I ran a tightly scoped, three-week pricing pilot that increased ARPU by 12% for a B2B SaaS product. I’ll walk you through exactly how I set it up, the hypotheses I tested, the metrics I tracked and the operational details that made a short experiment deliver reliable, reusable insight. This is the kind of pilot you can run in a few sprints with minimal engineering overhead and clear commercial impact.

Why I chose a three-week pilot

Three weeks feels short enough to keep momentum and limit exposure, but long enough to collect statistically useful signals from trial-to-paid conversion, upgrade behavior and churn within an active cohort. I wanted a pilot that could be operationalized without a full pricing overhaul: quick to implement, easy to rollback, and focused on behavioral levers rather than complex value-metrics.

My core hypothesis: simple framing changes (anchoring) and a measurable subscription add-on can increase average revenue per user (ARPU) without harming conversion materially.

What I tested

I ran two parallel experiments inside the same pilot window:

  • Anchoring variation: present a high-priced "Enterprise" anchor that makes mid-tier plans appear better value. The actual Enterprise option had limited availability and clear callouts about custom services.
  • Subscription add-on: offer a productized add-on — “Priority Support + Onboarding” — as a monthly or annual subscription that had a clear ROI statement (e.g., "reduce time-to-value by 40%").

I implemented a 2x2 design for a subset of new signups: control, anchor only, add-on only, anchor + add-on. This allowed me to isolate effects and test interaction between anchoring and the add-on.

How I selected the sample and rolled out

I targeted new paid upgrades and upgrade intents from existing trials over a three-week window. Key operational choices:

  • I limited the pilot to users in the UK and EU to avoid tax/pricing complications with other regions.
  • I excluded enterprise sales that require contract negotiation; the pilot focused on self-serve or inside-sales upgrades where the buying path is online and repeatable.
  • I routed users into experiment buckets using our feature-flag tool (LaunchDarkly) so changes were front-end only for the pricing page/modal—no product changes required.

Designing the offers

Designing the anchor and the add-on matters as much as the numbers. I used these principles:

  • Anchor: credible but slightly aspirational. I used a price point ~2-3x the top self-serve plan and included three differentiators: dedicated AM, custom SLAs, and priority roadmap input. The messaging stressed scarcity: "Available to a limited number of customers."
  • Add-on: immediately valuable and easy to explain. I priced the add-on at around 20% of the mid-tier plan's monthly price (or a two-month discount when billed annually). The copy focused on speed and outcomes: "Get live in 7 days, onboarding included."
  • Presentation: emphasize comparison and savings. The pricing modal showed the full list with the anchor on the right and clear per-feature ticks. The add-on had a persistent checkbox during checkout with ROI bullet points.

Metrics I tracked

To evaluate the pilot I tracked a short list of primary and secondary metrics daily and did a formal end-of-pilot analysis.

  • Primary: ARPU per paid account, conversion rate (trial-to-paid), upgrade rate from base to mid/top plan.
  • Secondary: add-on attach rate, churn at 30 days (where available), LTV estimate change, and customer support tickets related to pricing confusion.

Operationally, I used funnel tracking in Segment + Mixpanel for behavioral events and Stripe reports for revenue reconciliation. A simple table in Google Sheets consolidated cohort ARPU and conversion by bucket—this made it easy to spot differences day-to-day.

BucketConversionAttach rateARPU uplift
Control8.7%Baseline
Anchor8.3%+6%
Add-on9.1%18%+9%
Anchor + Add-on8.8%21%+12%

Results in practice

At the end of three weeks, the combined anchor + add-on bucket showed a 12% ARPU lift versus control. A few patterns stood out:

  • The anchor alone improved perceived value of the mid plan (ARPU +6%) but slightly decreased conversion — likely because some price-sensitive users self-selected out of upgrading when faced with a prominent high price. This was an expected trade-off.
  • The add-on alone increased ARPU by around 9% and slightly increased conversion: the add-on framed an instant gain (faster onboarding), which nudged trial users to convert earlier.
  • When combined, anchoring and the add-on had an additive effect: the anchor made the mid plan feel like a bargain, and the add-on gave a clear upsell path. That’s where we saw the +12% ARPU.

Statistical confidence and caveats

I pre-specified minimum detectable effect sizes and sample sizes before launch. We deliberately ran the pilot only on cohorts with sufficient traffic to reach ~80% power for a 7–10% ARPU uplift. For conversion differences that were smaller, we noted directional trends but treated them as hypothesis-generating rather than conclusive.

Key caveats:

  • Short pilots are vulnerable to timing effects (e.g., marketing campaigns during the window). I flagged concurrent initiatives and adjusted the analysis if a funnel spike overlapped with outbound campaigns.
  • We only observed early churn; longer-term effects on retention and expansion require follow-up cohorts and a plan to monitor 90–180 day cohorts.

Operational lessons — what made the pilot succeed

Three operational choices made the pilot fast and low-risk:

  • Front-end only changes: I implemented the anchor and add-on via UI copy and checkout toggles. No product changes meant rapid rollback and low engineering cost.
  • Clear segmentation: limiting to self-serve and inside-sales flows avoided complicating enterprise negotiations and kept the experiment clean.
  • Daily monitoring and a kill-switch: we monitored conversion and support volume daily. If conversion dropped >30% vs baseline or we saw spike in pricing tickets, we had a kill-switch to revert flags immediately.

Common mistakes to avoid

  • Changing multiple pricing levers at once without orthogonal buckets — this creates attribution problems.
  • Overcomplicating the anchor — if it’s not credible or clearly differentiated, it backfires and erodes trust.
  • Not defining success criteria beforehand — you need specific thresholds for ARPU uplift, conversion drag tolerance, and add-on attach rate.

Next steps after the pilot

Based on the pilot I pushed three immediate follow-ups:

  • Roll the add-on live with A/B rollout to all new signups, while continuing to track 90-day retention impact.
  • Refine the anchor messaging (tighter copy, customer case example) and test an alternative anchor with productized professional services to validate price sensitivity.
  • Instrument experiments to measure cohort LTV — not only ARPU — so we can quantify long-term effects of the add-on on retention and expansion.

Running a successful three-week pricing pilot comes down to tight hypotheses, simple implementation and disciplined monitoring. You don’t need to reprice your whole product to learn how buyers respond — a credible anchor plus a clearly valuable subscription add-on, tested in a controlled window, can give you a fast, measurable boost to ARPU and a roadmap for scaling pricing changes.