Market replication
We replicate markets: real buyers, real choices and real competitors, faithfully enough to try any move and see what it would earn. That is why big decisions bank on our models.
Every market, rebuilt to scale
Billions served
$1B+
Our projects have been worth more than a billion dollars to our clients, counting only what we know; much of our work runs behind research and consulting firms.
What you take home
Every engagement ends with a simulator of your own market: every product, channel, shopper segment and competitor at once. Change prices, launch or retire products, shift distribution and promotion, and let competitors respond until the market settles. See what the whole company earns, with how sure we are.
Today's questions are in it. So are the next ten.
Representative screen · illustrative values
Your line-up
| Product | Price | Shelf |
|---|---|---|
| Flagship 4-pack | $6.49 → $6.99 | On |
| Premium cold brew | New · $3.79 | Launch |
| Single 13.7 oz | $2.99 | On |
| Value 12-pack | $13.99 → $14.49 | On |
| Multipack 24 | $24.99 | Retire |
| Light 8-pack | $9.99 | On |
Competitors
Market
Firm earnings, net of cannibalization · after competitors respond
Earnings change by product and channel, $M a year
| Grocery | Mass | Club | Conv. | Online | Total | |
|---|---|---|---|---|---|---|
| Flagship 4-pack | +12 | +9 | +6 | +5 | +3 | +35 |
| Premium cold brew | +14 | +8 | +4 | +6 | +7 | +39 |
| Single 13.7 oz | +2 | +1 | 0 | +5 | +1 | +9 |
| Value 12-pack | +6 | +8 | +9 | −1 | +2 | +24 |
| Multipack 24 | −4 | −3 | −2 | 0 | −1 | −10 |
| Light 8-pack | +3 | +2 | +2 | +1 | +1 | +9 |
Where the market settles
By shopper segment, earnings change
Questions in this run
2,000 futures of the market · shoppers, channels and promotions drawn with their real correlations · calibrated to 52 weeks of sales · competitors move until no one can gain.
A scale model, to fit the page.
A real replica is far bigger than any screen can show. Client models carry every item and pack-price point, every retailer, region and channel, every competitor and private label, promotions week by week, and buyers in their true mix. The picture above is a slice.
What the model does
Choice 1 of 1
Which would you stake your next big decision on?
The evidence
Within 1 point of actual market shares. Attribute-based models missed by 13 to 22.
Just before founding Market Outcomes, at the American Marketing Association's Advanced Research Techniques Forum, Beau Martin presented An Empirical Test of Pricing Techniques: eight pricing methods, 1,620 buyers, one task: predict a month of the U.S. home gaming console market whose actual sales were already known.
Average share error
Median time to finish
Buyers who gave up partway
The work
Technology · a challenger's price cut
A challenger cut prices. The model showed buyers stayed with one supplier, so matching the cut would mostly have handed margin to buyers. The client team credited the work with about half of an unexpected billion-dollar gain.
The full story →
Product line-up · U.S. and EU
Every printer in the line replicated against its rivals and against each other, across the U.S. and EU. The optimized line went from 28 products to 14 or 15, raised expected net earnings by more than $300 million a year, and simplified the supply chain and support behind it.
The full story →
Pack and price by channel · MRIA Project of the Year
Pack sizes and price points replicated store type by store type and calibrated to real sales, to set the right ladder for each channel.
Every year since 2017
The whole menu board replicated: which items stay, which come off, which launch and at what price, with every order the change touches.
Brand valuation · Global strategy project of the year
Every market either name sold in, from local phone to wireless, internet and TV, replicated with five years of growth and margin. One name was worth billions a year more, and it became the name. It even beat the next target's own name in that company's home territory, and the next acquisition followed.
Specialty pharmaceuticals
Physicians' choices replicated for patients at their true clinical mix, so the launch team could see where the product would win, with whom and at what price.
Pick-and-pay television
Canada's regulator unbundled TV, and there was no history to forecast from. We replicated a market that did not exist yet, and the client positioned its line-up and prices for the most profitable result.
The full story →
Tax and small-business software
Customers who did not come back told us why. The reasons were largely fixable, and fixing them cut the share lost in half. Then the road map, optimized against staffing and uncertain task times, and later its small-business accounting editions.
