I want to walk you through a pragmatic, reproducible 90-day pricing experiment I’ve used (and refined) with B2B SaaS teams to prove—empirically—whether enterprise buyers will pay 30% more for a specific feature. This isn’t theory. It’s a tight experiment design, sales-friendly roll-out, and measurement plan that gives you defensible evidence you can use in pricing decks, investor conversations, or launch playbooks.
Why a 90-day experiment?
Ninety days is long enough to run meaningful sales cycles in many enterprise contexts, and short enough to iterate quickly. It forces discipline: clear hypothesis, focused segmentation, and coordination between product, sales, and legal. In my experience, shorter tests create noise; longer ones waste runway and decision-making momentum.
Define your hypothesis and success criteria
Start with a crisp hypothesis. For this experiment it’s:
"Enterprise buyers will pay 30% more in ARR for our product if we package Feature X as a premium, enterprise-only add-on with tailored value messaging and flexible contracting."
Next, set measurable success criteria. I use both revenue and behavioral KPIs:
- Primary KPI: Average deal value (ARR) for deals that include Feature X increases by ≥30% compared to control.
- Secondary KPIs: Attachment rate (percent of applicable deals that include Feature X); sales cycle length; win rate; churn risk indicators (e.g., discount requests, contract clauses).
- Statistical target: Minimum sample size to detect a 30% lift with 80% power (we’ll estimate this below).
Segment your accounts—don’t treat everyone the same
Enterprise buyers are heterogeneous. You must pick segments where Feature X is likely to deliver clear ROI. I typically split targets into:
- Strategic accounts (top 20% by ACV)—hand-sold, exec-level influence
- Mid-market enterprise (next 30%)—sales-led, predictable cycle
- Low-touch enterprise (lower ACV but complex requirements)
For the 90-day experiment, focus on the segment where Feature X solves a known pain and where sales contacts can champion higher price—a sales-enabled enterprise segment is often best.
Design the pricing variants and packaging
Keep the test simple: two arms only.
- Control arm: Current pricing (Feature X included or free trial as per status quo).
- Treatment arm: Feature X offered as a premium add-on priced at +30% relative to baseline ARR for comparable accounts, or as part of a higher-tier bundle that increases total ARR by 30%.
Variant tips I’ve learned:
- Use bundles rather than arbitrary surcharges—bundle Feature X with implementation services, SLAs, or analytics to justify price.
- Offer flexible payment terms (annual vs. monthly) and optional pilot discounts, but limit scope to avoid confounding variables.
- Maintain consistent discounting policy across arms. If sales historically discounts 20%, run the same discount allowance in both arms to see true uplift.
Enable sales with scripts, objection handlers and ROI tools
Sales must be armed with clear positioning and ROI calculations. I provide:
- One-page value prop that ties Feature X to quantitative outcomes (time saved, risk reduced, revenue enabled).
- Battlecards for common objections (e.g., "We don’t have budget", "We need it included").
- A simple ROI calculator spreadsheet that translates Feature X into dollar outcomes for the buyer.
Train sales in a 45-minute session and run roleplay scenarios. I also shadow early deals to ensure fidelity to the experiment.
Randomization and assignment
True A/B requires random assignment to avoid selection bias. Practically, we often use stratified randomization:
- Within each bucket (e.g., Strategic, Mid-market), randomly assign eligible opportunities to control or treatment.
- Exclude renegotiations or existing contracts where Feature X was previously promised.
Document the assignment in your CRM (custom field: Pricing Experiment Cohort) so reporting is clean.
Tracking and analytics
Key metrics to capture in your CRM and analytics stack:
- Account cohort (control vs. treatment)
- Deal stage progression timestamps
- Final ARR and line-item details (which includes Feature X)
- Discounts granted and contract term
- Sales rep and region (to detect confounders)
Weekly scrums to review pipeline and early signals prevent surprises. I like building a simple dashboard with:
| Metric | Control | Treatment |
| Average ARR | — | — |
| Attachment rate | — | — |
| Win rate | — | — |
Sample size and statistical significance
You need to estimate how many closed deals you’ll need. If your average deal is $100k ARR and you expect a 30% lift to $130k, plug inputs into a power calculator. As a rule of thumb:
- If your deal volumes are low (fewer than 20 closed deals per cohort), focus on directional qualitative signals (sales feedback, negotiation friction).
- If you have 30+ closed deals per cohort, you can often achieve statistical power to detect large effects (30%+).
When numbers are small, combine quantitative with qualitative evidence: consistent willingness to sign contracts at higher price, reduction in seat-based negotiation, and references from pilot customers.
Contracting and legal guardrails
Coordinate with legal early. Provide template addendums for the premium bundle to speed negotiation. Key guardrails:
- No backdating or grandfathering that could muddy numbers.
- Clear explication of the premium feature’s scope and SLAs.
- Defined pilot period terms (if any) and renewal pricing.
Timeline and weekly milestones
Here’s a lean 90-day plan I use:
- Week 0–1: Finalize hypothesis, segments, pricing variants, and legal templates.
- Week 2: Sales enablement training and CRM setup.
- Week 3–8: Live selling period—close deals, capture data, and perform weekly check-ins.
- Week 9–11: Analyze closed deals, run robustness checks, collect qualitative feedback.
- Week 12: Final report and decision meeting (scale, iterate, or shelve).
Interpreting results and next steps
When the 90 days end, don’t just look at the headline ARR uplift. I examine:
- Whether uplift came from increased license fees, contract length, or services.
- Changes in win rate and sales cycle—did price increase deter buyers?
- Customer feedback—are buyers describing the premium feature as "mission-critical" or "nice-to-have"?
If the treatment arm shows ≥30% uplift without material deterioration in win rate or increased churn risk, you have a strong case to roll out pricing changes more broadly. If uplift is ambiguous, iterate: refine messaging, adjust bundle components, or target a different segment.
Common pitfalls to avoid
- Allowing non-experiment discounts or exceptions—this kills data integrity.
- Mixing multiple changes (e.g., price + major feature UI changes) at the same time.
- Not training sales—poor positioning will bias results toward failure.
- Running experiments during atypical buying seasons (budget freezes, fiscal year-end).
I’ve run versions of this experiment with startups and scale-ups—sometimes the 30% target is conservative, sometimes optimistic. The real win is building a repeatable process where pricing decisions are driven by customer willingness to pay, not intuition. If you want, I can share a downloadable ROI calculator and a sample sales playbook I use in these experiments—let me know which segment you’re targeting and I’ll tailor the templates.
