Pricing Strategy

Why pricing is the single single best marketing decision

Pricing is the most underweighted decision in most marketing organizations. Hermann Simon, who has spent forty years studying it, puts the math plainly: a one percent improvement in price, holding volume constant, drives roughly an eleven to twelve percent improvement in operating profit for a typical industrial business. The same one percent improvement in volume yields three to four percent. The same in cost reduction yields two to three. Price wins, and it wins by a multiple, every time.

Most CMOs do not own the price. That is the problem. Price gets set by finance, by sales leadership, by a product manager pulling a number from a competitive scan, sometimes by the CEO on a whim. The marketing function inherits whatever number was set and is then asked to defend it, communicate it, and discount it under pressure. This is backwards. Price is the sharpest expression of brand positioning you have. If you do not own it, your positioning is being decided by whoever does.

The deeper issue is intellectual. Pricing is treated as a finance exercise (cost-plus margins) or a sales exercise (whatever the customer will pay today) when it is, properly understood, a value-perception exercise. And value perception is the marketing function's job. Until you treat pricing as a marketing discipline with the same rigor you apply to brand and demand, you are leaving the largest single profit lever on the table.

Cost-plus vs. value-based vs. competitive pricing

There are three pricing logics in common use, and only one of them is defensible at any scale.

Cost-plus pricing takes your unit cost, adds a fixed margin, and prints a price. It is mechanically simple and intellectually bankrupt. It tells you nothing about what your customer is willing to pay, anchors your entire commercial logic to your own internal inefficiency, and guarantees that any cost reduction gets passed straight to the customer instead of dropping to the bottom line. The only place cost-plus is appropriate is in regulated utilities or government contracting where margin caps are imposed externally.

Competitive pricing benchmarks against what others charge and positions a delta (typically below, sometimes above). It is the default for category followers and for any company that has not done the work to understand its own differentiation. Competitive pricing is reactive: it cedes the price-setting initiative to competitors and assumes their pricing is rational, which it frequently is not. If you set your price as a percentage of a competitor's price, you are letting that competitor's worst pricing decisions become your strategic ceiling.

Value-based pricing, the approach Simon and Madhavan Ramanujam both argue is the only defensible model, starts with the customer's perceived value and works backward. What is the economic value your product delivers to this customer (cost reduction, revenue gain, risk reduction, time saved)? What is the next-best alternative they would otherwise use, and at what cost? Where in that value spectrum should your price sit? Value-based pricing requires customer research, segmentation, and willingness-to-pay analysis. It is harder. It is also the only approach that lets you capture the value you actually create.

A practical rule from Ramanujam's Monetizing Innovation: if you have built a product without first understanding what segments will pay what amounts for what features, you have built a feature shock, a minivation, a hidden gem, or an undead. Four failure modes, all caused by treating price as a launch afterthought rather than a design input.

The price-anchoring instinct, and how it actually works

Daniel Kahneman's work on anchoring is the foundation of every modern pricing tactic. The brain does not evaluate prices in absolute terms. It evaluates them in relation to a reference point, and the reference point is whichever number it encountered first or most recently. Show a customer a $5,000 watch next to a $500 watch, and the $500 watch reads as accessible. Show them the $500 watch first, in isolation, and it reads as expensive.

This is not a quirk to exploit. It is the default operating system of human price perception, and you are either setting the anchors deliberately or letting your competitors set them for you. Every category has anchors: the dominant incumbent's price, the most-advertised promotion, the round number a customer remembers from last year. If you do not consciously place your price against a chosen anchor, the customer will choose one for you, and they will usually choose the cheapest reference point available.

The practical applications are direct. List a premium option first on a pricing page so the lower options read as deals. Show original-price-with-strikethrough next to your actual price so the customer registers savings, not cost. Quote large enterprise contracts in annual rather than monthly figures, then break them down to monthly for the comparison sale. Anchor your B2B proposal against the cost of the problem (lost revenue, headcount, opportunity cost) rather than against the cost of the solution. The anchor frames the entire comparison.

Anchoring is reinforced by Tversky and Kahneman's prospect theory: people feel losses roughly twice as intensely as equivalent gains. A discount framed as "save $200" outperforms one framed as "20% off" because losses-avoided activate the same brain circuitry as gains-realized, only stronger. The marketers who win pricing competitions are the ones who understand the brain is asymmetric about money and design every price exposure around that asymmetry.

The decoy effect (Ariely's wine pricing example)

Dan Ariely's research at MIT produced one of the most replicated findings in pricing: the decoy effect (also called asymmetric dominance in the academic literature, same idea, different label). When a customer is choosing between two options, adding a third option that is clearly inferior to one of them but not the other systematically shifts choice toward the dominated option.

