Introduction to Segmentation
Audience Segmentation is the process of grouping customers or prospective customers in meaningful ways, for three essential purposes:
To know who to look for and where to look for them
To identify which differentiated segment should be attached to every contact or prospect who crosses paths with your organization
To be equipped to engage with individuals in each segment using methods, messages, and offers tailored to each segment’s top choice (or rejection) factors
There are two approaches to segmentation used for very different purposes: database segmentation and target market segmentation. Target market segmentation typically looks beyond topline demographic, firmographic, and behavioral elements to incorporate emotional, psychometric, and situational elements at play for each segment (sometimes referred to as need states).
Internal Approach: Database Segmentation
Is Database Segmentation Right for Your Organization?
Many organizations maintain extensive database records for their existing customers and leads, including contact information, transactional history, customer service interactions, content engagement (email open/click/etc.), and more.
If your organization has been around for a while, and if your customer base represents a fairly comprehensive subset of market you serve, database segmentation may be all you need to get your arms around the pivotal customer segments you’re serving.
There are several cases where database segmentation is unlikely to serve your organizational goals fully or effectively:
If your organization is fairly new to the marketplace
If your customer data is limited or indirect (filtered through 3rd party distributors, for example)
If your your brand or product category has yet to reach broad market awareness (meaning your database may not reflect the market opportunity at large)
However, if your business is established, your CRM maintained, and your historical share of market appropriate, database segmentation will be the most efficient approach for understanding your customers and anticipating differences among prospective customers (lookalikes).
Database Segmentation Process
Database segmentation usually begins with a process of data enrichment, unless your organization’s CRM practices already incorporate routine data enrichment. Through data appends and/or custom studies with a subset of your database, a more robust and multi-faceted library of data “hooks” form the basis for subsequent analysis and segment modeling. Once data hygiene filters are applied to this enhanced dataset, it is possible to begin building models.
The modeling process often entails two or more analytical approaches, including latent class cluster analysis, 2-step cluster analysis, factor segmentation, and k-means clustering, along with a suite of supportive statistical methods. The selection of the appropriate approaches incorporates end use scenarios, dataset size and structure, and other criteria tailored to organizational objectives.
Delivering Your Database Segmentation Model
In general, we aim to deliver the segmentation model(s) that explains the greatest degree of variation in your database with the fewest number of segments necessary, unless your operations are equipped to handle (and serve) a larger and more detailed set of segments.
Appending Your Database
Now that you have a proprietary segmentation model for understanding and serving your database, it is essential that we make that segmentation insight useful. The most essential way we do that is by appending segment membership to each contact in your database (at least for those where sufficient data exists). For any records lacking necessary data, we will recommend data collection priorities for automated or CSR-led outreach, in order to increase the share of your database properly segmented over time.
Segment Personification Follow-Up
Optionally, we may follow your segmentation study with a qualitative contextualization & personification project, which is essential for adapting objective segments into fully fleshed personas. In this optional step, we recruit valid qualitative samples of database contacts now known to belong to each particular segment (perhaps excluding segments your organization deems as not a priority). A variety of qualitative methodologies are available for this work, and we will recommend the most suitable approach for your database segments that will maximize insight in a time- and budget-efficient manner.
Typing Tool Development
Though not always necessary, some organizations have operations that can benefit from a scalable typing tool — a brief set of criteria for accurately predicting the appropriate segment to apply to new contacts. In some cases, a typing tool may even be used for classifying or reclassifying certain existing contacts as additional data is collected. The abbreviated criteria enable streamlined and standardized classification, in exchange for classification accuracy that is necessarily less than 100%. Our iterative typing tool development process helps ensure you achieve standard 70% to 90% accuracy in your automated classification programs for lead capture programs, subsequent market research projects, or both.
Go-To-Market Approach:
Target Market Segmentation
Is Target Market Segmentation Right for Your Organization?
Many organizations, either new or established, face uncertainty regarding their current reach and share within their addressable market. Perhaps your organization’s revenue is strong and your customer or constituent base is growing, but you just can’t be sure if you’re leaving market share on the table.
