Should You Use Competitor Prices? A Decision-Making Framework for AI-Driven Pricing
- FutureUP
- 16 hours ago
- 9 min read
When competitor intelligence adds context, when it adds a true signal — and when it makes AI-driven pricing decisions worse.
Competitor prices seem like an obvious input into AI-driven pricing decisions, especially Price Optimization. After all, how can you determine the right price without knowing what the market is charging?
Or can you?
A recent case
A recent FutureUP case showed why the answer is not so simple.
The competitor information was close to a best-case scenario: the offering was straightforward, there were only a few relevant competitors, their prices were public and reasonably comparable, and the data coverage was good.
The initial AI pricing model — built without competitor prices — performed very well.
Then competitor prices were added.
Accuracy fell. A lot.
The issue was not simply bad or incomplete data. The competitor information was sufficiently clean and available. It simply did not add useful predictive information to what the model already knew.
But the case raised a broader question:
How should a company decide whether competitor pricing data is worth identifying, collecting, and using?
Knowing Your Competitors Is Not the Same as Following Them
Competitor intelligence can add valuable context. It can help you understand market positioning, identify unusual price movements, challenge internal assumptions, and explain changes in customer behavior.
But following competitors blindly is something entirely different. It assumes that competitors:
Understand the market better than you.
Have the same objectives and economics.
Serve the same customers in the same way.
Price rationally.
Make all the right decisions.
Those assumptions may be wrong.
Your competitors may be reducing their price because their product is struggling. Or increasing their prices because they improved the offer — or simply because they want higher margins. Or they may be reacting to inflation, cost increases, or simply following another competitor, even your own company.
Knowing your competitors is market intelligence. Following them without understanding why they price as they do is outsourcing your pricing strategy.
So, how can we determine whether competitor information adds useful context or genuinely new evidence to your pricing decisions?
Six Questions to Ask Before Using Competitor Prices
The following six questions can help determine whether competitor prices are important and feasible enough to use in AI-driven pricing decisions.
Importance
The first three questions define the importance of incorporating competitor intelligence:
Purpose: What Will You Use the Information For?
Relevance: Do Competitor Prices Really Matter to Your Customers?
Incremental signal: Does the Data Tell You Anything New?
Feasibility
The next three questions assess the feasibility of obtaining competitor intelligence that is actually usable:
Identification: Who and What Should You Compare?
Comparability: Are You Comparing Like with Like?
Availability: Can You Obtain the Data Properly?
These two dimensions — importance and feasibility — should be considered separately. A competitor price that is easy to collect is not necessarily important, while highly relevant information may still be too difficult or expensive to obtain systematically.
Let's explore each one in more detail, how to score it, and how to use the results to make a decision.
Importance
1. Purpose: What Will You Use the Information For?
Before collecting competitor prices, clarify why you need them. Possible purposes include:
Understanding your market position.
Monitoring major market changes.
Preparing for customer negotiations.
Predicting demand or sales performance.
Supporting pricing strategy.
Supporting price optimization.
These purposes are not equivalent. Competitor data that is useful in a strategy discussion may not improve a predictive pricing model. Information that helps a salesperson prepare for a negotiation may not be sufficiently consistent to use quantitatively at scale.
Ask:
What specific decision will the information support?
Will it be used as context, a benchmark, or a model input?
Evaluate Purpose based on whether you want to use competitor prices as a structured input for pricing guidance, rather than simply as a general benchmark for context.
2. Relevance: Do Competitor Prices Really Matter to Your Customers?
Competitor prices are more relevant in some environments than others.
Where it matters the most
They are more likely to influence demand when:
Prices are visible.
Products are directly comparable.
Customers actively compare alternatives.
Switching is easy.
Purchasing cycles are short.
Price is the strongest buying criterion.
Competitors change prices frequently.
This is common in many e-commerce, retail, and fast-moving B2C markets.
Where it could be less important
The situation may be very different in other markets, especially B2B.
Prices may be confidential, individually negotiated, or embedded in broader agreements.
Customers may have only a vague or limited knowledge of alternative offerings and their prices.
Price may not be the strongest buying criterion, and customers instead evaluate:
The value they receive.
Product performance.
Reliability and risk.
Service and support.
Existing relationships.