The canonical example is The Economist subscription page Ariely dissected in Predictably Irrational. The original page offered three options: web-only at $59, print-only at $125, and print-and-web at $125. Almost nobody chose print-only at $125 because print-and-web at the same price was strictly better. But the presence of that "useless" $125 print-only option dramatically increased the share of customers who chose the $125 print-and-web option over the $59 web-only one. When Ariely tested removing the print-only decoy, the choice distribution flipped: most people chose the cheap web-only option.

The decoy did no work as a standalone purchase. It did enormous work as a frame. Its job was to make the print-and-web option look like a deal by comparison, and it did. The decoy effect generalizes well beyond magazines: software pricing pages, restaurant wine lists, real estate comparables, automotive trim levels. Wherever a customer is asked to choose between options, an inferior third option engineered to flatter the target choice will move share.

The discipline is in using this without insulting the customer. A decoy that is too obviously a decoy reads as manipulation and damages trust. A decoy that is subtle (a slightly worse value at the same price, or a slightly better value at a noticeably higher price) does its work invisibly. Tiered pricing pages are the natural home for the technique. Three tiers, with the middle tier engineered to look like the obvious choice, is not lazy. It is a deliberate application of forty years of decision-science research.

Tiered pricing and segmentation

Tiered pricing is not a packaging convenience. It is a segmentation tool. Customers in the same category have wildly different willingness-to-pay based on use case, organization size, risk tolerance, and how central the product is to their operation. A single price asks every segment to pay the same. Tiered pricing, designed well, lets each segment self-select into the tier that matches their value extraction.

Ramanujam's framework in Monetizing Innovation is the cleanest articulation: build a "good, better, best" tier structure where each tier targets a distinct willingness-to-pay segment. The good tier captures price-sensitive customers who would otherwise not buy at all (or would defect to a cheaper competitor). The better tier is the volume tier, engineered to be the rational default for most customers. The best tier captures the high-value segment that wants premium features and is largely price-insensitive at the top of the range.

The mistake most teams make is to design tiers around features rather than around segments. The right sequence is reversed. Identify the segments, map their willingness-to-pay, then design feature bundles that match the value each segment perceives. Features in the best tier should be ones the best-tier segment specifically values and the other segments largely do not. Features in the good tier should be the minimum viable bundle that delivers core value without cannibalizing the better tier.

A useful diagnostic from Georg Tacke, Simon's co-author on Confessions of the Pricing Man: if your best tier accounts for less than ten percent of revenue and your good tier accounts for more than thirty, your tier design is leaving money on the table. The best tier should be aspirational but achievable; the good tier should feel limited enough that price-tolerant buyers move up.

Subscription, freemium, and usage-based models

Pricing is no longer just a number. It is a structure. The choice of model (one-time purchase, subscription, freemium, usage-based, hybrid) is itself a strategic decision with first-order revenue and retention consequences.

Subscription pricing converts a transactional relationship into a recurring one. The economics are well understood: predictable revenue, higher lifetime value, but also higher acquisition friction (customers commit to ongoing spend) and a permanent churn-management problem. Subscriptions favor products with ongoing value delivery (software, media, consumables, services). They are punishing for products with diminishing post-purchase value.

Freemium splits the user base into a free tier and a paid tier, with the free tier serving as a top-of-funnel acquisition mechanism. The model works only when the free tier delivers genuine standalone value (otherwise users do not stick), the conversion rate to paid is high enough to amortize the free-tier cost (typically two to five percent for consumer, ten to twenty for B2B with strong product-market fit), and the marginal cost of serving free users is low enough that the model is solvent. Freemium fails most often when teams optimize for free signups without engineering the bridge to paid.

Usage-based pricing aligns cost with value consumed. The customer pays for what they use (API calls, compute, transactions, seats). Done well, it removes the purchase decision (no per-feature gating), accelerates land-and-expand growth, and scales revenue with customer success. Done poorly, it creates unpredictable bills that erode trust and trigger budget-control reflexes in procurement.

ProfitWell's recurring-revenue research consistently finds that the highest-performing SaaS companies use hybrid models: a base subscription for predictability, plus usage-based components for expansion. Pure subscription caps revenue at the chosen tier price. Pure usage-based makes revenue volatile and customer-acquisition difficult. The hybrid captures both stability and expansion.

Discount theory: when never to discount, when discount is the right move

Discounting is the most overused tool in the marketing kit and one of the most damaging when applied without discipline. The mechanics are unforgiving. A twenty percent discount, holding volume constant, requires roughly a thirty to forty percent volume increase just to break even on contribution margin for a typical business. Volume increases of that scale, sustained by a discount, are rare. The default outcome of a discount is reduced profit per unit with insufficient volume compensation, plus a customer base trained to wait for the next sale.

There are conditions under which a discount is the correct move. Inventory clearance for time-sensitive product (apparel end-of-season, perishables, prior-version stock) where the alternative is zero recovery. Price discrimination across segments where you can ring-fence the discount so it does not cannibalize full-price sales (student discounts, geographic pricing, time-of-day rates). Customer acquisition trials where the discount is a finite cost of acquiring a high-LTV customer (first-month free, founding-member rates). Channel-specific promotions where the discount creates the listing or placement that volume requires.

