Using Choice Based Conjoint (CBC) to guide go-to-market and retention strategy
Product teams need to move faster than ever when making business-critical decisions. Whether implementing go-to-market strategies or seeking to improve customer retention, they need clear direction on consumer preferences and the likely impact of different product and pricing decisions.
Choice-Based-Conjoint (CBC) is a crucial survey analysis method to translate user preferences into a clear plan of action. Here’s what you need to consider when designing a conjoint study.
What is Choice Based Conjoint?
Choice Based Conjoint (CBC) analysis is a form of survey-based analytics method that helps determine how people value different attributes (features, benefits and price) that make up a product or service proposition. We use CBC analysis to help brands identify the optimal composition (features, packages, price) of a given product or range of products. Be the objective to increase market share, boost revenues or increase retention.
CBC analysis is a great way to inform decision-making around what to build and how to position it. And should be an essential component of the toolkit of any researchers involved in product and pricing strategy. Some examples of how it can be used would be:
To optimise product proposition design by testing different combinations of features and prices
To assess sensitivity to price and identify optimal pricing strategies
To identify which aspects of your proposition (eg brand, features) are most valued
To forecast market share and predict the adoption of new products or services
To identify the optimal range of products / service packages to launch
To segment customers based on their preferences and tailor marketing strategies accordingly
To compare products with competitors in terms of customer preference and value proposition
An Example:
To take a more specific example CBC could help answer the following question, how much would someone pay for a broadband plan with a certain speed, set up fee, monthly cost and contract length?
In CBC, respondents are shown several sets of alternatives (called concepts), each with different combinations of features (called attributes) and levels (the specific values of each attribute). For example, an attribute could be contract length and the levels could be 12 months, 18 months, or 24 months. Respondents are asked to choose their most preferred concept from each set taking into account variations of all of these factors. By analysing these choices, researchers can estimate how much each attribute and level influences the overall preference and willingness to pay for a product or service.
Is Choice Based Conjoint right for you?
Based on my own experience CBC can be underutilised within the MRX industry even amongst those who are involved in product development and pricing research. This can be due to misconceptions (I need to be a statistician), concerns (takes too long / costs too much) and lack of understanding about when and where it should be used. This can be useful for those already well-versed in the technique. For instance, for individuals, it can be a great strategic specialism to hang your hat on within a consultancy. And for consultancies, a seemly opaque and complex technique can be priced at a premium.
As ever though technology has been a great leveller. So if CBC was once the domain of statisticians or required specialist technical knowledge, the increased availability of self-serve software from the likes of Sawtooth and Conjointly has meant you now no longer need a PhD to run this kind of sophisticated analysis.
This is great, but a little knowledge can be a dangerous thing. It is cheaper, easier and quicker to run a conjoint study than ever before but like a lot of the more advanced tools and models available to market researchers it requires a user-centric mindset to deliver clear actionable and trustworthy data.
1. Ensure your CBC design is based on users’ needs
You are most likely using CBC to understand end preferences, in order to shape a product or service to increase take-up. It may be that you already have a clear view of possible benefits, features, prices and promotions. However, CBC often works best as the culmination of an iterative insight workstream. Starting with an exploratory qualitative phase to understand audiences’ Jobs to be Done and unmet needs. Followed by a generative phase to identify potential solutions in the form of features, functions and benefits. This process of foundational, exploratory research followed by ideation ensures several things:
You are placing your users at the heart of product development and ensuring you are building and testing propositions that are aligned with their needs.
You have a chance to prioritise ahead of the conjoint – some internal ideas and concepts may bomb at the qual stage and therefore are obvious contenders for the chop before the conjoint exercise
You have done your due diligence and ensured that features and benefits that have potential influence on consumer choices are not excluded
2. Lean into the benefits of choice-based conjoint
There are many great reasons to conduct a CBC analysis. For starters it is a trade-off technique so does a better job than a standard survey question of replicating real-world decision-making. The evaluative table at the heart of the exercise can be seen in everything from subscription services (broadband, mobile, digital news, etc) to product comparisons on Amazon. People are used to making decisions in this way. And because product options are assessed in the round, rather than at a component level, preferences are obtained indirectly and implicitly, therefore, overcoming the drawback of direct questioning (e.g. users may not be aware of their preferences or how they influence their decisions).
When building a CBC ensure you factor in the following for your analysis:
Add market context if needed: you can look at your product or product range in isolation, but you can also use a conjoint to explore market dynamics by including a representative set of brands that compete in the same product category. This is useful to explore how changes to your product may impact the wider market. However, the impact of brand preference can be strong so be mindful of this when designing your study. If your main priority is optimising your own product/product range rather than market modelling sometimes it is best left out (see the point on heuristics below).
Include a base case scenario: whilst you are using the CBC to understand what you should do in future make sure you ground it in the present if possible. If you are optimising an existing product or disrupting a market ensure your CBC can model the current status quo with a base case of the here and now. This allows you to better understand the impact of changes on sales and market share.
Get a read on pricing: include pricing in CBC can be useful, even if it is just to provide a broad guide of what the market finds acceptable. This means choosing a range of prices that reflect the market conditions and consumer expectations but also ideally extending beyond these limits slightly to ensure you measure the extremes of price acceptance.
Estimating take-up: CBC is about relative preferences and these alone can provide great insight and direction on what you should build and launch. But if you also want to understand likely adoption this can be done by including a follow-up question to measure this. And it can be a useful additional weight to add to any further data modelling. But watch out as this effectively doubles the amount of questions you are asking in an avery involved exercise, so try to be kind to your research participants elsewhere :-)
3. But be mindful of the limitations
A CBC better reflects real-world decision-making than a standard survey question. And it better simulates the the trade-offs we make when selecting products and services. However, it is still a survey. A CBC can provide great direction on what to build and how to price it to meet your objectives. But, consider the following:
Interpreting take-up data: When it estimating take-up, don’t take the analysis as a literal prediction of future market share or adoption. Use the CBC to understand your available market share of preference at any given price point, don’t expect it to accurately predict your market share as there will be many other factors to consider (marketing, social trends, distribution channels, etc).
Accounting for heuristics: People will often use shortcuts to help inform complex decisions. Whilst this is unavoidable, combat this by avoiding over-engineering the task or the inputs. As a rule of thumb, we recommend maxing out at test six product components in total, any more and responses will struggle to get their heads around what you are offering let alone be able to trade these off against other iterations.
Focusing on the big picture: If differences in pricing or proposition are too subtle and nuanced participants will struggle to notice the difference or respond accurately. Ensure what you are testing, is clear, simple and easy to parse to ensure good quality reliable data.
Ensuring there is a great survey UX: CBC exercises can be torturous if designed poorly – the more you want to test the longer they get and the quality of the UX varies widely from platform to platform. Make sure the experience is as engaging as possible. If not, as above, you run the risk of capturing poor quality unreliable data.
Survey analytic methods such as Choice Based Conjoint can provide fuel for your go-to-market and retention strategies.
They can provide clear direction and
the ability to model endless ‘what if’ scenarios to support decision making and avoid costly missteps. Get in touch to find out more.