Switching costs.
Contract terms.
The total cost of ownership.
Price positioning matters
Your own price positioning also matters.
A company competing primarily on price should generally pay closer attention to competitor pricing than a strongly differentiated or premium business.
A premium company still needs to understand the market. But its central pricing question is not necessarily:
"How close are we to the competitor average?"
It may be:
"Are we creating enough additional value to justify our position?"
Evaluate relevance based on the actual purchasing process — not on the assumption that customers always know and respond to competitor prices.
3. Incremental Signal: Does the Data Tell You Anything New?
Competitor prices may look relevant but still add little new information.
Why?
Correlation between your competitors' prices and your sales performance does not necessarily mean that competitor prices are driving demand. Both may simply be reacting to the same underlying market conditions:
Inflation.
Exchange rates.
Raw-material costs.
Energy prices.
Supply constraints.
Regulation.
Seasonality.
Changes in market demand.
In these cases, a market or macroeconomic indicator may be more informative than the competitor price that resulted from it.
Competitors may also follow:
The market leader.
One another.
Industry conventions.
Or even your own company.
Their prices may therefore contain little independent information.
Where competitor actions materially affect your demand, part of that effect may already appear in your own sales data through:
Volumes.
Win rates.
Customer churn.
Discounting.
Changes across segments or markets.
Adding competitor prices may then duplicate information the model already has.
Worse, it may add instability and noise.
Evaluate Incremental Signal based on how likely competitor intelligence is to contain differentiated information beyond your existing sales and market data.
Feasibility
4. Identification: Who and What Should You Compare?
Identifying competitor prices sounds easy until you try to do it properly.
First, who is the competitor?
Different competitors may be relevant for:
Different products.
Different customer segments.
Different geographies.
Different use cases.
New versus existing customers.
Premium versus value-oriented buyers.
Then there is the benchmark itself.
Should you use:
The market leader or your closest competitor?
Your three largest competitors?
All available competitors?
And what will be the price you use:
The lowest price?
The market average or median?
A weighted index based on market share, product similarity, or strategic importance?
You must also decide which price you are comparing, based on their availability and relevance:
List price.
Promotional price.
Observed online price.
Transaction price.
Total contract value.
Evaluate identification based on how easy it is to identify the right competitor offerings and what is the right price to track.
5. Comparability: Are You Comparing Like with Like?
A price comparison may need to account for differences, such as:
Product specifications and features.
Quality and ratings.
Warranty and SLAs.
Delivery times.
Pack size.
Bundled products.
Unit of measurement.
Contract duration.
For example, one competitor may charge a higher price but include installation, service and a longer warranty. Another may charge less because it offers fewer features, lower quality or more restrictive terms.
This leads to a critical principle:
A competitor’s price without its value proposition is not a benchmark. It is simply a number.
Evaluate Comparability based on how easy it is to compare "apples with apples" in terms of value proposition and pricing.
6. Availability: Can You Obtain the Data Properly?
Your own sales data is usually created as part of operating the business. Competitor data often has to be found, purchased, requested, scraped, inferred, or manually maintained.
This makes it more difficult to collect consistently.
Potential challenges include:
Prices are confidential.
Public prices differ from actual transaction prices.
Information is missing for important products or markets.
Competitors use different units or commercial structures.
Promotions and temporary discounts distort the benchmark.
Prices change before the database is updated.
Data sources disappear or change format.
Manual collection becomes too slow or expensive.
Different teams collect the information differently.
Recency is also critical.
A competitor price observed six months ago may still be relevant in a stable B2B market but almost useless in a dynamic online category.
Ask:
How recent must the benchmark be?
How frequently must it be refreshed?
Can it be collected at the required pace?
Is the source reliable?
Can the collection process be repeated?
What will the ongoing cost be?
There is also a legal and compliance question — are you allowed to collect and use competitor intelligence in the proposed way? Depending on the jurisdiction, industry, source, and collection method, different restrictions may apply.
Evaluate Availability based on how easy it is to collect the required competitor information at the right frequency, legally, and at reasonable cost and effort.
Getting Your Scores
Rather than collapsing all six factors into one average, I suggest looking at the two separate dimensions.
Importance
How important is competitor pricing information to the pricing decision you are trying to make?
