Solving a marketer’s top 3 segmentation dilemmas
Columnist Mary Wallace discusses the three primary reasons market segmentation strategies fail and explains what you can do to solve the problem.
Marketing automation tools empower marketers to create high-return campaigns that have a specific message based on a prospect’s needs. Using targeted messages coupled with strong content, marketers can effectively engage and escort leads down the buying cycle to a sale.
Campaigns are most effective when both content and message speak directly to the buyer’s needs. This means multiple campaigns are essential to address different (specific) needs. Each campaign focuses on leads that are grouped together based on their pain points.
Grouping or segmenting leads is the process of defining and subdividing leads into clearly identifiable slices that have similar needs, wants or demand characteristics. Two market segmentation strategies are using behavioral information (think digital body language) and using demographic information (think company size, industry and so on).
This is such a simple concept, but when the rubber hits the road, building and managing segments is rather complex. Here are three main reasons why segmentation often fails — and what you can do to solve the problem.
1. Lack of a data governance strategy
Data governance is a cornerstone to success in today’s digital world, where marketing automation is king. It comprises a collection of practices, policies and procedures, and it guides how an organization collects, manages and uses data.
Data governance that is forward-thinking and not reactive focuses on ensuring all marketing data is accessible and understandable. Its policies and processes make sure lead data is securely maintained, while its practices make sure this data is structured for analytics.
Without a data governance strategy, the ability to create accurate segments takes a big hit. That’s because it’s difficult to accurately build a segment based on data that might be inaccurate, inconsistent or stored in such a way that it’s almost impossible to interrogate.
What you can do to solve it?
- Define which data elements are needed for your segmentation — You don’t need to group on everything you collect. For example, age may be critical when grouping for car preferences, but not at all for type of business furniture.
- Establish and support the role of a Data Steward — There should be resources responsible for data standardization and segmentation; this can be a single resource, a cross-functional team or a governance body.
- Establish a data management strategy — Build a plan to normalize your database and manage database growth.
- Establish a database health baseline.
- Define and communicate your data processes — Get buy-in on the importance of data health.
- Monitor, learn and adjust — Monitor segmentation effectiveness, email deliverability and opportunity for pipeline growth, and then optimize based on what you learn
2. Data is not standardized
Many times, the data we want to segment on is confusing. Data in fields like industry, job role and company size have no consistency.
Think about a job title, for example: If some leads have data that show CIO, others that show C.I.O. and still others that show Chief Information Officer, it’s nearly impossible to build accurate segments.
Data inconsistencies appear for a variety of reasons. One reason is form submits. For example, leads might enter demographic data in web forms in text fields instead of selecting a value from a drop-down list or from checkboxes.
Another reason is the data source. The origins of lead records might be from third parties, like content syndication. They could also initiate from other systems within the company like a financial or ERP (enterprise resource planning) system.
What can you do to solve it?
- Start with a data review — Identify your current status and determine problematic areas.
- Establish an action plan — Do you need to standardize picklists? Do you need to update your data collection sources such as web forms or CRM (customer relationship management) system?
- Do a one-time data cleanse — This will clean and standardize all existing data based on your new data governance parameters.
- Build a data washing machine process to continuously catch and fix data coming in.
- Modify current input processes for receiving data — Update your web forms; build a data upload template; update picklists used by integrated systems such as your CRM and marketing automation software.
3. Data is not all in one location
On average, an organization’s data volume doubles about every two years. This means that data is not always in one place — not even in the marketing automation platform.
A variety of factors can cause this disenfranchised data: different security requirements; different systems doing different jobs and maintaining only the data associated with the function they assist with (ERP, financial, DMP); costs for maintaining records in some system necessitating archiving data for later use; or integration between two systems that should be communicating failing.
Regardless of the reason, today’s organizations maintain data across a variety of platforms, putting a huge cramp on effective segmentation.
What can you do to solve it?
- Identify the data you need and where it resides — Sometimes there is enough data in one place to do the segmentation.
- Cross-stack segmentation — If you find that data is indeed in multiple systems, you can turn to a new breed of segmentation tools that provides cross-system, or cross-stack, segmentation. Tools like 4Thought Marketing’s 4Segments enable you to segment contacts from one system and move them to another. A side benefit is that you can use data from multiple systems without having to move it all to one large data warehouse.
- Constant review and update — Part of role of the Data Steward mentioned above is to ensure that data sources stay current. You can also assign the job of documenting where the key segmentation data resides and provide documentation to the segmentation users so they know what is available and where it is. This documentation needs to be kept up to date or users will stop relying on the documentation.
Opinions expressed in this article are those of the guest author and not necessarily MarTech. Staff authors are listed here.