Pricing Strategy — Value-Based Monetization & Packaging

The strategy of what to charge, how to meter it, and how to package it. This is distinct from the pricing-*page* design patterns (3 tiers, middle highlighted, annual toggle) in conversion-ux.md and the mobile paywall benchmarks in paywall-benchmarks.md — those cover how to present a price; this covers how to set and structure one. Pricing is the highest-leverage growth lever there is: a 1% improvement in price typically flows almost entirely to profit, more than an equivalent gain in volume or cost.

The governing principle (Ramanujam & Tacke, Monetizing Innovation): design the product around the price, not the price around the product.

Part 1 — Value-Based Pricing & Willingness-to-Pay First

Three ways to set a price, in ascending order of sophistication:

Cost-plus (add a margin to your costs) — ignores what the customer will pay; leaves money on the table for high-value products and overprices low-value ones.
Competitor-based (match the market) — a race to the bottom; treats your product as a commodity and cedes strategy to rivals.
Value-based (price to the value the customer perceives and their willingness to pay) — the only method that captures the value you actually create. This is the default recommendation.

Design the product around the price

Monetizing Innovation (Ramanujam & Tacke, Simon-Kucher) studied why ~72% of new products miss their financial targets. The root cause is almost always pricing treated as an afterthought — built the product, then slapped a price on at launch. Their prescription: have the willingness-to-pay (WTP) conversation early, before and during development, and let it shape what you build. If a feature commands no WTP, it may not be worth building; if one feature drives most of the WTP, it may deserve to be the paid tier's centerpiece.

Their four failure archetypes (diagnostic vocabulary worth knowing):

Feature shock — cramming in so many features that value is diluted and the price is unjustifiable/confusing.
Minivation — under-pricing/under-scoping a genuine innovation; leaving huge value uncaptured.
Hidden gem — a feature with high WTP buried or given away free because the team didn't test it.
Undead — a product customers never wanted, kept alive despite no WTP.

Practical takeaway for SaglitzDesign engagements: pricing and packaging belong in the design/strategy phase, not the launch checklist. Ask "what would people pay for this, and which slice drives that?" while the product is still malleable.

Part 2 — The Pricing Metric (What You Charge For)

The pricing metric is the unit you bill against. Choosing it well matters more than the number — a good metric scales revenue with the value the customer receives, so customers who get more, pay more, and feel it's fair.

Per-seat / per-user — simple, predictable, easy to forecast; the historical B2B SaaS default. Weakness: it decouples price from value (a seat used all day and a seat used once cost the same), it penalizes adoption (buyers ration seats), and it collapses when AI does the work a "user" used to (agentic products have fewer human seats).
Usage / consumption-based — bill on the value-correlated unit: API calls, GB stored, messages sent, credits, compute minutes, tokens, transactions processed. Aligns cost with value and removes the adoption penalty. Weakness: revenue is less predictable for both sides; can create bill-shock anxiety that suppresses usage unless you add caps, alerts, and dashboards.
Hybrid (base + usage) — a platform/base fee for predictability plus a usage component that scales with value. Captures the upside of usage pricing while giving both parties a revenue floor.

2026 trend: hybrid/usage is now the mainstream

Per-seat pricing is in structural decline. Per Kyle Poyar's 2026 State of B2B Monetization survey (230+ software companies), hybrid is now the single most common primary structure (~37%), pure per-seat has fallen to roughly 15% of the market (down from ~21% a year earlier), and the large majority of SaaS companies now incorporate at least some usage-based component. The driver is AI: AI features create non-linear consumption (tokens, credits, compute) that seats simply cannot price, and AI agents reduce the number of human seats, breaking the per-seat model's core assumption. Customers report usage pricing aligns better with value received. Recommendation for new products in 2026: default to a hybrid base-plus-usage structure unless there's a strong reason not to, and pick a usage metric customers already associate with getting value.

Part 3 — Packaging: Good-Better-Best, Tiers & Fences

Packaging = how you bundle features into offers. The workhorse is Good-Better-Best (G-B-B) — three tiers:

Good — the entry offer; enough to be genuinely useful, deliberately missing the features power users need.
Better — the target plan; where you want most customers to land. Badge it "Most popular." Put the features that drive the highest WTP here.
Best — the premium plan; anchors value high and captures high-WTP customers. Some buyers self-select up.

