Pricing and Packaging
Design AI-era SaaS pricing models, packaging, margins, and governance.
Written and reviewed by Eric Han · Co-founder & CTO, reopt
Key takeaways
- AI changes pricing: seat-based pricing alone fails to capture AI usage and value, pure usage pricing makes budgets unpredictable, so most B2B AI SaaS needs a hybrid model.
- Price across three axes: access (who and what features), consumption (how much), and value (what work outcome).
- Pick a customer-understandable value metric instead of passing token or execution cost straight through.
- Check gross margin after AI execution and service cost before publishing a price.
- Use AI for discount approval, package recommendation, and renewal pricing, and validate packaging changes by cohort adoption and churn.
October 2026 HubSpot Credits: overage defaults depend on purchase date
HubSpot's October 20 documentation distinguishes default overage settings by the date additional Credits were purchased. Inspect the actual account settings and contract because customers can change defaults.
| Situation | Documented default | Operating check |
|---|---|---|
| Included credits only, no additional pack | Usage-based features pause when exhausted | Reset date and customer impact |
| Additional Credits purchased before September 16, 2026 | Auto-upgrade | Higher pack applies for the remaining commitment term |
| Additional Credits purchased on/after September 16 | Pay-as-you-go | Overage billing; original limit returns next cycle |
| Unused credits | Expire at usage-period end; no rollover | Track cash cost, consumption, and remaining credits separately |
Review account and feature monthly limits together. Feature limits do not reserve credits, so other features can consume the shared balance. Recheck contract rules for pricing changes, beta billing, and pack downgrades. These are HubSpot-specific rules, not universal CRM billing behavior.
Include additional packs, overages, human review, and rework in the cost model. Define vendor-billed leads/conversations/tasks separately from customer-accepted outcomes.
AI changes pricing strategy. Seat-based pricing alone often fails to explain AI usage, cost, automation value, and outcomes. Pure usage pricing can make budgets unpredictable. Most B2B AI SaaS needs a hybrid model.
Pricing Model Comparison
| Model | Strength | Risk | Best fit |
|---|---|---|---|
| Seat-based | Predictable and easy to buy | Weak link to AI usage or value | Collaboration and admin-heavy products |
| Usage-based | Connects cost and value | Harder budget forecasting | API, data processing, automation |
| Credit-based | Packaging flexibility | Can feel complex | Bundled AI features |
| Outcome-based | Direct value link | Attribution disputes | Clear cost reduction or revenue uplift |
| Hybrid | Balances predictability and scale | Harder to explain | Most B2B AI SaaS |
Three Pricing Axes
- Access: who can use which features
- Consumption: how much is used
- Value: what work outcome is achieved
Good packaging explains all three without making budget planning impossible.
Value Metric Selection
| Product type | Possible value metric |
|---|---|
| Sales AI | Accounts, prospected leads, meetings assisted |
| CS AI | Active customers monitored, tickets resolved, playbooks run |
| Data AI | Rows processed, workflows, data sources |
| Document AI | Documents analyzed, seats, projects |
| Agent platform | Tasks, runs, credits, connected systems |
Cost pass-through is not enough
Passing token or execution cost directly to the customer makes pricing feel supplier-centered. Connect cost units to customer-understandable value units.
AI Package Margin
AI features carry token, retrieval, tool-call, and human-in-the-loop costs. Before publishing a price, check gross margin after AI execution cost and service cost.
Manage Discounts and Exceptions With AI
| Area | AI role |
|---|---|
| Discount approval | Compare similar deals, margin, and win probability |
| Package recommendation | Use usage patterns and expansion potential |
| Renewal pricing | Reflect adoption, value proof, and risk |
| Pricing experiments | Detect segment response and churn |
| CPQ support | Generate guidance and approval rationale |
Packaging Principles
- Let buyers experience core value in free or lower tiers.
- Put expansion value in teams, data, automation, security, and governance.
- Use fair-use and overage rules for expensive AI features.
- Make enterprise tier security, audit, admin, and SLA explicit.
- Price pages should explain the operating problem solved, not only included features.
Operating Checklist
- Pricing reflects both AI cost and customer value metric.
- Customers can forecast monthly cost.
- Discount approval includes margin and value proof.
- Packaging changes are validated by cohort adoption and churn.
- Pricing experiment results update sales enablement.