Why Personalization Matters

Companies are facing a challenging time today wrestling with the requirement to personalize communications and offers.  What’s more, personalization is about much more than segmentation, it’s about calibrating offers to a level of detail only achieved by understanding the consumer’s mindset and behavior at the point in time when interacting with your brand. Three articles from MarketingProfs.com discuss personalization from the challenge of accomplishing real time personalization to how to implement radical personalization and the challenges of personalization in digital marketing. Enjoy! Marketers Struggling With Real-Time Personalization Understanding the Five Pillars of Radical Personalization Marketers’ Biggest Obstacles to Effective Personalization   Tell Me What You Think! DBMCatalyst

Real-time Customer Segmentation

I’ve been around long enough to have heard the term “real-time” associated with a variety of marketing technologies, how about you?  And I often wonder about the value of that type of immediacy let alone the possibility of actually interacting in real time with my customers.  It’s always been one of those marketing aspirations which seems never quite actualized.  Or at best, it’s realized in a very mechanical way, such as re-directing a consumer to a new offer branch in a campaign based on their real-time response in a campaign.  All of which is better by far than the “blast” form of email or direct mail – which is still practiced in many industries.

In my opinion, the only value in “real-time” is knowing who your customer is in that moment as they interact with your brand.  And by “who” I really mean not only their demographics or psychographics but also their past behaviors and transactions all of which tell me something about their capacity and propensity to buy as well as their interest in/affinity towards my product.  When properly modeled the result is a score that  effectively segments your customer according to their lifetime value to your brand which enables you to deliver the most personalized, effective offer at the most profitable price.  This is real-time segmentation.

There are numerous approaches to segmentation from the simple demographic/psychographic classifications to sub-segmentations which employ sophisticated modeling techniques like factor and cluster analysis.   Ultimately most segmentation models give you a piece of the picture but even when they come close to the full definition of “who” the customer is, the result is in the “rear-view.”  The learnings  are applied after the fact, in a sense they describe who your customer “was.”   The moment of connection with your customer is lost or delayed.  The most sophisticated companies will rigorously update their customer segmentation models, but again, they are not describing the customer in “real-time” purchasing mode.

Real-time segmentation and scoring is possible today and will soon be enhancing the customer experience in not only call centers and online, but also in social and mobile channels.

Ask me about how!

Ann McCartan, DBMCatalyst

Check into Intelligent Handshake




Mini-Segmentation, what can you do?

We’ve spoken about the practice of aggregating like customers into groups or segments to better target and market to them.  And how to utilize modeling – factor and clusters – to distinguish individual characteristics within the segment they are assigned to.  These are complex segmentations built with a lot of data on each customer.

Another approach is more organic and allows the marketer to build segmentation profiles a step at a time as more data is obtained.  The good news is that content and messaging can be created at each step based on specific information gathered.  The goal is an upward journey to 1 – 1 personalization and a true 2-way conversation but these mini-segments make effective communications possible.

You don’t have to build the whole segmentation before marketing effectively to your customers.

Tell Me What You Think!

Ann McCartan, DBMCatalyst


B2B Segmentation – the impact of behavior

Just a few more thoughts about segmentation…B2B segmentation specifically.  Segmenting B2B customers is both simple and more complex at the same time.  Simple because of smaller datasets and widely conforming variables, but more complex because of a greater number of disparate data sources – and aligning data definitions for each – and because of the very real problem of defining the “customer.”  Mostly I’ve seen companies overcome the above complexities in all but one regard – properly parsing and modeling customer behavior.

What does that mean?  consider this:  how often and how easily are customer tenure, i.e., length of time as a customer; specific product or service usage – i.e., name, SKU, quantity; and customer recency, i.e., last purchase date/ship date,  tracked and added to a customer or analytics database?  How many years of data are combined and stored in such a database:  1 year? 3 years? 5 years?  Modeled correctly, these few variables will provide actionable insight into your B2B customers’ behavior and when added to their firmographics profiles will impact the relevancy of communications, offers, promotions and service-calls.

Most of the B2B segmentation I’ve seen is based on a combination of firmographics and revenue, some even based on year-over-year revenue or 3 years of revenue.  Or sometimes the segmentation is based on product profiling.  But rarely are they enriched with behavioral data such as defined above.  Why care?  because this means that the company will forever be stuck knowing what their customers “did” but not what the customer will “do next.”

Wasn’t it Wayne Gretsky that said “I skate to where the puck will be not where it has been.”

Properly combined with customer firmographics and modeled for predictive outcomes B2B customer behavior will have a powerful impact on future growth.

Tell Me What You Think!

Ann McCartan, DBMCatalyst




Sub-segmentation – the power of relevancy

I recently read a research paper “Customization with cross-basis sub-segmentation”  which got me thinking about the real power of sub-segmentation.  And I’m betting that while segmentation is a common tool used by marketers, sub-segmentation isn’t widely understood or used.   Do read this article.  It’s an older article and a bit dense but the advice is timeless.

I’ve been involved with many companies who routinely group customers by homogeneous variables such as spend, profit, geography, trade-area, industry, product purchases or whatever variables are key to their business.  All good and prerequisite to effective offer targeting, but ultimately descriptive, i.e., who they are, what they have purchased but not how they act in the future.  Through various types of modeling customers can be scored as more or less responsive in different marketing channels – either to click or call or visit a store and/or to buy what you are selling.  But that only tells you WHO to market to.  I have to say that I, too, believed at one time that knowing WHO was enough.

