Pricing research and packaging plan

Scope note: The only facts I have are the three you gave me. Everything else below is labeled as a hypothesis or an unknown. Any numbers or tier names are placeholders to test, not recommendations to ship.

1. Evidence vs. hypotheses

Evidence (what we know)

Hypotheses (untested)

Unknowns to fill first

Current conversion, ARPU, logo and revenue churn (monthly and by segment), seat distribution per account, share of accounts that are agencies vs. mid-market, how many lost deals cite approvals, and the competitor price set. These come from internal data pulls, which are cheap, so do them before any survey.

2. Value-metric recommendation

Keep per internal user as the primary metric. Make client reviewers and approvers free and unlimited. Gate the approval workflow by tier.

Why:

• It matches E2. Prospects already understand per-seat. Changing the metric adds a new thing to explain just when we're adding a feature.
• It protects the approval feature's value. Approval is more useful the more clients are involved. Charging per client reviewer would push agencies to share logins or avoid inviting clients (H3).
• It scales with value for collaboration. More internal collaborators means more work flowing through the tool.

Challenger to test, not adopt yet: a secondary limit on active client workspaces/projects (H7). Agencies with many clients get more value from approvals than agencies with few, and seats alone may not capture that. Test it in research (Section 3) before deciding. A hybrid metric is more complex, so it has to show clear value over the simpler approach.

Metrics I'd avoid: per-approval or per-client-reviewer pricing (penalizes the behavior we want), and flat-fee pricing (loses the scaling link to team size).

3. Research plan

Phase 0: Internal data (weeks 1–2)

• Pull the unknowns above.
• Tag lost, stalled and won deals that mention approvals.
• Cluster customers by seat count, client-facing usage, and segment.

Phase 1: Qualitative (weeks 2–5)

• 12–20 interviews: current customers, recent prospects, churned accounts. Balance agencies and mid-market.
• Probe: what they do for client approval today, what it costs them (tools, time, delays), who approves purchases, and monthly vs. annual preference and why.
• Ask about the approval workflow before mentioning price, to avoid anchoring.

Phase 2: Quantitative (weeks 5–8)

• MaxDiff on features (approvals, guest access, branding, SSO, audit log, integrations, support) to inform tier packaging.
• Van Westendorp or Gabor-Granger on the per-seat price, run separately per segment. Caveat: these are stated-preference methods and tend to overstate acceptable prices, so use them to find ranges, not exact numbers.
• Add a tier-choice exercise (conceptual good/better/best with the approval feature in different tiers) if you can reach enough respondents.
• Sample size depends on whom you can reach. If you can only get dozens of respondents per segment, treat results as directional.

Phase 3: Live tests (weeks 8–20+) — see Section 5.

4. Packaging proposal (to be validated)

Starting structure, contingent on MaxDiff results:

Open packaging questions:

• Should basic approvals be in every tier (to answer prospects) with advanced approvals (multi-stage, audit trail, branded portals) gated higher? I'd lean this way, because gating the core of what prospects asked for may read as a price increase on the existing product.
• Does the existing $29 plan become Core, or stay as the middle tier? That depends on how much existing customers would value approvals. Keep existing customers' current price and features unchanged while testing (grandfathering as a transition, not a permanent promise).

5. Annual option test

Design

• New customers only. Don't touch existing accounts in the first test.
• Randomize at the visitor/account level across arms:
  • A: Monthly only (control)
  • B: Monthly plus annual at a discount in the typical range (about 15–20%, e.g. "two months free")
  • C (optional): Monthly plus annual at a smaller discount or with a non-price incentive (e.g., onboarding support), to see how much discount is actually needed
• Show a monthly/annual toggle with the monthly price visible so it can be compared.

Metrics

• Primary: annual take-rate, signup-to-paid conversion, and 90-day and 12-month retention. Annual cohorts need a full year to read reliably, so early reads are only leading indicators.
• Secondary: cash collected upfront, refund and downgrade requests, support tickets, discount-driven mix shift.
• Segment cuts: agency vs. mid-market (H4).

Cautions

• A discount lowers the effective monthly price. Annual should be judged on retention, cash flow and payback, not assumed to be a revenue gain.
• Higher annual take-rate among customers who would have stayed anyway is a cost, not a win (H5). Compare against the control carefully.
• Refund and renewal terms need to be defined before launch.

6. Rollout and guardrails

1. Test packaging and annual on new signups first.
2. If results hold, move 5–10% of eligible existing accounts to new plans or offers, and watch churn and support volume for a billing cycle.
3. Expand in waves. Announce changes with a clear reason and generous notice.
4. Guardrails: pre-agree stop thresholds (e.g., conversion drop, churn increase, support spike) before launch.
5. Don't set price points from competitor prices alone. Use them as a reference range.

7. Decision criteria and deliverables

• Decision rules written up front: what result in MaxDiff and WTP supports gating approvals; what annual take-rate and retention pattern supports keeping or changing the annual discount.
• Outputs: a segment-level WTP summary, a feature-priority ranking, a recommended tier structure and price range, an annual-test readout, and a list of what remains uncertain.
• What this plan won't give you: a forecast of revenue change. Results will show relative differences in conversion, retention and mix, with confidence limits. They will not predict outcomes after rollout.

8. Questions for you

1. What are current conversion, churn and seat-count distribution? These shape the sample sizes and test duration.
2. What's the agency vs. mid-market mix, and is the motion self-serve, sales-led or hybrid?
3. Are approvals already built or on the roadmap? That determines whether we can test with a real feature or a concept.
4. Who are the main alternatives prospects mention?