Most teams will spend a quarter shaving a few points off the cost of goods and a year fighting for incremental volume. The dial that beats both is sitting on the pricing page, untouched since launch, because changing it feels riskier than it is, and because nobody owns it.

The quick version

  • Pricing is the number you charge; packaging is how you bundle, tier and fence what the customer gets for it. Both are growth levers, not admin you settle at launch.
  • Price has outsized leverage on profit: McKinsey's analysis of S&P 1500 economics found a 1% price improvement, with volume held, lifts operating profit by roughly 8%, far more than the same move on cost or volume.
  • Good packaging captures the value you've already built. Tiered "good-better-best" structures and a sensible value metric (what you charge per) let different customers self-select into the right price.
  • Treat pricing as a repeatable practice: anchor it to customer willingness-to-pay, give it an owner, and revisit it on a cadence, not once a decade.

The idea in depth

Pricing and packaging are usually decided in a rush near launch, a number borrowed from a competitor, a feature list split into three columns the night before the page ships, and then frozen for years. That instinct is backwards. These are two of the highest-leverage decisions a commercial team owns, and they reward deliberate, repeated attention more than almost anything else on the roadmap.

Why price is the highest-leverage number you own

The arithmetic is stark. In their McKinsey Quarterly article "The power of pricing" (2003), Michael Marn, Eric Roegner and Craig Zawada modelled the average economics of the S&P 1500 and found that a 1% improvement in price, holding volume steady, produced about an 8% increase in operating profit, a larger effect than a 1% cut in variable cost or a 1% lift in volume. The mechanism is simple: a price increase that holds drops almost entirely to the bottom line, because it adds no cost. The same logic runs in reverse, they note that charging 1% below the optimal price forfeits roughly 8% of potential operating profit.

A number with that much pull deserves the same machinery as any other growth channel. So the move is to put pricing on a cadence and give it an owner. Name someone accountable for price and packaging, set a rhythm, quarterly review, annual list-price decision, and stop leaving it to whoever happens to be in the room when a deal stalls. Inertia is a strange thing to hand your most powerful lever to.

The honest limitation: that "1% → 8%" figure is an average across a broad index, and it assumes volume holds. In a price-elastic market, or against a well-funded competitor, a price rise can lose enough customers to wipe out the gain, the model says where the leverage is, not how much you personally can take. Which is why the next question is not "how high can we go?" but "how much is this worth to the customer?"

Packaging captures the value pricing alone can't

A single price asks every customer to be the same. They aren't, and that gap between what different customers will happily pay is money left on the table. Packaging is how you recover it: by offering structured choices that let high-value buyers pay more and price-sensitive buyers stay in, without you having to negotiate each one by hand.

The cleanest structure is the one most of us already navigate without thinking, good-better-best. In his Harvard Business Review article "The Good-Better-Best Approach to Pricing" (2018), Rafi Mohammed argues that a tiered line-up does two jobs at once: a strong "good" tier keeps price-sensitive customers from walking, while a premium "best" tier gives high-value customers something worthy of their wallet, and the mid "better" tier, framed against both, is where many land. Three options also reframes the customer's question from "do I buy or not?" to "which one?".

flowchart TD
    V(["The value you've built"]) --> P(["Pricing: the number you charge"])
    V --> K(["Packaging: how you bundle, tier & fence it"])
    P --> G(["Good: keeps price-sensitive buyers in"])
    K --> G
    K --> B(["Better: the default for most"])
    K --> X(["Best: for high-value buyers"])
    G --> R(["Customers self-select
into the right price"])
    B --> R
    X --> R
Pricing sets the number; packaging lets different customers sort themselves toward it. Leaders Loop

Tiers only work if the fences between them are real, the feature limits or usage caps that stop a high-value customer happily living on the cheap plan. And tiers need a sensible value metric: the unit you charge per (per seat, per GB, per transaction, per active user). A good value metric grows with the value the customer gets, so their bill rises as they succeed rather than as a renewal-time surprise.

The move here is to audit your line-up for two things: a value metric that tracks customer value, and fences that actually hold. If your power users could comfortably stay on the entry tier forever, your fences are in the wrong place and you are subsidising your best customers. If everyone picks the cheapest option, your "good" tier is too generous or your "better" tier isn't differentiated enough to be worth the step up.

Pricing is the number; packaging is the menu. Set them once and you are managing the most powerful lever on the business by accident.

The limitation worth naming: good-better-best is a default, not a law. Too many tiers paralyse buyers; the wrong fences annoy them; and a value metric that punishes engagement (charging per login, say) can quietly suppress the very adoption that drives expansion. Mohammed himself frames the tiers as a starting structure to test and tune, not a fixed recipe, the right answer is the one your own data, not a template, points to.