Beyond that, perhaps your in-house CRM database is anemic, for any number of technical or historic reasons. You want to be sure you’re building a segmentation understanding that will help you pursue and engage groups and subsets you may not be reaching today:
This is especially important in a couple of scenarios:
If your product offerings have shifted in recent months or years, in ways that could have broadened your potential market resonance
If you know your brand or product category has yet to reach broad market awareness (meaning your database may not reflect the market opportunity at large)
If your customer data is limited or indirect (filtered through 3rd party distributors, for example)
If recent business strategy or capitalization has your organization positioned to invest in growth at a more ambitious scale than previously
In these scenarios, database segmentation won’t be adequate to help you understand the untapped segments with potential. Perhaps more critically, database segmentation alone won’t enable you to identify segments that are likely to be more difficult or costly to engage. Instead, target market segmentation will help you focus your efforts and investment on expanding your reach with current and untapped segments most likely to fuel your growth objectives.
Target Market Segmentation Process
Target market segmentation usually begins with agreement on the full spectrum of individuals that comprise your Total Addressable Market (TAM). This is where we think about the absolute most essential criteria that might qualify someone for your product or services. We’ll work with you to make sure we’re not excluding any potential group unnecessarily.
Once we’ve identified the TAM “universe” criteria, we typically begin with a preliminary qualitative exploration with a range of individuals across that universe. Essentially, we’re surfacing potential hypotheses for demographic, firmographic, psychographic, use-case, lifecycle, transactions, or other items that may be valuable to incorporate into the segmentation study.
With those insights in hand, we develop a thorough, tailored survey instrument to be distributed to a representative or stratified sample of your TAM. We employ a sampling approach tailored to your organization, with sample size determined by the hypothesized items included and budget realities. With sample in hand, we sometimes append additional 3rd party data to enrich modeling and/or assist with data hygiene.
The target segmentation modeling process at this point looks fairly similar to modeling with a customer database. It often entails two or more analytical approaches, including latent class cluster analysis, 2-step cluster analysis, factor segmentation, and k-means clustering, along with a suite of supportive statistical methods. The selection of the appropriate approaches incorporates end use scenarios, dataset size and structure, and other criteria tailored to organizational objectives.
Delivering Your Target Market Segmentation Model
In general, we deliver the target segmentation model(s) that explains the greatest degree of variation in your database with the fewest number of segments required, unless your operations are equipped to handle (and serve) a larger and more detailed set of segments. Our delivered model and accompanying report includes robust overviews of each segment’s identifying or descriptive characteristics, relevant priorities (choice/rejection factors), and if needed, relative size-of-the-prize estimates for each segment. We will also support you in identifying likely top priority and/or de-prioritized segments for you to consider in operational, marketing, sales, R&D, and other functions across your organization. We insist on ensuring our segmentation models are truly actionable.
Segment Personification Follow-Up
As with database segmentation, we may follow your segmentation study with a qualitative contextualization & personification project, which is essential for adapting objective segments into fully fleshed personas. In this optional step, we recruit valid qualitative samples of segment representatives from the quantitative study known to belong to each particular segment (perhaps excluding segments your organization deems as not a priority). A variety of qualitative methodologies are available for this work, and we will recommend the most suitable approach for your budget and goals.
Typing Tool Development
As with database segmentation scenarios, some organizations have operations that can benefit from a scalable typing tool — a brief set of criteria for accurately predicting the appropriate segment to apply to new contacts. The abbreviated criteria enable streamlined and standardized classification, in exchange for classification accuracy that is necessarily less than 100%. Our iterative typing tool development process helps ensure you achieve standard 70% to 90% accuracy in your automated classification program. The options for leveraging a toolkit like this go far beyond standardizing your future research — build it into your lead generation forms, sales intake process, subsequent research recruiting, or other touch points to achieve powerfully segmented access to your customers on an ongoing basis.
With a typing tool in hand, we often assist our clients with applying their target market segmentation model(s) to their database. A variety of factors affect feasibility and ROI, which we will work through with you as you assess potential value for your organization.