Why three: too few tiers and you can't segment by WTP; too many and you induce choice paralysis (see conversion-ux.md). Three balances segmentation against cognitive load.

Tier/plan design principles:

Differentiate tiers by value metrics customers understand (seats, volume, advanced capability), not by obscure feature counts.
Each tier should have one clear reason to upgrade to the next — a "fence."
Align tiers to segments: solo/small, growing team, enterprise. Name them by who they're for when possible.

Feature fences

A fence is the barrier that makes a customer upgrade — the reason the cheaper plan won't do. Good fences track a dimension that grows with the customer's success, so upgrading feels earned, not extracted:

Volume/usage fences (records, seats, API calls, storage) — scale with adoption; the best fences.
Capability fences (SSO/SAML, roles & permissions, advanced analytics, audit logs) — gate the features specific segments (esp. enterprise) must have.
Support/SLA fences — priority support, uptime guarantees on higher tiers.

Ethical line: fence on features a segment genuinely values, not by crippling basic usability to coerce upgrades (a "roach motel" in reverse). Gating SSO purely to force enterprise onto "call us" tiers is a common, resented tactic — fence it, but price it sanely.

Anchor & decoy effects (used ethically)

Anchoring: the first/highest price a buyer sees sets their reference point. Showing the Best (premium) tier prominently makes Better look reasonable. For enterprise-leaning products, present the high anchor first.
Decoy (asymmetric dominance): a third option that exists mainly to make your target plan look like the obvious best value (the classic Economist subscription example — a print-only option priced near the print+digital bundle makes the bundle look free-upgraded). Use this only when the decoy is a real, honestly-available option — never fabricate a fake tier no one can/should buy. The ethical test: would you be comfortable explaining the tier structure to the customer? Dark-pattern pricing converts once and destroys trust and LTV (conversion-ux.md).

Part 4 — Researching Willingness-to-Pay

Don't guess the number — measure it. Three standard methods:

Van Westendorp Price Sensitivity Meter (PSM)

Ask respondents four open price questions about the product:

At what price is it so expensive you would not consider buying it? (too expensive)
At what price is it expensive, but you'd still consider it? (expensive / marginal)
At what price is it a bargain — great value? (cheap / good value)
At what price is it so cheap you'd question the quality? (too cheap)

Plot cumulative curves and read the intersections:

OPP (Optimal Price Point): where "too cheap" and "too expensive" curves cross — resistance is balanced.
IPP (Indifference Price Point): where "expensive" and "cheap/bargain" cross — the median perceived fair price.
Range of acceptable prices: from PMC (Point of Marginal Cheapness) to PME (Point of Marginal Expensiveness) — your viable band.

PSM is cheap to run and great for a first read on the acceptable range, but it measures price perception, not purchase behavior — treat it as directional, validate with the market.

Gabor-Granger

Show each respondent a series of specific price points and ask purchase likelihood at each. Produces a demand curve and a revenue-maximizing price. Better than PSM at estimating how demand falls as price rises, but it anchors on the prices you show and can't reveal a price above your highest tested point.

Conjoint analysis (briefly)

Present bundles of features-at-prices and force trade-off choices; statistically decompose which features drive choice and their implied dollar value. The most rigorous method for feature-level WTP and packaging decisions (what belongs in which tier), but heavier and more expensive to field — reach for it when packaging stakes are high.

Rule of thumb: PSM for a quick range, Gabor-Granger for the number, conjoint for the packaging. For early-stage products, qualitative WTP interviews (Monetizing Innovation-style) often beat all three — actually talk to buyers about value before you survey.

Part 5 — Freemium vs. Free-Trial vs. Reverse-Trial

The "front door" decision — how users experience value before paying. Tie this to the loop model in growth-frameworks.md and the hard-numbers in paywall-benchmarks.md.

Freemium — a genuinely-free tier forever. Maximizes top-of-funnel reach and fuels viral/content loops; low direct conversion (mobile benchmark ~2.1% vs ~10.7% for hard paywalls, per paywall-benchmarks.md). Choose when reach/network effects/virality are the growth model, or when the free tier is a genuine acquisition loop. Danger: giving away so much that no one needs to pay (a mis-set fence).
Free trial — full (or high-tier) access for a fixed window, then pay or lose it. Creates urgency and qualifies intent; converts far better per-user than freemium. Choose when time-to-value is fast enough to feel the value inside the window. On mobile, longer trials convert better (17–32 day trials → ~42.5% trial-to-paid vs ~25.5% for ≤4-day trials — yet nearly half of apps still use ≤4-day trials, the industry's costliest mistake; see paywall-benchmarks.md).
Reverse trial — start users in the paid/premium experience (no card required), then downgrade to a free tier at the end rather than cutting them off. Blends freemium's reach with a trial's value demonstration: users feel the premium value, and the fallback free tier keeps them in the funnel (and the loop) even if they don't convert immediately. Increasingly the default for PLG products in 2026 where a free tier makes strategic sense.

