What is customer segmentation? Customer segmentation is broadly defined as the process of dividing customers into groups based on common characteristics so companies can engage, market, reach and communicate to each group more effectively and appropriately. While there are many kinds of customer segmentation models, including device, referring source, B2B, etc., the traditional “birds of a feather” description generally segments customers four different ways: demographic (age, gender, income), geographic (ZIP, city, urban or rural), behavioral (purchase or spending habits) and psychographic (interests, attitudes, values).
Customer segmentation is defined, at its core, as finding similarities in customer data and exploiting those similarities to achieve business goals. Acquisition, retention, loyalty, and CLV are among the goals of discovering and marketing to customer segments based on similar situations, interests, location and/or behaviors.
CRM Segmentation Models and “Zero Segment Marketing”
Because a personalized customer experience has been shown to drive new revenue, there is one school of thought that personalization supersedes segmentation; if marketing to a customer based on gender or income is not relevant to a customer’s journey, the thinking goes, the only way to provide a relevant, personalized experience throughout a journey is with a “zero segment marketing approach”.
While it is certainly true that a personalized customer experience drives revenue and is central to a brand’s ability to acquire and retain high-value, loyal customers, what this school of thinking confuses is that personalization is not a zero-sum game. Rather, it occurs within segmentation models. The term zero segment, then, is somewhat misleading; it’s not that there are no segments, it’s that they are so granular that they no longer fit into neatly defined delineations. Age, gender and income may be a starting point, but technology now makes it possible to segment an audience 100 or more ways for a far more accurate “birds of a feather” approach.
Customer segmentation models without channel restrictions enhances personalization for the simple fact that engagements with an audience will always be relevant for each customer within a segment in any channel – or more precisely wherever a customer happens to be in an omnichannel customer journey. Customer segmentation analysis will provide additional visibility into which channels are used most often by your defined customer segments.
Benefits & Limitations of Traditional Customer Segmentation
When customer journeys were predictable, linear (research, evaluation, purchase) and limited to a few mostly physical channels, customer segmentation and personalization were for the most part thought of as identical. A sporting goods store in Cleveland with excess inventory of Cleveland Browns game jerseys might create a segment of 25-to-34-year-old mails and target that customer segment with a promotional email based on an assumption that young men are more likely to be football fans. The segment may have even been created based on some rudimentary analytics.
This type of segmentation was thought of as “personalized” – and to those who clicked on the email and redeemed the offer, perhaps it was. Maybe the campaign even outperformed its open-rate average.
For that type of segmentation, false assumptions were accepted as the cost of doing business. I might live near the store and fall into the age demographic, but maybe I despise the Browns because they never make the playoffs. Or maybe I’m a Cleveland transplant, and I’m a die-hard Bears fan. Or perhaps I just saved up and paid full price for a home jersey of my favorite player and receive the email a week later once the jersey’s already in my closet.
Customer segmentation modeling done the right way can still be effective, but dynamic, omnichannel customer journeys require far greater sophistication to ensure relevance and to guard against introducing friction into a customer’s holistic experience with a brand.
Adtech Customer Segmentation
The sporting goods example describes a typical martech engagement using basic customer segmentation analysis. Adtech customer segmentation works in much the same way. The sporting goods store might even outperform an email blast and secure a higher ROI with a paid TV advertisement for the same promotion that airs during a Browns’ game, putting the offer in front of the eyeballs of fans who are actively watching the game. A diehard Bears fan or someone who has given up on the Browns championship hopes is probably not watching. In the digital advertising world, an ad placed on ESPN.com could be segment-based by running the ad on the devices of known customers in that age demographic (assuming you have house-holding done right).
The downside is a comparative lack of tracking compared with an outbound email campaign, and less of a personalized experience relevant to a customer or prospect’s relationship with the brand. The store might run the TV ad on consecutive Sundays, or until the excess inventory is sold, but really won’t have true insight into the ad’s effect on KPIs such as retention or CLV. Likewise, the digital ad campaign on a single channel might not be personalized or relevant to a customer’s experience across every channel with the brand.
