Build an AI Price Increase Playbook: Six Chained Prompts for Raising Prices Without Losing Customers

Build an AI Price Increase Playbook: Raise Prices Without Losing Customers

TLDR: This tutorial walks through a six-prompt chained workflow for building a complete price increase strategy using AI. It covers value auditing, price setting, customer churn risk mapping, announcement email writing, objection handling, and a 30-day rollout plan with rollback triggers.

Summary: The framing here is good: most price increases fail in the delivery, not the number. You can pick a perfectly defensible new price and still lose customers at an alarming rate because the announcement was poorly written, the timing was wrong, or you treated every customer segment identically when they have very different risk profiles. The playbook structure addresses this by separating the strategic work, what is the right price and who is at risk of leaving, from the communication work, how do you tell them in a way they will respect.

The six-prompt chain is structured as a pipeline where each step feeds into the next. The first prompt builds a value delta report, an inventory of everything that has changed since your last price change, translated into customer outcomes rather than features. The second prompt uses that report as context for setting the new price with three scenarios: safe, recommended, and aggressive, each with revenue math. The third maps your customer base by churn risk and determines who gets grandfathered versus who receives the full increase. Then you write segmented announcement emails, prepare objection scripts for every pushback type you can anticipate, and build a 30-day rollout timeline with explicit rollback triggers.

The prompts themselves are written in a way that asks the AI to be honest about weak spots. The value audit prompt specifically says: if the value case is weak anywhere, flag it before recommending an increase. That is the right instinct. A lot of AI-assisted business advice skips the critical examination and goes straight to optimistic outputs. Telling the model to surface uncomfortable truths makes the output more defensible.

What the article does not address is the recursive problem of anchoring on competitor pricing. The benchmark-based pricing prompt asks you to compare against alternatives and competitors, which is fine, but if your competitors are also underpriced, you are anchoring your new price to a market floor that may be systematically wrong. Willingness-to-pay research, actual surveys or conjoint analysis, is the more defensible foundation for pricing decisions. The AI can help structure that research too, but the tutorial does not go there.

There is also an assumption that customers will receive these communications through email in isolation. In practice, price increase announcements land in inboxes alongside competitor offers, support tickets, and usage notifications. The emotional context at the moment of reading matters, and no prompt chain can fully account for that. The playbook is a starting structure, not a guarantee.

Key takeaways:

• Effective price increases require separating strategy from communication: get the number right first, then tailor the message by customer segment
• Chaining six sequential prompts creates a coherent, internally consistent playbook where each step uses the output of the previous one as context
• Asking the AI to flag weak spots in your value case before committing to a number produces more honest and defensible outputs
• Competitor benchmarking anchors to market floors that may themselves be systematically low; willingness-to-pay research is more reliable for the actual number

Why do I care: Prompt engineering for business decisions is where I see AI doing genuinely useful work that was previously locked behind expensive consultants. The chained prompt structure here is a good pattern for any multi-step decision process where context accumulates across steps. What I find underexplored is calibration: how do you know when the AI's framing of your value case is accurate versus optimistic? The honest-flagging instruction helps, but building in a reality check step, maybe structured differently from a customer's perspective rather than the seller's, would make this more robust.

Tutorial: Build an AI Price Increase Playbook (Raise Prices Without Losing Customers)