Decision heuristic: reach-driven loop → freemium or reverse-trial; revenue-driven, fast-TTV product → free trial (14–30 days) or hard paywall. Always pair a trial with a pre-charge reminder (trust, fewer chargebacks — paywall-benchmarks.md).

Part 6 — Psychological Pricing (Use Judiciously)

Charm pricing ($X9 / $X.99): the classic left-digit effect ($49 feels meaningfully cheaper than $50). It genuinely lifts response in consumer/impulse contexts — but caveat for premium and B2B: charm prices signal discount/value, which can undercut a premium or trust-heavy positioning. Round numbers ($50, $500, $2,000) read as more premium and more honest for high-consideration B2B. Match the tactic to the brand (see branding-identity.md).
Annual discounts: offer annual billing at ~2 months free (~15–20% off). Improves cash flow and slashes churn (an annual commitment can't churn monthly). Show the savings explicitly ("Save 20%") and default the toggle thoughtfully — but never hide the monthly option (dark pattern).
Framing: show price per the value unit the customer cares about ("$0.10 per 1,000 requests," "$8/user/month billed annually"). Present higher tiers first to anchor (Part 3). Bundle-vs-unbundle framing changes perceived value — a bundle at a round number can feel like a deal against summed à-la-carte prices.
Price = quality signal: especially where quality is hard to judge pre-purchase, too low a price destroys credibility (the "too cheap" curve in Van Westendorp is real). Don't underprice a premium product to win on price — you'll signal yourself out of the consideration set.

Part 7 — Common SaaS Pricing Mistakes

Pricing as a launch afterthought — the cardinal sin (Part 1). Have the WTP conversation during design, not the week before launch.
Cost-plus or competitor-matching instead of value-based — leaves money on the table or commoditizes you.
Setting it once and never revisiting — pricing is a living system; revisit at least annually as the product, costs (esp. AI/compute), and market shift.
Under-pricing (the minivation) — founders' most common instinct; fear of "too expensive" leaves the majority of created value uncaptured. Willingness-to-pay is almost always higher than founders assume.
One-size-fits-all single plan — no tiering means no segmentation; you overcharge small customers (losing them) and undercharge large ones (leaving money behind).
Too many tiers / feature shock — choice paralysis; buyers bounce. Three tiers, clear fences.
Value metric decoupled from value — per-seat on a product where value isn't per-person (the 2026 AI trap); revenue can't scale with the value delivered.
Discounting reflexively — every unearned discount resets the anchor and trains buyers to wait/negotiate; erodes the price integrity that protects margin.
Dark-pattern packaging — crippled free tiers, hidden fences, fake decoys, hard-to-cancel annual plans. Converts once, destroys LTV, increasingly illegal (conversion-ux.md).
No usage caps/alerts on consumption pricing — bill-shock kills trust and suppresses usage; add caps, alerts, and clear dashboards.

Quick-Reference Checklist

Was willingness-to-pay researched (PSM/Gabor-Granger/conjoint/interviews) before the price was set?
Is the price value-based, not cost-plus or competitor-matched?
Does the pricing metric scale with the value the customer receives (2026: default to hybrid base+usage; avoid pure per-seat on AI/agentic products)?
Three-tier Good-Better-Best with the target plan badged and one clear fence per upgrade?
Are fences on features segments genuinely value — no crippled-basics coercion, no fake decoys?
Front door (freemium / free-trial / reverse-trial) matched to the growth loop (growth-frameworks.md) and the numbers in paywall-benchmarks.md?
Charm vs. round pricing matched to premium/B2B vs. consumer positioning?
Annual option offered (~2 months free), monthly still visible, savings shown?
Is pricing revisited on a schedule, not set-and-forgotten?
Would you be comfortable explaining the entire pricing/packaging structure to the customer's face? (The ethics test.)