Choice Modeling: Polishing the Crystal Ball


When it comes to designing products that customers will buy in droves, the stakes have never been higher.  Despite billions of dollars of investment and countless hours of R&D, 90% of all new product launches will fail within 3 years of hitting the market. Furthermore, many products live as ‘walking wounded’, suffering from low market share, profitability and market differentiation.  The challenge is deceptively simple:  what is the ideal product that balances customer appeal, product profitability and supply chain fit?  The answer can be devilishly complicated, given the myriad of product combinations that could be delivered through different supply chains and sold in a range of markets.  

Fortunately, there are powerful, new analytical tools and methodologies that can help.   One of these techniques, Choice Modeling (CM), can improve new product success rates,  reduce business risk, increase customer knowledge and help define the optimal combination of features, services and prices for existing products. (Another powerful tool is CRM-driven data analytics)

Choice Modeling

As an approach, CM is part science and part art.  CM uses high-performance computer simulations and econometrics to understand and predict customer choice under various product configurations, market and environmental conditions.  In the past, marketers could only rely on simple statistical tools like regression analysis to understand a small set of cause and effect relationships between variables.  Thanks to CM, a firm can now dramatically accelerate the scope, depth and speed of their product analytical capabilities.

CM is being used to design products, services and supply chains in a wide variety of industries including consumer & industrial goods, financial services, hospitality, telecom and retail.

Using Choice Modeling

There are 3 basic steps to utilizing CM:

  1. The first step identifies the number of possible product, service or experiential features  (choices) that could influence a customer’s preference for your offering.  For a new car, the choices would include color, engine size, sales experience and options.  Information on choices can be gleaned from many sources including current product information, customer interviews, surveys and industry data.
  2. The second step is where the art comes in.  Marketers would design a series of simulations that ask customers to choose between a small number of choice options within a series of choice sets.  Using the car example, a simulation could be designed that asks customers to make choices between 2 different luxury packages (choice options) within a series of different feature collections (choice sets).
  3. The final step is where the science takes over.  Powerful econometric models are applied to a representative sample of respondents to identify empirical relationships between their selections of choice options and choice sets.  Unlike traditional tools, CM allows marketers to rapidly model and understand the relationships between hundreds of choices in hundreds of scenarios. Back to the car analogy, analysts would be able to test the impact of various option packages with different features on market share, segment profitability and customer satisfaction, before finalizing the product design and without guessing.

Poised for Growth

With the penetration of Web 2.0 technologies and higher bandwidth, it is now feasible to quickly gather key data and run simulations across multiple geographies, regulatory environments and customer segments.  Importantly, designers can now model unique and customized solutions to individual respondents or micro-segments using new advances in Bayesian statistics.

A Final Caveat

Like other analytical tools, CM is susceptible to “garbage in, garbage out” effects.  Problematic data, shaky assumptions and poorly designed simulations will inevitably lead to misleading results.  Furthermore, the most powerful CM models will not overcome incorrect findings arising from organizational effects like management bias or cultural influences.

 For more information on services and work, please visit the Quanta Consulting Inc. web site.

Advertisements

7 comments so far

  1. IP network camera on

    You certainly deserve a round of applause for your post and more specifically, your blog in general. Very high quality material.

  2. IP network camera on

    Hi, can I quote some of the content found in this entry if I provide a link back to your site?

  3. IP camera on

    Nice piece of facts that you’ve got in this website submit. Hope I can get some much more of this stuff on your weblog. I’ll occur back.

  4. IP network camera on

    I am so happy to read this. This is the kind of info that needs to be given and not the random misinformation that is at the other blogs. Appreciate your sharing this beneficial content.

  5. IP network camera on

    Thanks for the informative article, it was a good read and I hope its ok that I share this with some facebook friends. Thanks.

  6. […] and enables supply chain efficiencies. These other components include data visualization, choice modeling & mathematical optimization and simulation & scenario […]

  7. […] In these difficult times, gaining superior utility and value (including but not limited to low cost) is top mind of most consumers, even in premium categories. Yet, many companies have a poor understanding of their consumer’s core functional and emotional needs (i.e. the job to be done) and are unable to prioritize these needs against their product roadmaps and capital spend.  Gaining a deeper, more holistic understanding of consumers and the trade-offs they make oblige managers to go beyond basic research techniques to include analytical tools such as ethnography and choice modeling. […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: