Product guide / Strategy
your bets and assumptions, on the record
Agents propose. People decide. The record keeps both.
The company’s strategy spine — bets, charters, hypotheses, signals, decisions, experiments — where an agent can read everything, propose anything, and write nothing without a human signing for it.
| Bet | Stage | Status | Thesis | Confidence | Why now | Evidence | Stop criteria | Upside | Risk | Owner | Next review |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Vertical AI Reports | shape | committed | AI reports cut analyst time by 70%. | 4 out of 5 | LLM costs fell 80% in 12 months. | Pilot users saved 6 hrs/week on average. | Adoption <20% after 60-day trial ends. | Adds $2M ARR from premium tier. | Accuracy errors erode user trust fast. | Head of Product | 2025-09-01 |
| Self-Serve Onboarding V2 | decide | committed | Self-serve reduces CAC by 40%. | 5 out of 5 | Sales team at capacity; pipeline stalling. | Competitor PLG benchmarks show 35% CAC drop. | Trial-to-paid rate stays below 12% at 90 days. | Unlocks SMB segment worth $5M ARR. | Complex setup drives early churn spike. | VP Growth | 2025-08-15 |
| Data Connector Marketplace | capture | proposed | Native connectors increase retention. | 3 out of 5 | Zapier costs eat 8% of SMB margins. | Support tickets cite integration gaps weekly. | Fewer than 5 partners join in first quarter. | Platform lock-in lifts NRR above 120%. | Partner certification delays stretch timeline. | Head of Partnerships | 2025-10-01 |
| Enterprise SSO & SCIM | decide | committed | SSO unblocks Fortune 1000 procurement. | 5 out of 5 | Three $200K deals stalled on SSO gap. | Security reviews list SSO as blocker #1. | No enterprise close attributable within 90 days. | Enables $1.5M ACV deals in pipeline. | SAML edge cases inflate engineering scope. | Head of Engineering | 2025-07-31 |
| Annual Plan Incentive Push | review | validated | Annual discounts improve cash flow 30%. | 5 out of 5 | Q2 burn demands 3 months of runway buffer. | Last push converted 42% of eligible monthly users. | Annual attach rate drops below 25% in cohort. | Adds $900K to cash reserves by Q3. | Monthly churn risk re-emerges at renewal. | VP Finance | 2025-08-01 |
| Embedded Analytics Widget | shape | proposed | Embeds create viral B2B2C growth loop. | 3 out of 5 | Partner portals hungry for white-label data. | Three partners requested embed API in beta. | Fewer than 500 embed views per month by Q4. | Opens $3M white-label licensing market. | Maintenance overhead forks core product. | Head of Product | 2025-10-15 |
| Mid-Market CSM Playbook | shape | committed | Structured CSM reduces mid-market churn. | 4 out of 5 | Mid-market NRR dropped 8pts last quarter. | Churned accounts lacked QBR touchpoints. | NRR stays flat after two full playbook cycles. | Recovering NRR adds $1.1M retained ARR. | CSM headcount cost outpaces saved ARR. | VP Customer Success | 2025-09-15 |
| Usage-Based Pricing Pilot | capture | proposed | Usage pricing lowers entry barrier for SMBs. | 2 out of 5 | SMB conversion dropped 15% vs last year. | Exit surveys cite price as top barrier. | Pilot ARPU falls below $120 after 6 months. | Expands SAM by 3x into micro-SMB tier. | Revenue predictability drops for investors. | VP Growth | 2025-11-01 |
| In-App Collaboration Layer | capture | proposed | Comments add multi-seat stickiness. | 3 out of 5 | Teams share exports outside the platform. | Usage logs show 60% of shares are exports. | Seat expansion stays flat 90 days post-launch. | Seat-based upsell lifts ARPU by 25%. | Real-time sync adds infra complexity. | Head of Engineering | 2025-10-01 |
| APAC Localization Sprint | review | stopped | Localization unlocks Japan & Korea markets. | 2 out of 5 | Partners pushed back timelines by 6 months. | No signed LOI from APAC prospects in H1. | Zero qualified pipeline after 4-month push. | APAC TAM estimated at $8M in year three. | Translation costs exceed projected revenue. | Head of Partnerships | 2025-07-15 |
| Predictive Churn Alerts | review | validated | Early churn signals let CSM intervene first. | 4 out of 5 | Model training data now spans 18 months. | Beta alert saved two $80K accounts last month. | False positive rate exceeds 30% in production. | Saves estimated $600K ARR annually at scale. | Model drift needs continuous retraining budget. | Head of Data | 2025-09-01 |
| Partner Referral Program | decide | committed | Referrals drop blended CAC below $1,800. | 4 out of 5 | Paid CAC rose 22% quarter-over-quarter. | Word-of-mouth already drives 18% of signups. | Referred deals <10% of new ARR after one year. | Referral channel could fund 30% of pipeline. | Partner disputes require legal overhead. | VP Growth | 2025-08-30 |
Where this helps
Strategy documents go stale because keeping them current is nobody’s job. Handing that job to an agent is worse — unless the agent can only propose, and a person has to sign.
The question it puts in front of you
Do we accept this proposed change to what the company believes?
Where judgment belongs
Every strategy record is approved, edited, or rejected by a person, and what you changed while reviewing is recorded next to what was originally proposed.
What gets prepared for you
Agents read the whole spine freely and route every write through a proposal, so a suggestion arrives as a reviewable diff instead of a silent edit.
Strategy & decisions
How a proposal becomes a record.
Capture six typed records — bet, charter, hypothesis, signal, decision, experiment — on governed company lists.
Review each proposal against the original values, and against anything you edited mid-review.
Approve, reject with a reason, or let it expire — every resolution keeps its receipt.
Open the impact lens before retiring a bet, to see what depends on it.
Strategy in the real world
If this bet dies, what else just became suspect?
An agent can read every bet and assumption you hold, and change none of them. It proposes; the Workbench shows you the proposal next to what was there before, notes anything you edited while reviewing, and keeps a receipt either way.
The impact view reads the other direction. Pick a bet and it shows what depends on it, and what depends on that. Retiring a bet stops being a line-item edit and becomes a question the team has to answer out loud.
What holds up
What it does well, and what it was never meant to do.
Constraints are a design position, not an apology. Everything on the left ships today; everything on the right we decided against.
What you can rely on in Strategy
- Six record types: bet, charter, hypothesis, signal, decision, experiment.
- Agents can read all of it and change none of it without your approval.
- Each proposal is pending, approved, rejected, or expired, and keeps a receipt either way.
- The impact view only ever shows you what you already have access to.
What we chose not to build
- An agent cannot write a strategy record directly. That is the design, and there is no bypass.
- The impact view is a list today, not a visual graph.
- The six types are fixed. This is a strategy record, not a general-purpose database.
- Approving a proposal records a decision. It does not execute one.
Works with
Where Strategy hands the work on.
Decisions rarely stop where a product does. These are the guides that pick the work up next.
For your company
See how Operalta fits your company.
Bring one current decision or process. We will map it together.