Fler resurser
Would your business be enhanced if you knew your customers’ preferences? What if you could predict their behavior? You can gain access to this data with choice experiments.
Fortune 500 companies have used choice experiments for many years, and now you can use them, too. Communicate more effectively with customers, match product features to willingness to pay, and maximize your messaging—all with a discrete choice model.
A choice experiment looks to carry out a choice model, which is the decision-making process of an individual or group by highlighting preferences in a given situation. Using surveys, you can estimate the respondents’ relative value of the various attributes you provide.
For example, a company that manufactures baby food wants to determine what information should be emphasized on its product labels. They use a choice experiment to show various versions of the proposed labels—one features taste, one emphasizes the product is organic, another is focused on value/price, and one features nutritional values.
The results, collected via survey, will determine what attribute is most important to their target market. This information will be used to choose the label, inform product messaging, and potentially change the direction of their entire marketing strategy.
The most common uses of choice experiments are:
There are three commonly used types of choice experiments, discrete choice, conjoint analysis, and volumetric choice. Let’s take a closer look at each type.
A discrete choice model determines the probability that a consumer will choose a particular alternative when presented with products in a realistic marketplace scenario. This is the primary form of choice analysis.
Current and potential customers are given surveys that provide a view of competing products and are then presented with a variety of combinations of marketing strategies. They are asked to choose the product they would buy based on the marketing.
Example of a discrete choice survey question:
Please look at the packaging, and read the information provided for these two bags of coffee.
Brand A is a dark roast coffee from Nicaragua. It is whole-bean coffee with notes of cocoa, walnut, and a hint of citrus. The price is $7.99.
Brand B is a dark roast coffee from Ethiopia. It is pre-ground coffee with notes of hickory. The price is $8.49, and 5% of the profits are donated to provide villagers in Ethiopia with drinking water.
Based on the information and images, which brand would you buy?
Conjoint, or trade-off analysis, is often considered an alternative to discrete choice. Conjoint analysis breaks down products into a series of attributes that may influence a potential customer’s perceived value of the product. The analysis reveals the most influential attributes on purchase decisions.
Example of a conjoint analysis survey question:
Imagine you’re looking for a home near Lake Apopka in Florida. These are the first two homes presented to you. Of these two homes, which would you choose?
Attributes | House A | House B |
Size (in sq. ft.) | 1,852 | 1,985 |
Age (in years) | 3 | 12 |
Distance from lakefront (in miles) | 2.2 | 0.1 |
Price | $479,000 | $749,000 |
Your selection | ꠛ | ꠛ |
In this type of choice experiment, each attribute is assigned a value. During analysis, you can determine the importance of each attribute to consumers and the price they are willing to pay for their purchase.
Volumetric choice modeling is useful for industries where repeat or multiple purchases are common. The survey questions are similar to discrete and conjoint surveys, with a focus on questions about purchase volume.
Example of a volumetric choice modeling survey question:
You’re assigned to order paper for the 10 printers/copiers in your company. Read the information in the following table and indicate how many packages of paper you would buy for one month.
Product | Amazon Basics | Hammermill Copy Plus | Office Depot Multi-Use | Epson Bright White | HP Everyday Paper |
Amount | 3 reams | 10 reams | 3 reams | 1 ream | 5 reams |
Price | $20.36 | $37.99 | $21.99 | $7.19 | $28.99 |
Quantity you would purchase |
In this case, data analysis will reveal the importance of a brand name, price, and the number of reams per package in a situation where multiple brands can be purchased and a larger quantity is needed.
Choice experiments are useful in various settings, including healthcare, restaurants, packaged goods, and financial institutions. Let’s take a look at these four instances where choice experiments are of particular use.
According to the National Center for Health Statistics from the Center for Disease Control, about 37% of US adults eat fast food every day. With that in mind, it makes sense that fast food companies want to tailor their offerings to what their consumers really want.
Food choices are made in multiple contexts where choice is available. Using choice experiments, fast food companies can determine what attributes are most important to consumers. This can determine everything from what is offered in “value meals” to combinations of ingredients to special offers and promotional pricing.
For example:
On your way home from work, you decide to pick up a cheeseburger for dinner. You have the following choices. Which one would you purchase?
Meal name | Value meal 1 | Value meal 2 | Value meal 3 |
Contents | Cheeseburger, medium fries and choice of soda | ¼ lb. cheeseburger, large fries and choice soda | Double cheeseburger, choice of salad or fries and choice of soda or bottled water |
Price | $9.99 | $11.29 | $12.89 |
Approximate Calories | 1000* | 1600* | 1200* |
Special pricing | none | Buy one meal get one at half price | Free cookie with a donation to local charity |
Selection | 𛰐 | 𛰐 | 𐞺 |
*depending upon selections
The responses to this survey will help the restaurant determine the value consumers place on price, calories, special promotions, and meal combinations.