Anchor it to willingness-to-pay, not to cost

The most common pricing mistake is to start from cost, add up what it took to build, add a margin, post the number. That ignores the only thing that determines whether a price holds: what the customer is actually willing to pay. Pricing strategist Hermann Simon, co-founder of Simon-Kucher, puts it plainly, price is where value and money meet, and the discipline is to understand the value before naming the money.

You don't have to guess at willingness-to-pay; it can be researched. Two long-standing survey methods do exactly this. The Van Westendorp Price Sensitivity Meter (introduced by Dutch economist Peter van Westendorp in 1976) asks customers at what prices a product feels too cheap, a bargain, expensive-but-worth-it, and too expensive, and reads an acceptable range off where those answers cross. The older Gabor-Granger method (André Gabor and Clive Granger, 1960s) tests whether customers would buy at specific price points to estimate a demand curve and a revenue-maximising price. Neither is perfect, surveys overstate what people say they'll pay versus what they do, but both beat a number pulled from cost-plus or a competitor's page.

So talk to customers about value before you set the price. Even a handful of structured willingness-to-pay conversations, the discipline Madhavan Ramanujam calls having "the talk" early, surfaces which features people would actually pay for and which they expect for free. (This is the same muscle as customer needs identification: you are pricing the need, not the feature.) The limitation is that stated willingness-to-pay drifts from real behaviour, so treat research as a strong hypothesis you confirm with a live test, a price experiment on a cohort, a new tier for new customers, before you bet the whole book on it.

A worked example

The figures below are illustrative, chosen to show the mechanics rather than to report a real company.

Picture a small B2B analytics product with one flat plan: £50 per user per month, everyone on the same features. Growth has stalled. Enterprise buyers churn because the product feels "consumer-grade"; tiny startups bounce off the price entirely. A single number is failing both ends of the market at once.

The team does three things in sequence. First, willingness-to-pay research: a Van Westendorp survey and a dozen customer conversations reveal that large accounts value security, audit logs and SSO far more than the headline dashboards, while small teams mostly need the core charts. Second, repackage into good-better-best: a Starter tier at £25/user (core charts, capped at 5 users, the fence) to keep startups in; a Team tier at £55/user as the default; and a Business tier at £90/user bundling the SSO, audit logs and security that enterprise buyers named. Third, a value-metric tweak: usage-heavy data exports move to a metered add-on, so the customers extracting the most value pay proportionally.

flowchart LR
    A(["One flat plan
£50/user, all features"]) --> B(["Research willingness-to-pay
(survey + customer talks)"])
    B --> C(["Starter £25: core only, 5-user cap"])
    B --> D(["Team £55: the default"])
    B --> E(["Business £90: SSO, audit, security"])
    C --> F(["Blended ARPU up
churn down at both ends"])
    D --> F
    E --> F
From one number that fits no one to three that fit the market, same product, repackaged around value. Illustrative. Leaders Loop

Nothing about the product changed. The same software, repackaged around what different customers value, now lets startups in at a price they accept, holds the middle, and lets enterprise buyers pay for what they were always going to need. The blended revenue per account rises and churn falls at both ends, not because the team built more, but because they stopped charging everyone the same.

Frequently asked questions

What's the difference between pricing and packaging?

Pricing is the number, what you charge. Packaging is the structure around it: how you bundle features, split them into tiers, set fences between those tiers, and choose the value metric you charge per. You can change packaging without changing your headline price, and vice versa. Most "pricing" problems are really packaging problems in disguise, the number is fine, but everyone is on the wrong plan.

How often should we change prices?

Not constantly, but far more than "never." The practical answer is to review pricing and packaging on a cadence, many software teams do a serious review yearly and minor adjustments more often, and to give one person clear ownership of it. The failure mode is treating launch pricing as permanent and only revisiting it in a crisis. Set a rhythm, watch the data between reviews, and change deliberately rather than reactively.

Should we just price below the competition to win?

Rarely, as a default. Underpricing forfeits the profit the McKinsey analysis shows is hardest to recover, trains customers to expect cheap, and signals lower quality. Competing on price is a viable strategy only if you have a genuine, durable cost advantage. Otherwise, compete on value, understand what your product is worth to the customer and price toward that, using packaging to serve the price-sensitive segment without dragging your whole price down.

What is a "value metric" and why does it matter?

It's the unit you charge per, per seat, per GB stored, per transaction, per active user. A good value metric rises with the value the customer gets, so their bill grows as they succeed, which feels fair and drives natural expansion. A bad one is disconnected from value (a flat fee regardless of usage) or actively punishes the behaviour you want (charging per login suppresses engagement). Picking the right metric is often a bigger lever than the price level itself.

How do I find willingness-to-pay without guessing?

Combine talking and testing. Structured customer conversations and survey methods like Van Westendorp or Gabor-Granger give you a research-backed range. Then confirm it with a live experiment, a new price for new customers, a tier tested on one cohort, because what people say they'll pay drifts from what they do. Treat research as the hypothesis and a controlled test as the proof.

Related in the Toolkit

Where to go next