Area | The question to answer about competitor intelligence* |
How strong is the reason for using it as a structured model input for pricing guidance? | |
How strong is its relevance to your customers' decision-making? | |
How strong is your competitor intelligence differentiation against existing sales or market information? |
* Score: Low (1 - Very weak, 2 - Weak), Average (3), High (4 - Strong, 5 - Very strong)
Feasibility
How realistically can you obtain competitor pricing information that is usable and comparable?
Area | The question to answer about competitor intelligence* |
How easy is it to identify the right competitor offerings and prices to monitor? | |
How easy is it to compare competitor value proposition and price against yours? | |
How easy is it to collect the necessary information? |
* Score: Low (1 - Very difficult, 2 - Difficult), Average (3), High (4 - Easy, 5 - Very easy)
The scores are not intended to decide automatically whether competitor prices belong in your pricing model. They help determine whether the information is important and feasible enough to justify the effort of collecting and testing it.
How to Interpret Your Score
Calculate the average separately for each dimension and use the matrix below to determine whether competitor pricing data is worth pursuing for your AI-driven pricing model.
Feasibility | |||
Importance | Low | Average | High |
Low | No | No | No |
Average | No | Maybe | Yes |
High | Maybe | Yes | Yes |
No Do not make competitor data a priority
Importance is either too weak or not high enough to justify substantial investment. Use competitor information selectively for qualitative context where it is available. But do not make it a prerequisite for pricing analysis, modeling, or decision-making.
Focus first on:
Your own sales and transaction data.
Pricing and discount behavior.
Value drivers.
Customer and product characteristics.
Market and macroeconomic indicators.
Maybe Use selectively
There may be value, but the case for broad competitor-data investment needs to be justified.
Start with a focused test:
One product group.
One market.
A small number of directly comparable competitors.
One clearly defined decision.
A limited collection period.
Then assess whether the information changes the conclusion or improves the result.
Yes Worth collecting and testing
The data is likely to be commercially relevant, sufficiently available and reasonably comparable. It may justify systematic collection.
But a high score does not mean that competitor prices automatically belong in the predictive model. It only means they have earned the right to be tested.
The Final Test: A Practical Rule of Thumb
If you get No or Maybe
If competitor pricing data is difficult, expensive, or unreliable to obtain — as in many B2B markets — or its importance is lower, start without it.
Build the strongest model you can from your own sales, customer, product, and market or macro information.
Test its accuracy against unseen data not used in the training of the model.
If sufficient, keep it and stop any effort for collecting competitor information for the model's sake.
If you get Yes
If competitor data is readily available and scores highly across the framework, collect it and treat it as another candidate factor.
Then compare the predictive model's performance with and without it.
Keep it only when it materially improves the accuracy of the model.
What if competitor prices don't make it onto the predictive pricing model?
Excluding competitor prices from a predictive model does not mean we should never use them anywhere.
Competitor intelligence may still help:
Provide context for decision-makers.
Review strategic positioning.
Identify emerging trends.
Prepare customer discussions.
Identify case exceptions.
Challenge recommendations.
Highlight areas requiring additional investigation.
The information can remain valuable qualitatively even when it adds no predictive power quantitatively.
That is an important distinction.
Competitor intelligence does not have to become the pricing anchor — or a model variable — to be useful.
Conclusion
Companies often assume that more data must lead to better pricing decisions.
It does not.
Competitor prices can add context, reveal market movements, and support strategic discussions.
They can also be difficult to identify, expensive to collect, misleading to compare, legally sensitive, duplicative, or irrelevant to how customers make decisions.
And even when the data looks ideal, it may still reduce predictive accuracy.
The goal is not to build the pricing model with the most information. It is to build the model that makes the most reliable decisions.
So before investing heavily in competitor pricing data, ask six questions:
Why do we need it?
Does it matter to our customers?
Have we identified the right competitors and benchmark?
Are the offers genuinely comparable?
Can we collect it properly?
Does it tell us anything our existing data does not?
Then test it.
Knowing your competitors can make your pricing stronger. Following them blindly — or assuming that their prices must improve your model — can do the opposite.
AI-driven pricing insights to unlock hidden revenue & margin
Using your data. Minimal involvement. No implementation needed.
Comments