Advances in Understanding Consumer Decision-Making

One of the workhorse tools used to understand consumer decision-making is Conjoint Analysis. From its roots in psychology and transportation science, Conjoint Analysis has evolved from a simple rating of one experimentally-designed product/service prototype to choice tasks which reflect a broader range of decision-making situations.

These models have their roots in microeconomic models of utility maximization. And while some have questioned the basic assumption of utility maximization, most academics and practicing researchers agree that this assumption continues to characterize human behavior in most situations.

Additional Choice Situations

Important advances have been made since the introduction of Choice-Based Conjoint Analysis (read more in our Knowledge Center). Several of these can be seen in the below figure. in4mation insights has the tools, talent, team, and technology to help you understand what drives consumer decisions, whatever the buying situation.


Single Item

Multiple Units of One Item

Multiple Items (One Each)

Multiple Units of Multiple Items

lotion 120

beer 120

laptop 120

babyfood 120

Complementary Items

Shoppers Influenced by Promotions

Shoppers Looking for Quantity Discounts

Shoppers Constrained by Other Factors

razor 120

coupon clipping 120

papertowel 120

wallet 120

 

Categories Divided into Quality Levels (Trade-Up or Trade-Down Opportunities)

 

 

ring1
ring 2 120

 


Single-Item Choice:

This choice situation is the standard workhorse conceptualization of Choice-Based Conjoint Analysis. The consumer is presented with several products/services and is asked to pick the one they would buy or perhaps choose "none." What has become clear is that this choice situation covers only a fraction of decision-making situations. Some others follow.

Multiple Units of One Item:

This situation covers the purchasing of multiple units. For example, when buying beer, would the consumer rather buy a six-pack or a thirty-pack? The larger packsize requires a larger outlay of scarce funds. Under what condition will the larger or smaller package be favored, and by whom? Read the published paper "A Choice Model for Packaged Goods: Dealing with Discrete Quantities and Quantity Discounts," for more on this approach.

Multiple Items (One Each):

This case covers the idea of buying several separate products at the same time. The classic case is shopping for a computer online. Configuring the exact computer you want is a series of single choices, with each choice being priced separately. The end price is the sum of the individual items selected, plus any discounts that may be applied. This idea has been called "mass customization" or "user-configured design" and has been applied by many companies to great effect. Read the published paper, "Have it Your Way: Menu-based conjoint analysis helps marketers understand mass customization," for more on this method.

Multiple Units of Multiple Items:

Imagine you are a marketer of yogurt or baby food. Does the classic "choose one" CBCA method work here? Of course not. People buy multiple units of multiple flavors. Or when studying a family's total beverage choices, how does the price of, say, bottled water affect the price of beer or carbonated soft drinks. The standard model can be tricked to handle this situation, but the newer methods use more appropriate assumptions and statistical tools.

Complementary Items:

Many product offerings consist of a "base product" -- the handle -- and sets of attachments/refills that go with them. Examples of this are electric toothbrushes, razor blades and Swiffer™ mops. Should you give away the handles to get the refill stream of cartridges? Can you get someone to switch if they have cartridge inventory on hand? Will consumers upgrade their cartridges if they cost more but fit the same handle? These are questions that require a specialized model that handles the mutual impact of handles on refills and refills on handles.

Influence of Promotions:

Many companies have used promotions as a way to introduce a product, increase awareness, gain new users, or temporarily increase share. The question becomes, "What are the best promotional styles and tactics?" in4mation insights has pioneered new ways to compare the reactions to different promotions, to forecast their in-market performance, and to quantify the benefits of the winning promotion, not only to the manufacturer, but also to the retailer -- effectively identifying a win-win situation. This includes measurement of discounts, shelf-talkers and coupons but also new ways to measure promotions that drive quantity increases such as "buy one get 50¢ off the second."

Categories are Divided into Quality Levels:

As the number of entries into product/service categories has proliferated, many companies have tried to differentiate their products by either "premiumizing" their products or by becoming an economy brand. These tactics have squeezed the marketing and pricing of mainstream brands. in4mation insights has developed methods to understand what can be done to help in the premiumization, or alternatively, the value-branding process. What are the features, positioning, branding, and pricing signals that convey these ideas to the consumer? We identify which consumers are so influenced and in which price tiers they are willing to purchase in which product categories. Read the published paper "A Model for Trade-Up and Change in Considered Brands" for more on this topic.

Shoppers Look for Quantity Discounts:

Imagine the consumer in this situation: they can buy the 16 ounce bottle for $3.99 and the 32 ounce bottle for $6.99. On a price-per-ounce basis, the consumer should buy the larger size. Yet, on a relative price basis the smaller size should be purchased. This illustrates that a model that properly measures the effect of price in the quantity discount situation must be used. See the Knowledge Center for more details.

Constraints on Purchasing:

Modern economics teaches us that consumers make choices under constraints. Yet the most common statistical model used in choice modeling (e.g. multinomial logit) has the property that, even if you increase the price of a product to astronomical levels, it will predict that someone will still buy. Yet, we know that people have budgets -- constraints -- when shopping, which they may exceed in some circumstances. But a budget is a "hard stop" -- exceed the budget and the likelihood of purchase should be zero. In other circumstances, the constraint may be physical. Some people cannot purchase a large size, despite an attractive price, because it is either too heavy to carry or too large to store at home. Calories are another type of constraint. The shopper may want the benefits of a granola bar but may limit himself to less than 200 calories. We understand and can apply new methods of introducing and quantifying constraints on the choice process.

Other Advancements:

The above discussion covers new models for understanding consumer choice. Additional advancements in marketing research methods for choice situations that we can apply include:

  • Models for purchasing small numbers of units (e.g. count models) with corrections for excess zero purchasing;
  • Joint choice and volume models;
  • Allocation models, often used in pharmaceutical and financial investing applications;
  • Advanced choice designs which incorporate managers' prior beliefs about the results, which have been shown to be more efficient than standard linear choice designs; and,
  • Our own implementation of Bayesian individually-adapted choice designs, which has been shown in the academic literature to outperform commercially-available solutions.

Paper

Read more about CBCA in Steve's paper, "Perfect Union."

Click here to download