How to build a segment for marketing automation? Why is it important to segment your users? We explain it in 5 points.
Segmenting the market, the customers, the leads is an exercise we do on a daily basis. The goal is always the same: to identify a point of attention in people, an extra leverage to propose content as close as possible to their needs. Today, companies can collect a lot of information about users in a very complicated scenario. This information has the characteristics of big data: it is a lot, it changes quickly, and it is of different types. That is why it is good, before implementing a marketing automation project, to integrate the data as best as possible to build refined segments.
Building segments for marketing automation requires approaching data in a certain method. These are the steps to get started on the right foot:
Mapping
The first question to ask is: what data, in what platforms? A marketing automation project cannot ignore a preliminary analysis of a company’s information asset management. It is therefore essential to check on which platforms all the data potentially available reside: crm, ecommerce, loyalty cards, erp, cash systems, etc. Next you need to understand which of these are functional to business objectives and focus on what is really useful, distinguishing it from what is “noise.” You are probably in possession of a lot of information, but not necessarily all of it contributes to segments. Some may be irrelevant from the point of view of journey analysis.
Normalization
Marketing automation platforms typically handle different data sources and also take care of their normalization: they check the goodness of the data and capture it only if it is correct. In some cases they intervene by correcting fields following predetermined rules (for example, they introduce the country code on the phone number field). In others they return lists of errors that can be verified. Finally, the technology automates data matching by tracing information back to a single user and discarding/marking duplicates.
Synchronization
Data synchronization is something that typically needs to be customizable to the business need, and the timing is often linked to the business model. Real-time synchronization can be costly on high volumes, but in many cases it is necessary to return users to an experience that is always up-to-date on their preferences and engaging on a communication level. It certainly cannot be a manual activity, but if it cannot happen in real time, it must be scheduled according to strategy.
The creation of the segment from the data
As we mentioned in this post, segments can be of two types. We consider static data those that are based on information with a low level of update, which certainly can change, but with a low level of speed and not with a high incidence (for example, gender, age, but also residence). Dynamic, on the other hand, is information about the context of interaction with the brand for each individual user: people browse online, click, read, enter stores, purchase, and judge products publicly. Managing this complexity requires technologies that can not only handle different data formats, but also know how to analyze and aggregate them in real time. Information to create segments typically falls into three types:
- socio-demographic – gender, age, education level, etc;
- purchase history and RFM-where an attempt is made to calculate the “value” of the customer, based on how much and what they have purchased, how often and how far in time since the last time;
- contextual – userengagement at that precise moment: pages viewed, for how long, clicks, opens, shares, etc.
How then is a segment created? By using all three of these dimensions. For example:
- Women surfing from mobile, 9 a.m. in Milan.
- Inactive users for more than 10 days (on all channels) with a potential spend x
- Regular customers, age 35, who have already purchased for x value and are viewing a certain product
Recognizing users and thus belonging to a certain segment allows marketing automation to offer content close to the needs of each of them.
Algorithm-based segments
As we have said repeatedly,artificial intelligence is the engine that can unleash the personalization of experience. Algorithms are constantly learning from segments, and the wealth of information about them can help us create new ones. Marketing automation can highlight certain similarities between users, which marketing has not thought about: for example, discovering that all women browsing from mobile at 9 a.m. also buy a certain type of shoes. Of course, this is just an example, but it is useful to understand that there are similar user segments and profiling is not a static thing. Segmenting is a continuous learning logic: integrating data with marketing automation is therefore the key to continuously improving the brand experience.
