The customer decides, 1 on 1 marketing

Within marketing, personalization has become increasingly important in recent years. Where much traditional communication is often taken for granted, online media gives us the chance to properly identify target audiences.

‍

We know a lot about (potential) customers and can therefore tailor the message more and more to their needs, which has a proven positive effect on sales. Moreover, personalization is now also expected of us by customers. Although the possibilities for personalization are becoming more and more obvious and there is now talk of hyperpersonalization, we see that in practice it is still almost to not applied.

‍

Personalization gives companies a competitive edge. Companies are able to turn the market on its head by delivering a hyper-personalized offering, or creating a very good personal match between a provider and consumers, who, after all, have an increasingly shorter attention span. So the success of the Uber Eat and Airbnb's of this world is largely due to the level of personalization they can deliver.

‍

Collecting and storing data

Traditional companies realize that they have been overtaken left and right in terms of personalization in recent years and want to keep up with the times for that reason as well.... But if we look at what personalization projects are actually being set up, there are only a handful of clients where we are working on 1-to-1 personalization using machine learning and algorithms, for example.

‍

Admittedly, for many companies this may still be too big a step and a good first step is to centrally collect and store the data from all channels and points of improvement with the customer. This can be done in a data warehouse, a repository or through a DMP(data management platform), so that there is an overall picture of the customer. Having those data sets in order is the foundation. When a company has its data warehouse and cloud set up properly, they are already well on their way. DMPs are currently rather underestimated and really only become interesting for large companies that are growing fast and with a large gap in data networks.

‍

From segmentation to machine learning

The first step is to neatly map all the data. Then marketers and CRO experts can start breaking that data down into separate components. Think male-female classification, where someone lives, interests and whether someone is already signed up for the newsletter.

‍

Once that data is enriched and a complete user profile is created, much more complex components can be created. Eventually, those parts and scenarios become so specific that it becomes difficult for humans to identify and analyze them based on rules anymore. That's when the change starts happening, and that's when machine learning should come in handy.

‍

Personalization Maturity

As an "industry," we have now reached the stage where the application of machine learning is normal and we are creating many new systems based on what the customer wants. For example, we then predict what people are interested in based on previous purchases or searches. That prediction is used in personalizing the website and media through the use of advertisements. So the development in the maturity of data and personalization go hand in hand. When that development is ready, it may make sense to start working with a DMP. That makes it possible by accurately identifying target audiences and constituents and offering them the same, personalized message both cross-device and cross-channel. In this process of "personalization maturity," different teams - marketers, data scientists, CRO experts - have their separate focus, but work together toward the same goal: making it as interesting as possible for consumers.

‍

It should be emphasized that this involves the personalization of both media and platforms, such as the website and app. Both are important touch points for many businesses.

‍

Machine learning as a commodity

Personalization cannot be separated from developments in machine learning. We are already at a stage where many tools and software for machine learning by the likes of Google, Sales force and Adobe are included by default in their software packages. These are applicable to the same kind of business and can be activated at the push of a button. In that sense, machine learning has become a mass product, something that is widely used by companies, as opposed to a developing technology in which there is still a lot of investment. Those tools are just available in the CRM and data ware house, and can basically be used by anyone.

‍

Data scientist out or more important than ever?

The question then is what this means for the work of the data researcher, who in this process of sales forecasting etc is actually no longer needed. But at the same time, companies are growing in their personalization maturity and more "problems" arise that machine learning as a commodity cannot solve. So this means that in the long run, more data researchers are actually needed to apply machine learning to all these unique issues and challenges.

‍

In short, with growing demand, machine learning is being automated and made available in large software packages, but eventually there will be many more user cases that need a data scientist to interpret correctly. The movement of the Internet is unstoppable. It won't be long before malls provide a complete picture of every step visitors take. That will give huge amounts of data, which will make it possible to launch very specific personalization pathways. In this, the role of the data researcher will only increase.

‍

Babysteps

Admittedly, this is really not yet an option for many companies. A more realistic plan is to have data sets in order. The first question an organization can ask itself is: how do I become more interesting to my target audience? And thus specifically for all parts within that target group. From there, more and more steps can then be taken toward personalization. It is a process, one step at a time.

‍