Excise tax · Thailand · for an MBB strategy firm
Thailand taxes alcohol at the higher of a specific and an ad valorem rate. Before rates rose, we replicated how drinkers would switch among specific brands of beer, wine and spirits, each with its own price and alcohol content. The company used it to weigh acquisitions and to make the case for a tax design that served both the treasury and the company.
Learning toys · Grand Effie
A far cheaper rival came for the category leader. Head-to-head ad testing showed which ads won with which viewers, and where to place each one. The campaign won the Grand Effie, and the rival soon left the market.
The full story →
Express toll lanes
Drivers choosing between free lanes, express lanes and other routes, at prices that move with traffic, so the operator could see traffic and revenue at every price.
Passenger rail and transit
Travelers choosing between rail, bus, car and plane at real fares and schedules, so each agency could weigh ridership against fare revenue.
Clients
The markets these companies lead are markets we have replicated. Some engagements are direct; many are as the modelers behind the research and consulting firms that lead them.
Pharmaceuticals and health
Food, beverage and consumer goods
Technology, industrial and automotive
Telecom and financial services
Transportation
Higher education
“We spent just 4 hours with Beau and in that time we came out with a simple, but effective framework and desired set of outcomes that we need to shape a major change in our strategy. We all have the data, Beau makes it make sense.”
“Probably the best ever project we’ve done with your firm.”
“In over 10 years in both market research and strategy consulting I have never met anyone with a greater depth of knowledge.”
How an engagement runs
Some decisions cannot wait; others deserve every market, segment and scenario. Design, fieldwork, replication and optimization run on PansophIQ, our own platform, so the time goes into the thinking, not the logistics.
The call, every move on the table and what rides on it. Bring sales data if you have it; we work from what exists.
Every study is built around your market and reviewed before it fields. Once the design is agreed, fieldwork can start the same day.
The replica must reproduce today's sales before any scenario is trusted.
The best move, what it earns, how sure we are, and a simulator of your market to keep asking.
For research and consulting firms
When your client's decision needs the market itself modeled, bring us in. Your name on the work, your client stays yours, and we handle the hardest part.
How we work
Most studies ask buyers to choose among products that do not exist. We replicate the market itself, prove it against today's sales, then search every move for the one that earns the most.
The shelf, the menu board, the channel line-up, the prescribing decision: the options buyers actually see, the choice to buy nothing, the steps they take.
Real competitors, available where they are really sold, with buyers, patients and occasions in their true mix and true correlations.
Tuned to the market as it stands and tested on choices it never saw, before any scenario is trusted.
Volume, revenue and margin for every product, netted for cannibalization into what the whole company earns, with its range.
Because the model is the market, clients stack their strategic questions into it: price, line-up, launch, distribution, a competitor's move. Every answer comes from the same market, so the answers agree.
Prices, line-ups and launches searched for the best result, with competitors free to respond, to find where the market settles.
Market optimizer · illustration
Your line-up today vs. the optimized line-up
| Product | Move | Earnings |
|---|---|---|
| Flagship | Price +9% | +$38M |
| Line extension | Launch | +$41M |
| Value pack | Wider distribution | +$27M |
| Promotions | Fewer, deeper deals | +$13M |
| Legacy item | Retire | +$6M |
| Core line | Cannibalized | −$19M |
Illustrative figures. What you take home: the best move, what it earns across the whole line-up, and how sure we are.
Inside our models
Line-ups of any size, sold where they are actually sold: each retailer, region and channel with its own shelf, distribution and prices.
Shoppers, households, patients or travelers in their true mix and true correlations, never one average buyer. Every result breaks out by segment.
Each rival can hold, match or play its best reply, round after round, until the market settles. Entrants and exits too.
Volume, revenue, margin and company earnings net of cannibalization, each with its range from thousands of simulated futures.
The model searches prices, line-ups, distribution and promotion for the most earnings within your real limits: capacity, price ladders, retailer rules.
Dozens of strategic questions answered in the same replica, so the answers agree, and the simulator stays with you for the next one.
Beyond market replication
Built on PansophIQ, our own platform for designing, fielding and modeling studies.
Every answer comes with its range
Never a single number: thousands of plausible futures, so every move comes with what it would earn and the odds it pays.
Illustration: one product against a competitor and the choice to buy nothing, 2,000 simulated futures. In client work the same picture covers every product, channel and competitor response together.