The interesting piece, and, yes, I know, not new news, is raising the bar through offer customization.  Customization enables relevance.  Simple.  But what I want to discuss here is the effective customization generated by a process known as sub-segmentation.  And further that it’s quite likely that sub-segments even within one group (segment) can often act differently, display different buying behavior and responded differently to offers that you believe are germane to the whole group.    Through sub-segmentation it is possible to tie previous purchase behavior so closely to each customer that it becomes virtually predictive and definitely more relevant.  And, that’s WHAT to say.

Let’s look at two sub-segmentations, RFM (recency, frequency, monetary) and the development of interest scales.  RFM is a well-known, often utilized method of segmenting customers based on purchase recency – how long since their last purchase; frequency – the number of previous purchases; and monetary – the total amount spent by the customer in any timeframe but often either over their tenure or over the last twelve months.   Once customers are sub-segmented by this method a customer contact strategy becomes apparent, e.g., best customer (retention and loyalty), growth customer (upsell), and maintenance (recent buyer, new customer.)  But again, all best customers or new customers are not equally likely to respond to one offer message even when targeted to their segment.  RFM offers little relevance.

The development of interest scales requires more sophisticated modeling.  It typically involves measurement of how often the customer buys various product categories, the ranking of each customers’ interests into a scale and uses this analysis to create sub-segments.  So the frequent shopper of books online – who also slots into a single segment based on the types of characteristics above – proves to be not just a book lover but a devotee of Science books based his interest scale.  He may also buy other book categories, but buys Science considerably more often.  It is also likely that the Science category far outstrips all other categories.

Of course the development of the interest scale isn’t so simple and requires sophisticated analytic tools like factor analysis which enables the identification of the purchased products and cluster analysis which is used to group products in such a way that products in an assigned category are more similar to each other than to those in other categories, thus reducing the overall product set.  And, of course, scoring customers accordingly.

By combining the two sub-segmentation approaches, you become able to communicate more relevant offers, speak more knowledgeably to your customer about things he in interested in and convert more sales in the process.

Let Me Know What You Think!

Ann McCartan, DBMCatalyst

Segmentation, a Deeper Look

Recently I posted a brief, basic explanation of Segmentation.  I mentioned some of the most common applications for segmenting customers for marketing purposes.  Clearly the more narrowly customers are grouped by commonalities, the more able the Marketer is to direct content and offers to meet their particular interests or behaviors.  It’s basic targeting, yes?  and the ideal outcome is always greater incremental revenue, profits and the sense by the customer that the company understands them and is providing relevant messages.   As a first step, I always look to customer characteristics and behaviors as prerequisite to defining segments.

As a second step in the process, before a customer segment is acted upon it should meet several other business-focused criteria.

  • It is possible to measure.
  • It must be large enough to earn profit.
  • It must be stable enough that it does not vanish after some time.
  • It is possible to reach potential customers via the organization’s promotion and distribution channel.
  • It is internally homogeneous (potential customers in the same segment prefer the same product qualities).
  • It is externally heterogeneous, that is, potential customers from different segments have different quality preferences.
  • It responds consistently to a given market stimulus.
  • It can be reached by market intervention in a cost-effective manner.
  • It is useful in deciding on the marketing mix.

If the business can identify distinguishable customer differences and commonalities which can be can justified from a marketing and financial perspective then solid customer segments can be derived.

Tell Me What You Think!

Ann McCartan, DBMCatalyst




Customer segmentation is a major factor in establishing appropriate, successful customer interactions.  Knowing how customers are different or the same – and their latent value – enables you to plan, cost, and execute campaigns to drive the highest response, conversion and ROI.

Segmentation models can be based on simple demographics such as industry, geography or number of employees. More detailed groupings can be developed by using customer spend, profit, store trips or products purchased.

Segmentation will provide a basis for understanding the overall value of each customer, as well as impacting the assignment of each to the proper customer strategy:  Retain, Grow, Maintain or Defect.  Segmentation enables you to:

  • Target segments according to their potential and the company’s ability to serve them in a proprietary way;
  • Invest resources to tailor product, service, marketing and distribution programs to match the needs of each target segment;
  • Measure performance of each segment and adjust the segmentation approach over time as market conditions change decision making throughout the organization.

Contact Me at 339-227-7591 or ann@dbmcatalyst.com

Building Customer Insight

I didn’t write the article I’m going to present, but I could have.  Which is not to compare or detract from Joseph M. DeCosmo’s wonderfully concise “survey course” in using customer insights to target relevant messages to the most receptive customers or prospects.  The Goal: make money from your customer data, right?

I could have written it because I have experienced every step along the path he describes both as an employee of or consultant to large corporations who have decided that customer data is the key to relevant marketing.   [Read more…]

B2B Marketing: Digging Into Customer Behaviors, Segmentation and Three Top Benefits

I’ve recently read about new trends in B2B Marketing:  the use of behavioral analytics, psychographic segmentation and data overlays.   (B2B Magazine, “Transformation via Sophistication” ) There’s no doubt that behavioral tracking and analysis shines a light on the pathways that lead from research to interest to engagement and on to conversion.  The trick is understanding which activities most strongly correlate to the ultimate conversion process and how to move the prospective buyer from point A to point “C.”   And when to disengage from prospects, i.e., Visitors, who endlessly participate in downloads, webinars and even frequent site click throughs without any intention of purchase.  The cost to maintain interaction with “Visitor” isn’t insignificant. [Read more…]