{{item.videoDuration}}
{{item.title}}
{{item.text}}
{{item.videoDuration}}
{{item.text}}
Many retailers still try to predict tomorrow’s shoppers with yesterday’s methods, but that approach might no longer be sufficient. Organizations are required to make critical decisions every day related to assortment, new product introduction, pricing, promotions, and even the sustainability of products. Consumers are moving faster than surveys can capture, and faster than test markets can trial. To keep up, modern companies need something different—not just faster and more narrow feedback, but a smarter way to see what’s coming next.
What if you could test a pricing strategy before it ever reached the shelf, and get feedback from millions of shoppers in minutes? Sooner than most executives think, retail strategies won't be tested in stores or surveys, but in simulations built on intelligent data foundations.
Video
Synthetic customers: Rewriting retail’s playbook
1:11
More tools
Transcript
A grocery retailer evaluating a protein-packed frozen meal promotion doesn’t have to rely on guesswork. With advances in AI, “synthetic customers” can provide reliable proxies on how real people shop—and explain their decisions in clear, conversational terms.
For example, in response to the promotion detailed above, a persona modeled on a millennial GLP-1 user might say: “At that price, I’d buy three meals and freeze them,” an indication of the pantry effect. Another, representing Gen Z parents, might push back: “My kids won’t eat the store brand—the discount doesn’t matter,” potentially leaving the brand or the category.
These aren’t survey quotes. They’re responses from synthetic customers—AI-enabled digital personas trained on real-world behavioral data. And they’re giving retailers something they’ve long lacked: actionable foresight before decisions go live.
Synthetic customers won’t buy anything—but they could tell you who will.
Many companies are missing the ‘why’ behind what customers buy
Legacy analytics can reveal past behavior, not future drivers. While retailers own the point-of-sale, consumer packaged goods (CPG) manufacturers typically own the research—and hospitality brands rely heavily on historical booking trends. But none offer clear foresight into future decisions. As loyalty shifts, budgets lighten, and customer priorities evolve, traditional methods are proving too slow, too narrow, or too siloed to guide what comes next.
For decades, retailers have had access to purchase data, but they often lacked insight into why customers buy certain products.
All of that leaves retailers operating at a disadvantage, making multimillion-dollar bets with incomplete visibility into shopper motivations. But synthetic customers can change that. They help simulate future behavior—grounded in data, not assumptions—giving retail leaders a clearer picture of what’s next.
Synthetic customers can simulate behavior and explain why
PwC’s proprietary synthetic customers draw on household-level data, including demographics, spending patterns, and location signals, to help generate decision-ready personas. And when organizations integrate their own proprietary datasets, such as loyalty, digital engagement metrics, shopper insight surveys and purchase history, these models become even more tailored to reflect how their actual customers think and act.
These aren’t static dashboards. They’re AI agents you can engage directly—designed to simulate how real customers think, choose, and change.
Executives can ask questions like: Would this discount increase a shopper’s basket size? Could swapping this product trigger unexpected churn? Where should I launch this new product? And instead of just outputting probabilities, synthetic customers can respond with AI-enabled logic in natural language, linking expected behaviors to motivations.
They can also simulate campaign messaging, packaging updates or offer design—and understand how different segments might react. Want to know how a new tagline might land with Gen Z? Whether bundling a free upgrade changes business traveler behavior? Synthetic customers can surface these reactions in real time, in plain language.
And these simulations aren’t just for retail. A CPG team can explore how a product launch might perform across price tiers and retail partners. A hospitality brand could model how leisure travelers might respond to loyalty changes or sustainability messaging—and pressure-test trade-offs before making the shift.
Each persona can also evolve with context: shifting household priorities, regional dynamics, economic shocks, or changes in life stage. You can model behavior over months or years, then pressure-test outcomes across categories, channels, or formats.
Retailers no longer have to guess how customers will respond. They can ask.
Synthetic testing can drive revenue, margin and speed
Whether launching a promotion in retail, a new flavor in CPG, or a tiered room package in hospitality, teams often rely on slow pilots or conflicting insights. But by the time the results arrive, the market may have already moved on. Gen Z, for example, might redefine “value” before the campaign even concludes.
Synthetic customers significantly compress this cycle. Retailers can simulate promotions, pricing strategies, loyalty mechanics, and more—virtually, safely and at scale. They can also open up a broader range of high impact simulations, including:
- How will Gen Z respond to sustainability messaging and carbon labeling?
- What happens if inflation spikes or buy now, pay later fatigue sets in?
- How might GLP-1 drug adoption shift demand?
- Will business travelers value flexible check-out more than bundled loyalty offers?
This isn’t just speed. It’s foresight—and sharper signal detection. In modeled environments, synthetic customer testing has shown measurable upside as a growth driver, not just an efficiency play. And that’s just the beginning. Synthetic customers don’t replace traditional research methods—they multiply their value by integrating specific company insights with market data. By testing smarter, retailers can capture growth ahead of competitors, protect the core, and reinvest savings into innovation.
What value do synthetic customers deliver?
- Sales growth: By testing new products, promotions, and marketing messages that resonate, retailers can capture new segments.
- Expanded margins: Retailers can expect to see a reduction in failed launches, costly resets, markdowns, and wasted spending.
- Accelerated speed-to-market: Innovation teams can move from a concept to a confident launch in weeks, not months.
- Increased ROI: Every dollar in assortment, pricing, and marketing can be spent in smarter and more productive ways.
- Financial upside: Synthetic customers can help drive profitable growth through higher revenue, stronger margins, and faster inventory turn.
This is a planning advantage, not just a tech feature
For retail teams, synthetic customers enable more than faster feedback. They offer a new way to connect customer insight to strategic planning and decision-making—across merchandising, operations, and marketing.
Executives can test assumptions early, explore edge-case scenarios, and challenge legacy playbooks. And because synthetic customers are based on generalized panels of data, not live customer files, they help reduce compliance, privacy, and consent risk.
And for marketing and innovation teams, synthetic customers offer something traditional pilots can’t: speed at scale. Creative concepts, product ideas, and even loyalty incentives can be tested in hours—helping teams move from hypothesis to action, without waiting for a survey or store test.
It’s important to note, however, that synthetic customers simulate intent, not emotions. They don’t replace the nuance or empathy of real human behavior, but they can model how groups might behave at scale, helping retailers prioritize more confidently.
CPG and hospitality leaders can benefit from synthetic customers, too. Whether testing campaign messaging or evaluating format shifts, they can help leaders move from retrospective analysis to proactive planning.
Retailers don’t need to rip and replace systems. They can start with focused-use cases, such as campaign or assortment planning. Synthetic customers can integrate with existing tools, and new personas can be built in days—not quarters. There’s also potential for shared insight. With synthetic customers, retailers, and CPG companies could work from a shared behavioral model—or at minimum, align on the “why” behind customer actions. In fast-moving categories, this kind of mutual visibility could ease friction, speed up decisions, and replace conflicting data narratives with forward-looking logic.
Synthetic customers can become retail’s new competitive edge
Retailers pulling ahead are the ones making more informed decisions—faster. Synthetic customers help them go beyond reacting to past behavior and instead plan around what customers are likely to do next.
They won’t replace real customer relationships. But they do surface customer logic—early enough to change the outcome.
In a market defined by speed, complexity, and shifting loyalty, knowing what customers did is no longer enough. Retailers should understand what they will do—and why. The most valuable customer insight your team gets next may not come from a dashboard or a focus group—but from a customer who isn’t real.
{{filterContent.facetedTitle}}
{{contentList.loadingText}}
{{contentList.loadingText}}
Contact us
Follow us