Grocery stores offer tons of packaged goods, with a plethora of choices for each type. Choice experiments help manufacturers of various products determine what attributes affect willingness to pay and customer preferences. The focus of these studies is on product features, ingredients, pricing, and promotions.
For example, if the store offers four types of milk—organic, homogenized, almond, and chocolate—which is the consumers' choice? Are they willing to pay a premium for chocolate or organic milk? These questions can be answered with choice experiments like this:
Milk type | Organic | 2% milk | Almond | Chocolate |
Price/gallon | $4.74 | $4.36 | $6.21 | $6.89 |
Calories/8 oz. | 150 | 103 | 60 | 209 |
Promotion | none | Limited time free cookie with purchase | Company donates to charity with every purchase | Limited time free branded plastic cup and straw |
Selection | 𛰐 | 𛰐 | 𐞺 | 𛰐 |
Choice experiments in financial institutions examine what combinations of rates, fees, and perks for financial products are most attractive to potential customers. This can be used for multiple financial products, including credit cards, financial advisement, mortgages, loans, and investments.
A financial institution’s use of choice experiments could look like this:
You wish to open a new checking account. Study the choices below and select the bank you would choose.
Bank | Citibank | M&T Bank | Citizens Bank |
Offer for opening new account | $2000 | $100 | $400 |
Monthly maintenance fee | $12 | $4.95 | $0 |
Required minimum deposit | $1500 | $25 | $5000 |
Selection | 𛰐 | 𛰐 | 𐞺 |
Discrete choice experiments are used extensively in the healthcare industry to evaluate health technology, resource allotment, provision of services, staff attitudes, and more.
An example of a discrete choice experiment in healthcare is looking at factors of importance when choosing a provider with attributes like waiting time for an appointment, reviews of the healthcare provider, location of treatment (private office or clinic), and type of staff providing care (doctor or nurse practitioner).
A healthcare practitioner’s office might send out a survey that asks:
You’re seeking a new healthcare provider. Please review the following office attributes and choose the one you would select.
Office | Clinic 1 | Private office | Clinic 2 |
Attributes | - nurse practitioner (NP) to fill in when a doctor is out- guaranteed to be seen within 15 minutes of arrival- online self-serve appointment scheduling | - doctor only (no NP)- no information on wait times- appointments taken by phone | - multiple doctors and NPs - advertises “short wait times”- online and telephone scheduling |
Selection | 𛰐 | 𛰐 | 𐞺 |
It’s simple to design your choice experiment with SurveyMonkey. We have an abundance of resources available to help you create a professional, thorough survey to gather the data you need. Here are a few tips to keep in mind as you start designing your experiment.
Before you begin your experiment, you need to identify your target market segment. You’ll collect the most relevant data from a group that’s made up of your potential customers.
Need help finding people to participate in your choice experiment? SurveyMonkey Audience is a useful tool for obtaining survey responses based on the characteristics of your target market.
Choice experiments will help you determine the relative importance of each attribute you test. So, how do you decide which features to include? Choosing attributes that will provide actionable insights are your best bet. Run a short screening survey with Likert scales or ranking questions to find out what attributes are important to consumers
We suggest including no more than seven attributes or features in your study. More than that will feel burdensome to participants and be more difficult to complete on a mobile device.
Use your experiment to gather demographic information from your respondents. This is useful for further segmentation of your target market responses when you analyze your data.
Decide if you’re going to add an opt-out option to your choice experiment questions. This would be an option such as “none of the above” and would be relevant when the answer choices potentially don’t apply to everyone. For example, a survey question asking about preferences in types of whisky would not apply to someone who doesn’t drink alcohol.
State the survey’s purpose, instructions, and privacy information before you introduce any questions in your study. Make these sections and your questions easy for your respondents to read and comprehend.
Avoid providing answer choices that contradict each other or restate the same thing. And be careful not to repeat questions. These types of survey questions are longer and take more concentration—respondents can experience survey fatigue, which leads to unfinished surveys and less data for your study if they aren’t concise.
Discover your customers’ preferences and shopping behaviors with choice experiments. They are used effectively across several industries to inform decisions about features, bundles, and even pricing. Everything you need to create your first survey is here in our SurveyMonkey market research services.
Explore all of our research solutions to amp up your marketing, innovate your products, and boost your brand.
Collect market research data by sending your survey to a representative sample
Get help with your market research project by working with our expert research team
Test creative or product concepts using an automated approach to analysis and reporting
To read more market research resources, visit our Sitemap.