A major fast-food chain faced a critical decision about changing their beverage supplier. A challenger supplier was offering them a deal to switch. Their traditional research showed customers cared about drink choice – but by how much? Using choice modeling, we uncovered that changing suppliers would cost them $0.25 per meal in value – significantly more than the discount the challenging supplier offered. The data gave them the confidence to maintain their current partnership, protecting both their brand equity and their bottom line.
This is the power of choice modeling – turning complex decisions into clear action.
The decision challenge
Consider this: People consistently rate price as “very important” when buying laundry detergent. Yet the market leader, Tide, continues to dominate. Why? Because what people say drives their decisions often differs from what actually drives their choices.
Humans often struggle to explain what drives their decisions. When asked directly, we tend focus on what we believe should matter or rationalize our choices after the fact. This happens because our decision-making processes are influenced by both emotional, intuitive reactions and more deliberate, rational thinking. When making decisions, we often rely on our emotions and intuition first and then use our reasoning to justify the choice later. This is why someone might say price is a key factor in a survey and still opt for higher-end brands when shopping.
The true drivers of our choices emerge when we observe actual decisions being made. The difference lies in context: without a specific scenario, we often overestimate how much certain factors, like price, actually matter. For example, a price difference of a dime may not affect our detergent choice, but a ten-dollar difference could be a deciding factor. By placing options in direct competition, we give ourselves the context needed to reveal what truly influences our decisions.
Challenging convention
Traditional rating scales – those familiar 1-to-5 or 1-to-10 options – have served researchers well. But they have clear limitations:
- People struggle to make precise distinctions on rating scales
- Individual scale usage varies dramatically (your “7” might be my “5”)
- International comparisons are particularly challenging because different cultures use scales differently
- There is typically a “halo” of positive correlations among items, because the correlations reflect how people use scales more so than how they feel about the individual items
- Most importantly, they often fail to reveal true decision drivers
The choice solution
Choice modeling breaks through these limitations by putting options in direct competition. Think of it this way: Ask people to rate how much they value being wealthy, intelligent, or good-looking on a 5-point scale. Most will rate all three highly. But ask them to choose among these attributes? Now you uncover what truly matters to them.
The trade-offs
Rating scales are comfortable and familiar, so they remain popular. Researchers and clients know how to track them over time, and they let us seemingly ask about what we want to measure. But this comfort comes at a cost of precision and actionability.
Choice models also have their challenges. Everything in a choice model typically sums to 100%, meaning when one thing goes up, others must go down. This can make tracking changes over time more complex than with traditional rating scales. Not that this is intractable, but it is an issue. Choice models also often require more sophisticated analytics, which can make mid-wave reporting more challenging than simple average scores.
At PSB, we’ve pioneered an integrated approach to choice-based research that breaks through traditional limitations. By maintaining consistent anchors across waves and carefully designing our choice sets, we can track changes over time while maintaining the power of forced trade-offs. Our advanced analytics reveal not just what people choose, but why those choices matter for your business.
Our original approach works with both complex Discrete Choice Models and the simple MaxDiff. We leverage information to extend the shelf life of expensive Conjoint models or DCMs, as well as make MaxDiff techniques not only a viable tracking tool, but a preferred one in many situations.
The decision impact
When we put this approach into action, we see breakthrough results:
- A fast-food chain avoided a costly supplier switch by discovering that changing beverages would cost them $0.25 per meal – more than the proposed supplier discount
- A national coffee chain uncovered the true barriers preventing customers from visiting during key dayparts, enabling them to make strategic operational changes
- A healthcare company identified which marketing messages would both motivate action and differentiate their teeth whitening product, leading to more effective communications strategy
The decision guide
Choice modeling proves most valuable when:
- Stakes are high
- Multiple factors influence decisions
- Traditional research shows everything is “important”
- You need confident, actionable results
When stakes are high, you need more than opinions – you need to understand what truly drives decisions. Choice modeling cuts through the noise, revealing clear priorities that give you the confidence to act. It’s not just about what people say they’ll do – it’s about understanding what they actually choose when it matters.
Ready to break through? Connect with Rob Kaiser, Chief Methodologist at robkphd@psbinsights.com to discuss how choice models can drive your next big decision.
