customer, in one sentence, why the price is what it is.
• Human review. Someone monitors outcomes, reviews unusual prices and owns the result.

9.5 Lumen Cycles: pricing the Cargo for a national launch

Back in Denver, Dana and Marcus sort the pricing questions into three decisions.

The Cargo's national price. At $2,999 the national launch needs 1,000 online Cargo sales to cover its $1,350,000 of added fixed costs; at $2,799 it would need 1,174. Dealer sales change the math. When a dealer sells a Cargo, Lumen receives about $1,949 and its own variable cost is about $1,569 (manufacturing, warranty reserve and $60 of freight to the dealer, but no customer shipping or card fees), so Lumen keeps about $380 per bike instead of $1,350. The more Cargo sales go through dealers, the more bikes the launch must sell to break even, a trade-off you'll examine in Chapter 10. Channels, Retail and the Customer Journey.

The case for keeping $2,999 is mostly value-based: for the target families, the comparison is a second car, not another bicycle, and a lower price could make the certified battery and build quality look less credible. The case for an introductory discount is that Lumen is unknown in its new cities and needs early riders to generate reviews. Dana leans toward keeping $2,999 everywhere and offering a launch bundle instead: the "family starter" bundle of Cargo, child seat and a year of Lumen Care for $3,199. A bundle protects the reference price, adds value families care about and gets new customers onto Lumen Care.

Defending the Commute. A $200 cut would need about 31% more sales to earn the same contribution and would still leave the Commute nearly twice the price of rivals. Dana's team prefers value-added pricing: keep $1,899, and make the reasons for the extra money visible, especially the certified battery, dealer test rides and service, and the app. Marcus suggests testing a lower-priced Commute trim later, but only if research shows a large segment won't pay $1,899 for the value Lumen offers.

Dynamic pricing on the website. Lumen's policy is that its website and dealers advertise the same price for each bike. If the website changed prices by the hour, dealers would lose sales whenever it undercut them, and customers who saw a different price than a friend paid would feel cheated. A $3,000 family purchase is not a hotel room that perishes overnight. Dana concludes that Lumen won't use personalized pricing and won't use the app's ride data to set prices. The team will consider AI for narrower jobs: forecasting demand by city, monitoring competitors' prices, and timing end-of-season markdowns on accessories and older models within floors and ceilings that people set. A person approves the rules.

Notice what drove each decision. The arithmetic ruled some options in and out, but the choices themselves came from positioning, channel relationships and judgments about fairness.

Your marketing plan. Draft the Pricing strategy section of your plan, in about one page. If you're using Lumen, cover the national Cargo and the defense of the Commute. If you're writing for another company, cover its main product.

1. State your pricing objective and how it follows from your positioning statement.
2. Recommend a price (or price range) and explain it using all three approaches: customer value, competitors and substitutes, and cost.
3. Show a break-even calculation with your assumptions stated. For Lumen, you may use this chapter's illustrative costs or change them, but say which you used.
4. Choose your price adjustment strategies: bundles, discounts (including dealer and fleet pricing), promotional pricing and any launch offer, and explain how they protect the reference price.
5. Take a position on dynamic or algorithmic pricing: what you would use AI for, what you would not, and what guardrails you'd set.

If you use AI to help, follow the rules from Chapter 1. Marketing in an AI Era: note where you used it, and check every number it gives you.

9.6 Try it with AI

AI assistants are good at laying out a pricing analysis and explaining the formulas. They also make confident arithmetic mistakes, confuse markup on cost with markup on price, and slip in assumptions you never gave them. In this exercise you'll have an AI build a pricing analysis, then rebuild it in a spreadsheet and compare. The numbers are this chapter's illustrative Lumen figures, so no confidential data is involved. If you adapt the exercise for a real employer, don't paste real cost data into a public AI tool.

Start a new conversation and type:

Now check the reply.

1. Rebuild it in a spreadsheet. Put each input in its own cell and write formulas that refer to those cells. Use the worked examples in this chapter to check your formulas.
2. Compare line by line. Mark each AI figure as correct, rounded differently or wrong. Did it round break-even units up? Did it compute the dealer's markup on cost, or repeat the 35% markup on price?
3. Check the mixed-channel calculation. A 50/50 mix needs a weighted average contribution margin. A common mistake is to average the two break-even volumes instead. If the AI's answer differs from yours, find the step where it went wrong.
4. Audit the assumptions. Did the AI add assumptions you didn't give it? Did it miss ones that matter, such as costs staying the same at every volume?

Then ask a follow-up that goes beyond arithmetic:

Verify the required sales increase in your spreadsheet. Then judge the AI's suggestions using the chapter: which are examples of value-added pricing, which are product-mix strategies such as bundling, and which would actually tell Lumen something about price elasticity? Finally, write two or three sentences, in your own words, recommending whether Lumen should cut the price, and note in one line how you used AI, as the disclosure rule asks.

9.7 Summary

Price is the only element of the marketing mix that produces revenue. Customer perceptions of value set the ceiling, costs set the floor, and competitors and substitutes help decide where in that range a price falls, so value-based, cost-based and competition-based pricing work best together. Contribution margin, break-even volume and the sales increase needed to justify a price cut let you test a price before you commit to it. Within the range, pricing depends on objectives and positioning, the rest of the mix, costs, demand and elasticity, competitors, resellers, regulation and customers' sense of fairness.

Companies use skimming or penetration for new products; line, optional, captive, two-part, by-product and bundle pricing for product mixes; and discounts, segmented, psychological, promotional and geographical pricing to adjust prices. AI has made dynamic and algorithmic pricing widespread. It can match prices to demand across huge assortments, but it brings risks of unfairness, collusion through shared software and errors at machine speed, which call for guardrails, clear rules about data and a person who owns every pricing decision.

9.8 Review Questions

1. Explain why customer value sets a ceiling on price and cost sets a floor. Where do competitors' prices fit in?
2. A shop buys a helmet for $45 and sells it for $75. Calculate the markup on cost and the markup on selling price. Which one would the shop's owner most likely quote, and why does the difference matter?
3. A product sells for $120 and has variable costs of $70 per unit. Fixed costs are $200,000 a year. Calculate the contribution margin per unit, the contribution margin ratio and the break-even volume in units and in dollars.
4. Using the same product, how much would unit sales need to increase to keep total contribution the same if the price were cut to $110? What would you want to know before recommending the cut?
5. Choose a product you bought recently. Was it priced with a skimming or a penetration strategy at launch, or neither? What evidence supports your answer?
6. Give an example of captive-product or two-part pricing from your own life. Do you think the pricing is fair? Why or why not?
7. Distinguish dynamic pricing based on demand from personalized pricing. Why do customers tend to accept the first more readily than the second?
8. A company's pricing software, supplied by a vendor that also serves its competitors, recommends a price increase that competitors adopt at the same time. What questions should the company's managers ask, and what does "own the decision" require of them?