In my 30+ years of data-driven marketing experience, I can make the following statement with the utmost confidence: Combining customer segments, third party data and predictive modeling will significantly improve your marketing results.
Most companies use one.
Some use two.
Far fewer use all three.
When you combine all three…customer segments, 3rd party data, and predictive models…you will achieve significant gains in your marketing results.
In this primer, I will discuss all three and how you can combine them to optimize your results. I’ll conclude with specific recommendations to help you achieve more, regardless of your starting point.
Let’s get started.
I – Market Segmentation
Definition: “Market segmentation is the subdividing of a market into homogeneous subsections of customers, where any subsection may conceivably be selected as a market target to be reached with a distinct marketing mix.” (Philip Kotler)
Note: The words “cluster,” “cohort,” “target group” are often used synonymously with “segment.” For the purposes of consistency and clarity, I will use “segment” throughout this blog post.
It is impossible and impractical to target the entire market. So, segmentation allows you to:
- Prioritize your efforts and customize your products, pricing, and marketing, so you can…
- Provide your customers and prospects with more relevant and desirable offers, so they will…
- Buy more products at higher prices with greater long-term retention and profitability.
Clearly, segmentation is a vital part of every modern marketer’s toolbox. There are several ways to segment your market or customer base, including:
- Demographic Segmentation: In this approach, you segment your market using demographic variables such as age, gender, children, education, assets, income and more. Using this approach, marketers focus their attention on demographic segments that generate disproportinately large percentages of sales and profits.
- Psychographic Segmentation: In this approach, you segment your market using lifestyles, lifestages and personalities. These segments are often based on the interests, opinions and activities of customers. When psychographic segments are closely correlated to spending in your category, they can be very useful in your marketing.
- Behavioral Segmentation: In this approach, you segment customers based on their unique spending patterns, uses, preferences or responses. In many cases, the actual purchase histories and brand preferences are combined with psychographics and demographics to create the behavioral segments. Marketers often use this approach to target non-purchasers of their product who appear to be classified as a heavy spender for their product.
- Benefit Segmentation: In this approach, you segment customers based on the unique wants and needs they are trying to satisfy using your product. Different segments will have different wants and needs that drive product/brand preference and purchase. In addition to targeting and marketing benefits, this segmentation approach helps marketers develop winning products and features.
II – Consumer Data: third party
While segmentation holds great power for marketers, you often need to target your customers and prospects more granularly. This is where consumer data from compiled sources comes to the rescue.
Third party data, once appended to your files, improves virtually every aspect of your research, product development and target marketing. This includes customer acquisition, cross-selling and retention.
By using consumer data from many different sources, marketers tap a wealth of powerful marketing insights. Compiled consumer data consists of the following types of information:
- Demographic
- Geographic and Proximity to Competitors
- Lifestyle
- Lifecycles and Generations
- Psychographic
- Financial and Credit
- Purchase History
- Purchase Propensity
- Channel Preference
- Brand Preference
- Ethnicity and Language Preference
- Medical and Ailments
The sources for compiled data include consumer transactions, public sources, online surveys, telephone directories, auto registrations, warranty cards, sweepstakes and predictive models. Consumer compiled data can be classified into specialty groups or lists such as newly married, ethnic groups, or new parents. These consumer specialty lists help niche marketers who precisely target their prospects.
When combined with segmentation, consumer data helps you:
- Identify sub-segments of prospects
- Further refine copy and creative
- Customize offers and pricing based on financial and credit profiles
III – Predictive Modeling
Segmentation and consumer data are vital to your marketing success. But, you can go farther and faster.
Predictive modeling gives you the unique ability to anticipate a specific prospect or customer need, behavior, or preference before it happens. This allows you to communicate with each person at the right time, with the right offer and using the right message.
Predictive modeling uses behavioral data, consumer data and modeling techniques to predict the future behavior of an individual or household.
For example, predictive modeling is often used to predict the likelihood that a customer will make a purchase. Many companies have a portfolio of predictive models that are optimized for:
- Customer acquisition
- Campaign response
- Loyalty and share of wallet
- Cross-selling and product preference
- Market basket analysis
- Customer retention
It is not uncommon for our predictive models to use hundreds of data points to generate the most reliable models for scoring a prospect or customer file for marketing.
Combining Customer Segmentation, Consumer Data & Predictive Analytics
We have established the power and importance of segmentation, consumer data and predictive modeling. Now, let’s discuss how you can combine them for maximum impact.
While there are dozens of ways to leverage the combination, I will touch on just two:
- Segment-specific models
- Segment/data-specific communications
Segment-Specific Modeling
Many companies develop a single predictive model and deploy it across their entire customer or prospect base, regardless of their segments. While this gives them better results vs. the absence of modeling, they are leaving a ton of lift on the table. Enter segment-specific modeling.
Segment-specific modeling refers to the process of developing different predictive models for (or within) each customer segment. The objective is to improve your marketing results by developing models that are more predictive and result in higher response and conversion rates.
While segment-specific modeling is far more sophisticated, the marketing improvements can be eye-popping. It’s not uncommon for this approach to yields dozens or hundreds of different models vs. a single predictive model. By using the right model for the right segment, predictive accuracy soars, and so does your campaign performance and marketing ROI.
Segment/Data-Specific Communications
Many companies deploy models to score each customer’s likelihood to buy a product or respond to an offer or campaign. But many companies fail to customize those offers. Enter segment/data-specific communications.
Segment/data-specific communications refer to the process of customizing your communications to each customer based on their segments and/or detailed data profiles.
Again, while segment/data-specific communications are far more sophisticated, the marketing improvements are dramatic. Customers respond far better to creative, copy and offers that resonate with their unique personal profiles. We all like it when a company caters to our unique desires and preferences. Your customers are no different.
Concluding Recommendations
I hope you found this primer helpful.
Depending on where you are today, here are some quick recommendations to help you achieve better results.
Market Segmentation
- If you already have a segmentation scheme, use it. Make sure you have deployed it at the customer-specific level. In other words, build a classification model (using behavior and consumer data) to place each customer and prospect into their best segment. Without this, your segmentation scheme is useless for direct marketing purposes. If you don’t know how to develop a scoring model, contact us for ideas.
- If you do not have a segmentation scheme, get one. For starters, you can use one of several syndicated segmentation schemes that we offer through our DecisionPoints consumer data offering. Custom segmentation is also an option to consider.
Third Party Data
- If you have already appended consumer data to your file, use it. Use the data to improve your communications and to develop better predictive models. Use the data to build a segmentation scoring scheme. Use the data to improve your market research and product development. Use it throughout every stage of your marketing process. Also, make sure you are conducting robust data hygiene before you append consumer data, otherwise you will have significant errors in 20-30% of your data. Don’t make that mistake.
- If you have not appended 3rd party data to your file, get started today. The use of supporting data will have an instant impact on your marketing results. Our DecisionPoints consumer data is the leading solution in the market, with hundreds of data points on over 270 million consumers in the U.S.
Predictive Modeling
- If you already use predictive modeling to enhance your marketing, consider using segment-specific models to take your results to the next level. In addition, use consumer data to enhance the performance of your models and/or to extend their use to prospect files for acquisition campaigns.
- If you aren’t using predictive modeling, get started today. Make sure you start with a clean data source that uses behavioral and appended consumer data. Also, make sure you fully test your models before widespread deployment.
Combining All Three
- If you are already combining all three, congratulations!
- If you are not combining all three, get started today. They are far more powerful together, I guarantee it. Let me know if I can help.