Cross-Channel Customer Segmentation
One reason for the misconception about the synergy between segmentation and personalization is that marketers struggle to create a segment that can be used across channels. Traditionally, siloed data and siloed channels prevent an audience segment created for the purposes of, say, an email marketing campaign to be carried through to visitors to a landing page. Hence, many marketers have the mistaken belief that personalization must be bound by the same restriction, and they fail to grasp its power to create a holistic, omnichannel experience.
With cross-channel audience segmentation, omnichannel personalization is enhanced because restrictions are lifted. Dynamic audience segmentation that is unbound by channel is the Redpoint differentiator with the rg1 solution. Once a segment is defined and created, it is universally applicable to a campaign on any channel dynamically. This is really what is meant by “zero segment marketing”. It’s not just that micro-segments are far more granular, it’s also that there are fewer restrictions because the customer segment is unbound by channel. Zero limitations is a better description; removing siloes and channel restrictions allows for personalization within a segment throughout an omnichannel journey. This means that an audience can be segmented almost infinitely precisely because selection is applied dynamically.
Rules-Based Customer Segmentation
With next-level customer segmentation with Redpoint, audience selection is not only unbound by channel, dynamic selection also means that segments are rules-based, not list-based. As customers move through a customer journey, various rules apply depending on an action taken or not taken, or the inbound or outbound channel of engagement. A customer, then, moves in and out of a segment as a journey is ongoing.
Consider what this means for offering a relevant experience. A basic customer data platform (CDP) may claim to offer segmentation, but once a list-based audience segment is created it is usually exported to a CSV file, uploaded to an email service provider or another channel of engagement, and sent from there. It is not difficult to see the fault; the moment the list is created its relevance wanes because it does not accurately capture the entire audience – either those leaving or joining – as the audience interacts with a brand across channels. A simple example is receiving an email promotion for a product a customer may have just purchased on the mobile app. Enhanced customer segment analysis is therefore required to improve the experience along the customer journey.
Customer Segment Validation
Regardless of customer segment size – from a large cohort to a micro-segment of one – testing and validation are key to ensuring the accuracy of predictions. Testing confirms whether you’re engaging with a customer segment in a way that is not just relevant a specific point in a customer journey, but relevant in terms of what is most likely to trigger a positive response. Testing is a prioritization mechanism; a customer Golden Record provides a single view of the customer – preferences, behaviors, transactions, devices, IDs, social, etc. – and as customer segments before more defined, stringent testing will reveal which behavior, interest, trait or combination of traits form the best response predictor. This is true for any channel, particularly as the customer segment changes throughout a customer journey.
Next-Level Segmentation with AML
Automated machine learning (AML) makes dynamic audience selection across channels possible, with algorithms that render a next-best action or offer for a customer according to the rules in place in whatever direction the customer journey unfolds. AML is the key to finding granular segments for analysis at scale. A manual selection of cohorts is usually fairly basic; with AML, marketers can drill down to find and create segments that are better aligned to profitability goals – members most likely to respond to an offer, most likely to churn, etc.
The Redpoint AML solution facilitates audience selection with visualizations, Venn diagrams that find and narrow an audience from nearly any number of sets of data points. This capability puts audience selection in the hands of the everyday marketer in a sandbox environment without having to pull from an information database or complicated tables or spreadsheets. Clustered audiences is another capability of the Redpoint AML solution, enabling companies to identify sub-segments within a larger set of customers – discerning granular patterns that would otherwise remain unknown left to the devices of the everyday marketer.
The misconception that customer segmentation modeling is going by the wayside, replaced by the urgent need for a personalized omnichannel customer experience, is understandable to a degree. Dynamic audience selection immediately and universally available and applicable to all channels, governed by rules that change an audience in real time according to the vagaries of a customer journey, is heady stuff. Hiding the complexity with AML models that create and segment audiences on the fly is so powerful it almost seems segmentation isn’t happening at all. In reality, it allows for an unparalleled personalized and relevant experience across the entire customer journey. That’s the power of Redpoint. It’s why ambitious marketers choose Redpoint to help them lead markets.
Frequently Asked Questions
Many companies define a 360-degree view of an individual customer as that customer’s golden record, because of the robustness and depth of the data. Having a continuously cleansed, correlated, merged, and up-to-date view of the customer based on data from across systems and channels—a golden record—is a must for accurately predicting customers’ preference and actions and for delivering relevant messaging and experiences.
