Automated Pricing Optimization Engine

Introduced in v1.0.43 · Category: Sales & Revenue · Severity: High

Overview

The Automated Pricing Optimization Engine closes a critical mission gap: pricing optimization is now a fully autonomous loop, not an advisory one. The revenue agent no longer stops at generating suggestions — it evaluates, proposes, tests, and executes price changes end-to-end, including syncing updates to Stripe.

Background

Before v1.0.43, the revenue agent produced pricing suggestions (e.g. pricing_model, upsell recommendations) and persisted them to the revenueSuggestions table. That table was effectively read-only — suggestions sat idle until a human operator clicked Add to Queue. This made pricing optimization advisory rather than autonomous, which is inconsistent with the platform's core mission of zero-human-bottleneck revenue operations.

How It Works

1. Suggestion Evaluation

The revenue agent continues to generate pricing_model, upsell, and related suggestions as before. These are written to revenueSuggestions. The new actioning pipeline picks up unprocessed suggestions and evaluates them against current subscription analytics, churn signals, and revenue targets.

2. Price-Change Proposal Generation

Based on evaluated suggestions, the engine generates concrete price-change proposals. Each proposal includes:

• Target product / price tier
• Proposed new price point
• Justification derived from revenue signal data
• Confidence score
• Proposed rollout strategy (immediate, A/B, or staged)

3. A/B Test Scaffolding

For proposals that fall below a confidence threshold, the engine automatically scaffolds a pricing A/B test rather than applying the change directly. This includes:

• Splitting eligible new subscribers into control and variant cohorts
• Defining the success metric and minimum sample size
• Scheduling an evaluation checkpoint

No manual configuration is required to initiate a pricing experiment.

4. Stripe Price-Update Workflow

Once a proposal is approved (autonomously, based on confidence thresholds and guardrails), the engine triggers a Stripe price-update workflow that:

• Creates the new Stripe Price object against the existing Product
• Updates active subscription references where appropriate
• Records the applied change back to the revenueSuggestions row, marking it as actioned

5. Observability

Every stage of the pipeline — evaluation, proposal, A/B scaffold, and Stripe update — is logged and surfaced in the Revenue dashboard. Operators can inspect what changed, why, and when, without needing to intervene.

Key Components

Component	Role
Revenue Agent	Generates raw pricing suggestions
Actioning Pipeline	Consumes revenueSuggestions, drives proposals
A/B Test Scaffolder	Creates experiments for low-confidence proposals
Stripe Price-Update Workflow	Executes approved price changes in Stripe
Revenue Dashboard	Observability for all pipeline activity

Guardrails

The engine operates within configurable guardrails to prevent runaway price changes:

• Maximum price delta — proposals exceeding a configured percentage change are held for review
• Minimum evaluation window — the engine will not re-price the same product more frequently than a defined interval
• Confidence threshold — proposals below the threshold are routed to A/B testing rather than direct application

Mission Alignment

The SaaS Factory mission calls out automated pricing optimization as a core Sales & Revenue capability. This engine fulfills that requirement: from suggestion to live Stripe price, no human is required in the loop.