Ready for AI in Pricing? Your First 90 Days.
A simple plan to go from readiness to first results.

From score to action
In October, we asked a simple question: Is your organization ready for AI in pricing? Seven dimensions, a score out of 35, and three bands: AI-ready, on your way, or start with a quick win. The scorecard is now also a free online Readiness Check.
A score tells you where you stand. It doesn't tell you what to do on Monday morning.
This post helps with exactly that. It sets out a simple 90-day plan to get started, and shows how to adapt it to your score.
The most common mistake: starting with the technology
Many AI pricing initiatives start with a question about the technology. Which AI model? Which platform? Which provider or expert?
It feels like progress, and choosing the technology looks like the right first action.
It isn't.
Technology alone can't tell you which decision to improve, what "better" means for your business, or who will act on the result. Without those answers, even a good AI model produces recommendations nobody uses.
So start with the decision, not the technology. Which pricing decision do you want to make better? List prices for one product line, discounts for one customer segment, or renewal prices for key accounts are all good candidates.
One decision, improved and proven, teaches you more than a broad program that never leaves the planning stage.
Your first 90 days
The plan has three phases of about 30 days each. Each one ends with something concrete.
Days 1–30: Foundations
Pick one pricing decision, and map how it's made today and who owns it. Who sets the price, who approves exceptions, and who answers for the result? Many teams find the answers are less clear than they expected.
Set one or two business objectives and the real constraints. Margin, revenue, volume, or the right balance between them. Constraints are the limits any new price must respect, such as price corridors, minimum profitability, or capacity.
Audit the data that the decision needs and identify the gaps. Only that data. Waiting for perfect data is one of the most common ways AI pricing initiatives stall.
Educate stakeholders on what AI can and cannot do. It can find patterns across thousands of products, customers, and deals that no one can see by hand. It can't set your strategy or replace commercial judgment. Realistic expectations now prevent disappointment later.
Consider outside help for the first step. Few teams have done this before. An experienced partner shortens the learning curve, while your team brings the market knowledge.
Days 31–60: Proof of concept
Choose a few products, categories, or markets. Small enough to manage, large enough to matter.
Choose an expert or a tool, and favor simplicity and speed to value. For a first test, what matters is how quickly you get a credible answer, not which solution you'll use for years. Our free Pricing Solution Advisor can help you weigh the options.
Agree success measures before you start. Decide what result would count as a success, and what would make you stop. Margin uplift, growth, accuracy — whatever matters.
Hold out a period the model never sees. Keep back recent data, then check the model's predictions against what actually happened. It's the simplest way to know whether you can trust it.
Compare with today's pricing, document lessons, and decide go/no-go/improve. Today's approach is your baseline. Write down what worked and what didn't, report the results to leadership in commercial terms, and decide whether to scale. A clear "not yet" or "needs improvement" is also a good outcome: you learned it cheaply.
Days 61–90: Scale
Expand to more segments, products, and use cases. Build on what the proof of concept showed, one step at a time.
Decide if periodic AI analysis is enough, or if frequent changes and volumes need software and integration. For many companies, a focused AI analysis repeated at regular intervals delivers the value they need. Deploying software makes more sense when pricing is highly dynamic, volumes are very large, or integrations are essential.
Standardize the AI–human workflow, including who owns exceptions. Recommendations only create value when they are used. Decide who reviews them, and who has the final word when a recommendation and a salesperson's judgment differ.
Train the team to interpret the output. People need to understand, question, and explain the recommendations, including to customers.
Finalize the plan for further expansion after the first 90 days. Which decisions come next, and what must improve to get there?
At day 90, you should have evidence of value from your own business, a team that knows how to use AI-driven pricing insights, and a plan for what comes next.
How the plan changes with your score
The three phases stay the same. What changes is how ambitious each one is and where you should focus more.

Under 20. Your first 90 days are not about a big AI project. They are about early, visible wins that build data, confidence, and internal support.
20 to 29. This is the most common position. Don't wait for every gap to close. Fix only what your chosen decision needs, and run a first test.
30 or more. Your risk is moving too slowly or too broadly. Choose a decision with real commercial value, set success measures up front, and plan how you will scale.
Let your lowest scores shape the plan
As we said in October, the total is only half the story. Your weakest dimensions tell you where to put extra effort. For instance:
Strategy and ownership: spend more of days 1–30 agreeing the objective and naming the owner. Everything else depends on it.
Data: fix only the fields your chosen decision needs, not every source.
Experimentation: agree the testing rules before day 31, including scope, risk limits, approvals, and success measures.
Executive support: plan the leadership review from day one, and build it around a problem leadership already recognizes, such as margin pressure or discount leakage.
What 90 days can and can't do
Ninety days won't transform your pricing. That isn't the goal.
The goal is to replace a debate about AI with evidence from your own business. Once you have that, the next decision gets much easier. AI in pricing is a capability that grows over time, not a one-off project.
Work through it in Amsterdam
On 2 December, we'll be leading a full-day workshop at PPS profitABLE26 in Amsterdam: Where AI Price Optimization Fails and Succeeds: Real-World Lessons. Participants assess their own readiness and scope a 90-day pilot for a pricing decision of their own, alongside framing, data, and validation.
Your next step
Your score tells you where you stand. The next 90 days decide what happens next.
Haven't scored yet? Take the free Readiness Check. It takes 1–3 minutes, with no sign-up.
Not sure which approach fits your decision? Try the free Pricing Solution Advisor.
Curious what this would look like with your data?
AI-driven pricing insights to unlock hidden revenue & margin
Using your data. Minimal involvement. No implementation needed.



