Fewer, Bigger, Bolder: From Mindless Expansion to Focused Growth by Sanjay Khosla and Mohanbir Sawhney
April 6, 2014
4 Challenges in Customer Segmentation
November 20, 2008
Customer segmentation is probably the most important concept in marketing. But it is also one the most difficult to do well. In this post, I reflect on four key issues that marketers must address in their segmentation work:
Scope of segmentation – What is the appropriate scope at which segmentation studies should be conducted? Specifically, how do we reconcile the level of depth required by engineering with the breadth required by cross-business alignment?
Inbound versus Outbound Use of Segmentation – How do we reconcile segmentation variables that are useful for inbound product development decisions versus segmentation variables that are useful for outbound product marketing decisions?
Refresh Cycle and Stability of segmentation – How we ensure that segmentation remains relevant between the time the product is initially conceptualized, through development and into go-to-market? How often should we “refresh” segmentation studies to ensure that they are tracking changes in the market structure or demand?
Segmentation for Multiple audiences – how we approach segmentation and targeting for products that cater to multiple audiences?
1: Segmentation Scope:
The broader the scope of segmentation research, the less actionable the results are for specific business groups and products. On the other hand, the more focused the segmentation research, the less the “interoperability” and reusability of the research for other business/product contexts. If the scope is defined too broadly, you can end up with a very generic segmentation approach. At the other extreme, segmentation of a very narrowly defined audience will lead to very precise segments, but it would be of no relevance for other products/markets.
Specific questions that you need to think about related to scope:
What are the different levels of scope at which segmentation is currently done?
How “interoperable” are the segments/segmentation variables in segmentation studies done for different products/business groups?
How feasible is it to “standardize” segmentation across business groups/products?
What is the “sweet spot” in scope between too broad and too narrow?
What can we learn from other companies or other industries in terms of best practices?
2: Inbound vs. Outbound Segmentation
When segmentation is done at the inbound stage to guide engineering decisions, it will emphasize variables such as:
Depth and breadth of product use
On the other hand, when segmentation is done at the outbound stage to guide marketingdecisions, it will emphasize variables such as:
Goals and outcomes
Additionally, segmentation done at the outbound stage will tend to emphasize actionability in terms of the ability to reach customers through marketing communications. And it will need to include various players in the Decision Making Unit, such as the economic buyer, technical buyer, end user etc. On the other hand, segmentation for product development may focus on how the product will get used, deployed, managed and supported. A key issue is to reconcile these two approaches to segmentation, and to decide how the segmentation approach can be “transitioned” over the product lifecycle so that there is a consistent view of segments and value propositions.
Specific questions on this issue include:
How is segmentation currently done at the inbound and outbound stages?
What are the uses of segmentation at the inbound stage? What information is most valuable to developers?
What are the uses of segmentation at the outbound stage? What information is most valuable to marketers?
Can these two sets of variables be reconciled? Can we have one “end-to-end” segmentation approach? Or can we at least think about “transition” from the inbound to the outbound segmentation approach as the product progresses through the product lifecytlce?
3 – Stability and Refresh Frequency of Segmentation
Segmentation studies eventually become obsolete. This may be due the evolution of the market (from early adopters to the mainstream market), economic changes (affecting price-sensitivity), to disruptive market redefinition (e.g., the iPhone and iPod Touch). This raises an important question – how frequently should segmentation research be revised or refreshed? Should it be every time a new product is introduced (3-4 years), or sooner?
Specific questions to think about:
How frequently are your segmentation studies currently refreshed/updated?
How does this frequency vary with type of business or purpose of segmentation?
What is the “optimal” refresh cycle for segmentation studies that will be a good compromise between the cost to refresh segmentation research and the risk of using outdated segmentation research for marketing/development decisions? How will this vary by product, business, market or geography?
How can you reduce the cost of refreshing segmentation research (possibly using online data collection on a continuous basis)?
4 – Segmentation for Multiple Audiences:
Several products (like Microsoft Windows and Microsoft Office) are purchased and used by multiple audiences. This suggests that segmentation of these audiences should be common or at least related. Yet, the products are used for different purposes, so it is not essential that a “power user” of Windows will also be a “Power User’ of Office. And it is not essential that we the same variables will even be relevant. So, to what extent can segmentation approaches be made consistent across products?
Specific questions to consider on this issue:
How different is the segmentation approach used for different products aimed at overlapping audiences?
How different or similar are the variables that are used for segmentation of products aimed at multiple audiences?
What are the most useful variables that can become the basis for segmentation of products aimed at multiple audiences? These might include simpler/more actionable variables like industry, company size and demographics versus more complex/more meaningful variables like lifestyle, behaviors, needs etc.