An engagement
We start from the call you have to make, every move open to you, and every move competitors could make back.
Buyers, choices, competitors and where each is sold, measured and rebuilt as the market actually is.
The replica must reproduce today's sales, and choices it never saw, before any scenario is trusted.
The best move, what it earns across the whole company, how sure we are, and a simulator to keep asking.
The work
Many ran behind the research and consulting firms that led them; we tell only what is ours to tell.
Children's learning toys
The leader in children's learning technology
Its stylus-based reader led the category. The toy industry's giant launched a rival at a far lower price: much less capable, and built to take the market anyway. Our head-to-head ad testing showed which ads won with which viewers, so its agency could pick the right ads and place them in the right channel and time slot for each audience.
The campaign won the Grand Effie, the top award for advertising effectiveness. Within a year or two the challenger was in retreat, and it soon left the market.
Pick-and-pay television
Canada's largest wireless and cable company
Canada's regulator required pick-and-pay channels, ending the bundles the industry had always sold. There was no history to forecast from. We replicated a market that did not exist yet: how households would choose channels one by one, in small packs and in bundles, at every price.
The client used the model to position its line-up and prices for the most profitable result in the new market.
Technology, and a challenger's price cut
The challenger cut prices. Follow, and how far? The model showed buyers stayed with one supplier to avoid running mixed systems, so matching the cut would mostly have handed margin to buyers. The same model priced the whole line-up to what the factories could make.
The client team credited the work with about half of an unexpected billion-dollar gain.
The client team's own attribution, as told to us.
Product line-up, U.S. and EU
A printer leader: the inkjet and multifunction line
The inkjet and multifunction line had grown to 28 products across the U.S. and EU, many competing with each other as much as with rivals. We replicated the whole market, every one of the client's models alongside the competition, and optimized which products to offer.
The answer was 14 or 15 products, depending on the goal, with one model on the bubble. The smaller line raised expected net earnings by more than $300 million a year and took much of the complexity out of the supply chain and support behind it.
Pack and price by channel · MRIA Project of the Year
Pack sizes and price points replicated store type by store type and calibrated to real sales, to set the right ladder for each channel.
Every year since 2017
The whole menu board replicated: which items stay, which come off, which launch and at what price, with every order the change touches.
Brand valuation · Global strategy project of the year
Every market either name sold in, from local phone to wireless, internet and TV, replicated with five years of growth and margin. One name was worth billions a year more, and it became the name. It even beat the next target's own name in that company's home territory, and the next acquisition followed.
Specialty pharmaceuticals
Physicians' choices replicated for patients at their true clinical mix, so the launch team could see where the product would win, with whom and at what price.
Tax and small-business software
Customers who did not come back told us why. The reasons were largely fixable, and fixing them cut the share lost in half. Then the road map, optimized against staffing and uncertain task times, and later its small-business accounting editions.
Excise tax · Thailand · for an MBB strategy firm
Thailand taxes alcohol at the higher of a specific and an ad valorem rate. Before rates rose, we replicated how drinkers would switch among specific brands of beer, wine and spirits, each with its own price and alcohol content. The company used it to weigh acquisitions and to make the case for a tax design that served both the treasury and the company.
Express toll lanes
Drivers choosing between free lanes, express lanes and other routes, at prices that move with traffic, so the operator could see traffic and revenue at every price.
Passenger rail and transit
Travelers choosing between rail, bus, car and plane at real fares and schedules, so each agency could weigh ridership against fare revenue.
Every admissions cycle
Admitted students choosing among real competing offers, so each school could set aid and net price for the class it wants.
For research and consulting firms
Firms bring us in when a client's decision needs the market itself modeled. We welcome it.
We work under your brand, inside your deliverables, or beside you in the room. Whichever serves your client.
We never pitch a partner's client. The relationship, the account and the next project are yours.
Design, fieldwork, the replica, calibration, the optimizer and a simulator your client keeps.
The design decides what the model can answer. We help shape the study around the decision, so nothing has to be patched later.
Already fielded? We can work from your data and still replicate the market around it.
Price, line-up, launch and competitive response in one market, so every answer you present agrees with the others.
Our own platform: design, fieldwork, replication and optimization as fast as the decision needs and as deep as it deserves, every figure traceable.
The evidence
Eight ways of predicting a market, the same buyers, and a month of actual sales to score them against. Market replication came within a point. Attribute-based models missed by 13 to 22.
The test
Just before founding Market Outcomes, Beau Martin presented An Empirical Test of Pricing Techniques at the American Marketing Association's Advanced Research Techniques Forum. The design was simple and unforgiving. 1,620 buyers were split across eight pricing methods. Each method predicted the shares of the U.S. home gaming console market for a month whose actual sales were then published, and each was scored on how far its shares landed from the real ones.
A home console market is a hard test: a few strong brands, real price differences, and buyers who care about far more than a list of features.
The result
Market replication put every brand within 0 to 1 point of its actual share, in the right order. The attribute-based versions missed by 13 to 22 points on average, and the worst of them barely ranked the brands at all. In the months that followed, market replication's error stayed under 2 points.
Why the gap
Attribute-based models treat every product as an interchangeable bundle of features. Cut down to brand and price alone, the same approach missed by 4 to 7 points instead of 13 to 22. The features were doing the damage, not the buyers.
They underreacted to price. At the market leader's actual price, they put its share at half of what it really sold. A pricing decision rides on exactly that number.
The task took nearly three times as long and lost more than three times as many buyers partway. Fatigue and drop-out bend the answers before any model is fit.
What makes a replica trustworthy
The products and packs actually on sale, at their prices, where each is sold, to buyers in their true mix.
Each product competes the way it really does, so a premium launch takes from premium and cannibalization is counted.
The replica has to reproduce today's sales before it is allowed to predict anything.
Choices held back from the model have to be predicted too, before any scenario is trusted.
Real choices, shown as buyers meet them, so people finish and answer as they would in the store.
Thousands of simulated futures, so every move comes with what it would earn and the odds it pays.
Martin and Rayner, An Empirical Test of Pricing Techniques, American Marketing Association Advanced Research Techniques Forum.
Privacy
Effective October 2026
Market Outcomes ("we") operates marketoutcomes.com. This policy covers information collected through this website.
When you use the contact form, we collect what you enter: your name, work email, company and message. Our hosting provider keeps standard server logs (such as IP address, browser type and the pages requested) to run and secure the site.
This site sets no advertising or analytics cookies and does not track you across other sites.
We use contact-form information only to reply to you and to discuss the work you asked about. Server logs are used only to operate, secure and troubleshoot the site.
We do not sell or rent personal information. We share it only with service providers who host this site and deliver our email, under obligations to protect it, or when the law requires.
We keep contact-form messages as long as needed to respond and to keep a record of our business correspondence, then delete them. Server logs are kept for a limited period set by our hosting provider.
You can ask us what we hold about you, ask us to correct or delete it, or ask us to stop contacting you. Depending on where you live, including California and the European Economic Area, you may have further rights under local law. Send requests through the contact form and we will respond within 30 days.
Data we handle for clients in the course of an engagement is governed by our agreement with that client, not by this policy.
If we change this policy, we will post the new version here with a new effective date.
Terms
Effective October 2026
By using marketoutcomes.com you agree to these terms. If you do not agree, please do not use the site.
The content on this site describes our work and approach for general information. It is not advice for any particular decision. Figures shown in illustrations and representative screens are illustrative unless stated otherwise. Engagements are governed by a separate written agreement.
The site's text, design, images and graphics belong to Market Outcomes or its licensors. Client names and logos belong to their owners and appear only to identify companies whose markets we have worked on. You may view and share pages for your own reference; any other use needs our written permission.
The site is provided "as is". We work to keep it accurate and available but do not guarantee that it is complete, current or free of interruption.
To the extent the law allows, Market Outcomes is not liable for any loss arising from use of this site or reliance on its content.
Links to other sites are provided for convenience; we are not responsible for their content.
We may update these terms; the version posted here applies from its effective date.
About
Beau Martin has replicated markets since 2000 and founded Market Outcomes in 2008. His background is operations research: building models that hold up when real money rides on them. The practice works directly with companies and as the modeling partner to research and consulting firms, often under their name.
Our models now run on PansophIQ, the platform we built to design, field, replicate and optimize a market as fast as the decision needs and as deep as it deserves, with every figure traceable to its source.