Kustomer's Shift to Outcome-Based Pricing for AI

We ran the work in four phases: discovery, research, design, and operationalization.

In discovery we held open conversations with the leadership team to define the problem, surface internal and external headwinds, risks, and early hypotheses. That phase was short but critical because leadership had the deepest historical view of pricing.

Research combined deep competitive analysis with primary customer insight. We examined emerging players, legacy CX platforms, and—crucially—disruptive usage- or outcome-based models outside our category (Snowflake, Databricks, Twilio) to identify useful units of value. We then surveyed more than 200 CX leaders (customers, competitors’ customers, and evaluators) on pricing-model preferences, adoption timelines, key decision-makers (often CFO or CEO in smaller firms), and willingness to pay across features.

Design turned those insights into packages and price points. We set clear principles for upsell paths and add-ons, modeled AI, hosting, and operational costs to protect gross margins, and defined overage behavior. Anything with a numerical limit became a pre-packaged, discountable add-on plus a pay-as-you-go safety net so usage never stopped; binary “checkmark” features simply unlocked in higher tiers. We validated packaging by literally walking through the customer’s shoes in cross-functional workshops.

Operationalization moved the model from theory to every touchpoint—product limits, billing systems (including Zuora price books), website, and documentation—so nothing was ambiguous. We structured stakeholders in layers: the full executive team for regular progress checks, a smaller pricing committee (including the VP of Finance) to pressure-test recommendations, and a broad operationalization group spanning sales, CS, product marketing, enablement, systems architecture, FP&A, legal, and billing. Post-launch we stood up a SWAT team (suggested by our new COO) with explicit adoption goals so we could course-correct positioning or packaging quickly from field feedback.

We launched in October as a major high-visibility moment—LinkedIn Live, full website overhaul, and debut of our new AI agent—while explicitly giving customers time to transition at their own pace, using the Netflix DVD-to-streaming multi-year shift as our internal analogy.

For sales enablement we built a focused set of assets:

• A comprehensive FAQ covering when new packages could be sold, package differences, and common questions.
• An internal interactive pricing tool (spreadsheet) limited to five inputs—number of seats, number of conversations, proposed package, and desired add-ons—that produced a clean output chart. (A customer-facing web version is planned next.)
• Pricing 101 training built around real-world scenarios across verticals and company sizes so reps practiced recommending the right package.
• An objection-handling guide and standard talk tracks; messaging for existing customers emphasized renewal support and transition clarity, while messaging for new logos focused on value, flexibility, and modern CX alignment.

These assets, plus tight coordination with billing and systems teams, kept the rollout consistent.

We successfully introduced an outcome-based pricing model in which customers pay when the platform actually helps them rather than for seats they may under-utilize. The October launch was treated as a flagship moment rather than a quiet update, and we deliberately allowed customers to adopt on their own timeline instead of forcing an overnight switch. A dedicated post-launch SWAT team now tracks adoption goals and feeds rapid refinements back into packaging and positioning. The lasting lesson is that pricing is never “set it and forget it”—continuous feedback and the willingness to adjust keep the model working for both the business and our customers.