The conditions under which discounting is wrong are more common. When the discount becomes seasonal expectation (holiday sales, end-of-quarter sales) you have not generated new demand, you have shifted full-price demand into discount windows. When the discount is permanent (perpetual "sale" pricing) you have simply lowered your price while still paying the brand cost of looking like a discounter. When the discount is across-the-board rather than ring-fenced, every full-price customer who learns about it feels overcharged. When the discount is reactive to a competitor's promotion, you have ceded pricing initiative and trained your customers to expect concessions.

Hermann Simon's rule: if you are going to discount, do it deliberately, with a defined start and end, a defined segment, and a defined commercial outcome (volume, acquisition, clearance). If those are not all defined in advance, you are not discounting strategically. You are panicking.

Price elasticity and the Van Westendorp Price Sensitivity Meter

You cannot set price intelligently without an estimate of price elasticity. Elasticity is the percentage change in volume that results from a percentage change in price. An elasticity of -2 means a ten percent price increase produces a twenty percent volume decline. An elasticity of -0.5 means the same increase produces only a five percent decline. The first product has lots of substitutes and price-sensitive customers; raising price is dangerous. The second has differentiated value and price-tolerant customers; raising price is probably overdue.

Estimating elasticity rigorously requires either historical price-variation data (rare in B2B, common in consumer retail) or controlled experimentation (which carries the legal and ethical complications discussed below). The Van Westendorp Price Sensitivity Meter, developed by Peter van Westendorp in 1976, is the most widely used survey-based approximation when experimentation is not feasible.

The method asks customers four questions about your product: at what price would it be so expensive that you would not consider buying it? At what price would you start to think it is expensive but still consider buying it? At what price would you think it is a bargain, a great buy? At what price would you think it is so cheap you would question its quality? The intersections of the resulting cumulative curves identify a range of acceptable prices: the optimal price point (OPP), the indifference price point (IPP), and the points of marginal cheapness and marginal expensiveness.

Van Westendorp is not a substitute for revealed-preference data (what customers actually do when offered the product at a given price). Stated willingness-to-pay is systematically inflated relative to actual willingness-to-pay. But as an early-stage triangulation tool, especially for new products without sales history, it gives you a defensible range to test against rather than a number pulled from intuition.

A more rigorous extension is conjoint analysis, which forces respondents to make trade-offs between feature bundles at different prices and infers utility weights from their choices. Conjoint takes more research investment but produces a willingness-to-pay model that survives scrutiny, which is what you want before committing to a launch price.

A/B testing prices (the legal and ethical landmines)

The temptation in any digital business is to A/B test prices in production. Show one cohort $49, another $59, measure conversion and revenue per visitor, pick the winner. The math is clean. The legal and ethical exposure is not.

Charging different customers different prices for the same product at the same time is, in most jurisdictions, legal if the segmentation is not based on protected classes (race, gender, religion, national origin) and if it is not deceptive. It becomes illegal under U.S. law if it constitutes price discrimination in violation of the Robinson-Patman Act (B2B contexts), if it violates state consumer protection laws against "dynamic pricing" that targets vulnerable customers, or if it crosses into deceptive trade practice (advertising one price and charging another). In the EU, the Digital Services Act and GDPR add transparency requirements around personalized pricing.

The ethical exposure is independent of the legal exposure. When customers discover they paid more than a peer for an identical product, the trust damage is severe and durable. The Amazon dynamic-pricing controversy of 2000, the various airline pricing complaints, and the Uber surge-pricing backlashes all share the same root: customers will accept that prices change over time, but they will not accept that they personally were charged more than the customer next to them for the same thing in the same moment.

The practical approach: A/B test pricing structure (number of tiers, feature bundling, billing cadence) rather than price level on identical SKUs. Test price changes longitudinally (price was $49 last month, $59 this month for all customers) rather than cross-sectionally (price is $49 for cohort A, $59 for cohort B simultaneously). Use price experimentation in new-product introduction or new-segment expansion where there is no prior anchor to violate. If you must run a simultaneous price test on identical product, ring-fence it tightly, document the experimental rationale, and offer a price-match guarantee to anyone who paid the higher price during the test window.

Ramanujam, who has run more pricing experiments than almost anyone in B2B, recommends a stricter posture: test on willingness-to-pay surveys and conjoint, not on live prices, until you have a model that justifies the commercial decision. Then change the price for everyone. The information value of a live price test rarely justifies the trust risk.

Further Reading

• Hermann Simon, Confessions of the Pricing Man (Springer)
• Madhavan Ramanujam, Monetizing Innovation (Simon-Kucher)
• McKinsey: The power of pricing
• ProfitWell pricing strategy blog
• Dan Ariely, Predictably Irrational (free chapter on decoy pricing)