Having a 360-degree view of the customer means having real-time access to everything that’s knowable about each customer in one place; that is, maintaining what some companies call a “golden record” for each customer. The data that comprises a single view of the customer will be unique to you based on what’s important to your business. It should include first-, second-, and third-party data from internal and external systems and channels, such as:
- Call center transcripts
- CRM systems
- Loyalty programs
- Preference centers
- POS systems
- Social media
- Website activity and behavior data, including shopping cart data
These data sources may be batch or streaming; both will support real-time data access.
Customers are increasingly demanding more personalized offers from the brands they engage with. To react, brands must be able to segment their customer base accuately. As the demands for personalized attention increases, segments are getting smaller, and many brands now treat each individual customer as a segment. This level of segmentation is known as one-to-one marketing. Similarly, marketers may have goals related to acquisition, retention, loyalty, and CLV – accurately segmenting your audiences directly supports the attainment of these goals.
A customer profile is a 360 degree view of every engagement a customer has had with your brand. To achieve this, data is collected and collated on each individual customer by integrating all sources of customer data into a Customer Data Platform (CDP). The output from a CDP is a Golden Record, or single view of the customer, that can be used for personalized marketing purposes.
Redpoint’s Digital Advertising & Acquisition solutions enable companies to optimize their ad campaign targeting and media spend and improve overall workflow efficiencies. This enables companies to manage all their customer and prospect data – as well as information purchased from data providers – in a single environment which, in turn, can be used to orchestrate advanced customer journeys that include digital advertising campaigns.
Customer centricity is a strategic approach to customer experience, leveraging data to provide highly relevant experiences across the entire lifecycle. Customer centricity is the idea that every moment of interaction a consumer has with your brand is extremely relevant and that gaining insight into customer context and cadence enables you to maintain that level of hyper-personalization. Ultimately, it’s a value exchange: if you treat each individual as a segment of one, they will reward you in revenue. Prioritizing a customer-centric approach to customer experience (CX) increases acquisition, retention, and lifetime value.
- Know all That’s Knowable About the Customer: personalize engagement with meaningful, reliable data from multiple systems into a single view of the customer – a Golden Record – that’s more accurate, more complete and more contextual than the data from any single source.
- Customer Profile View: a web user interface that displays key data such as personally identifiable information, demographic and behavioral information, and metrics like web visits, transaction history, customer lifetime value and more.
- Clienteling: the Single View of the Customer is a dynamic view—as customer preferences and transaction history change, the record changes as well. View previous campaigns and drill into content to view offers and treatments and gain insights into how to take the next best action with product recommendations.
- Improve Customer Lifetime Value and Revenue: know what the customer wants through an always on and always updating Golden Record that gives product recommendations that align with customer expectations.
- Increased Personalization: gather a precise customer view that gives you the information to have meaningful conversations with your customers, reducing frustration and friction, and providing a clienteling experience that complements omnichannel customer journeys.
- Reduced cost and improved profitability: maximize ROI for your best customers by understanding and optimizing their journey with up-to-date and accurate information.”
Identity resolution is the method of ingesting all types and sources of customer data, processing that data, and producing a full representation of individual customers. This capability increases customer understanding across online and offline touchpoints – and anonymous to known stages – which fuels data-driven experiences that customers value. While identity resolution is sometimes misunderstood to mean finding a name, contact details, or even a person, its primary goal is to gain a better understanding of customers by reconciling different records for a customer across different engagement systems. Marketers must resolve identity proxies across different systems while faced with an explosion of data sources and types – streaming, batch, structured, semi-structured, unstructured and third-party data – all of which are potential sources of customer data to be reconciled across different engagement systems. An obstacle to getting there is trying to process and match all customer data in the timeframe needed to be able to engage with a customer at their pace and speed. A brand that can gauge positive social sentiment and link it via a device identifier to a recent e-commerce transaction can proactively shape a customer engagement in a segment-of-one approach – but only if it can do so according to the customer’s timetable. Data matching at the pace of the customer adds context to interactions, and without this context a brand may lose the opportunity to provide a next-best action in the proper cadence.
What is Automated Machine Learning
Clear the Hurdles for a Personalized Customer Experience
Customer Journeys are Dynamic, Your Engagement Technology Should Be as Well