Steps to building a Switch Readiness Score

This checklist walks you through the autonomous email workflow Adam Cain, VP of growth at ElectricityRates.com, built to drive a 200% increase in monthly conversions. By the end, you’ll have a working readiness score, a custom field in your email platform, and event-triggered campaigns that run without manual intervention. Setup time is roughly 3–5 hours depending on how your data is structured.

Before you start

Identify 3–5 data sources you can export or connect to (CRM, website analytics, transaction history, etc.)
Confirm you have a Google Gemini, ChatGPT, or Claude account — any LLM that accepts file uploads works
Have admin access to your email marketing platform (ActiveCampaign recommended — custom objects are required for score storage)
Know your key conversion event: the action that signals a customer is ready to buy, renew, or switch

The workflow

Phase 1: Identify your conversion signals

After this phase, you’ll have: a documented list of 8–12 variables that predict customer intent in your business.

List every data point you collect per customer: contract or subscription end dates, engagement history (email opens/clicks, site visits), past purchase behavior, potential savings or value delta, and any external market signals (price changes, competitor activity)
Identify your top signals: narrow to the 10–12 variables most likely to correlate with conversion. Include at least one time-based signal (contract end, trial expiry, renewal date) and one value signal (potential savings, pricing delta)
Export a sample dataset: pull 500–1,000 customer records with all signals included. Anonymize if needed

“We started by telling Gemini to just have at this and tell us what you think. Look at this data, tell us which customers you think have the highest likelihood to switch to a new energy provider and why.”
Adam Cain

Phase 2: Feed your data to AI broadly

After this phase, you’ll have: AI-surfaced patterns you hadn’t explicitly programmed, and a baseline understanding of what your highest-intent customers look like.

Upload your customer export to Gemini (or your preferred LLM): include all signals, not just the ones you think matter
Ask a broad, open-ended analysis question. Copy and edit the Starter prompt below:
Note every pattern the AI surfaces: pay attention to combinations you hadn’t considered — Cain’s team discovered Gemini could identify customers who should be emailed a specific week because a good rate had just appeared in their area
Ask a follow-up question about scoring:

Open-ended analysis

I’m sharing a dataset of [NUMBER] customers for [BUSINESS TYPE].

Look at this data and tell me: which customers do you think have the highest

likelihood to [YOUR CONVERSION ACTION] and why? What patterns do you see?

Follow-up prompt

Based on the patterns you found, assign each customer a readiness score from 0–100

that reflects their likelihood to [YOUR CONVERSION ACTION].

Explain your scoring logic.

Phase 3: Layer in your business rules

After this phase, you’ll have: a refined scoring formula with specific thresholds you can test and measure.

Review the AI’s scoring logic and identify gaps: what constraints did it miss? (minimum value thresholds, timing requirements, eligibility rules)
Add your first business rule — a timing constraint: customers must meet a deadline or window to be actionable (use Refined prompt V2 pattern)
Add your second business rule — a value threshold: there must be enough upside to justify outreach (use Refined prompt V3 pattern)
Test different threshold combinations: run the model on your dataset with 2–3 variations and compare how many customers qualify at each threshold. Validate against actual past conversions if you have that data

Refined prompt (Version 2 pattern)

Assign a score above [THRESHOLD] only IF [YOUR TIMING CONSTRAINT].

For example: contract end date is less than [X] days away.

Refined prompt (Version 3 pattern)

Assign a score above [HIGHER_THRESHOLD] only IF [TIMING CONSTRAINT]

AND [VALUE CONDITION]. For example: potential savings are [X]% or greater.

Phase 4: Build your readiness score into your platform

After this phase, you’ll have: a live score field in your email platform feeding automated segmentation and campaign triggers.

Create a custom field (or custom object) in ActiveCampaign for your readiness score — store as a number field (0–100). See ActiveCampaign’s custom objects guide
Set up a sync process to write scores from your AI/scoring layer into that field — options include Zapier, Make, direct API, or a scheduled script
Build your segmentation lists: ElectricityRates.com maintains 25+ lists organized by utility area, state, and customer status — create the equivalent for your business (geography, product line, customer tier, status)
Configure campaign triggers based on score thresholds: high scorers (e.g., 85+) enter an active nurture sequence; lower scorers go on a low-frequency drip

Phase 5: Connect external triggers to internal signals

After this phase, you’ll have: fully autonomous campaigns that fire when real-world events match high-intent customers. No manual scheduling required.

List the external events that should trigger a campaign for your business. Examples include price changes, new inventory, competitor news.
Build an event listener or webhook to detect each trigger (utility API, competitor monitoring tool, inventory system, etc.)
Connect the trigger to a filtered segment: when an event fires, filter your list by readiness score first — only high-intent customers get immediate outreach
Configure follow-up cadence logic: after the initial send, adjust follow-up frequency based on opens and clicks. High scorers who engage should receive more frequent follow-up; low engagement drops frequency automatically
Enable predictive sending in ActiveCampaign so each email goes out at the optimal time per recipient — Cain uses this alongside the score to let the platform handle frequency decisions

Quick reference

Total setup time: 3–5 hours (longer if data exports require engineering support)
Tools needed: Google Gemini (or any LLM), ActiveCampaign, a sync mechanism (Zapier / Make / API)
Key output: An autonomous email system that detects external triggers, filters for high-intent customers, and sends personalized campaigns without manual drafting or scheduling

Ready for the full story?

Read The autonomous email engine behind 200% conversion growth, featuring Adam Cain of ElectricityRates.com, published on the AI Lab by ActiveCampaign.

Get the checklist