Webinar Notes: Why Customer Data Platforms; Why Now

April 18, 2018

CDPThird party data is under siege. Yes, you read that right... third party data is on its way out, and marketers need to evolve their strategies accordingly. 

This recap of our April 18th webinar looks at the shift from third to first party data, as well as dives into the five stages of CDP analytics. 

Why CDPs; Why Now Webinar

Watch the Webinar

 The GDPR Is Imminent

The General Data Protection Regulation (GDPR) will take effect on May 25, 2018, and even if you don't operate out of the EU, this new law will impact any company who markets to, or collects data from, individuals located there. If you are unfamiliar with the GDPR, you can find the full text of this new regulation here.

Educated Consumers

Thanks to recent events like the Equifax data hack or the Facebook and Cambridge Analytica scandal, consumers are becoming more cognizant of who has their data and how it's being used. As a result, consumers are becoming more reluctant to share their information with companies, and when they do, they expect said company to keep it safe and only market to them with relevant, timely content without invading their privacy.

The Rise of First Party Data and the Need for a CDP

First Party Data

With the ever increasing regulations and wariness around third party data, marketers are turning to their first party data as a solution. Studies show if a consumer gives you information, and they trust that you will use it properly, they will continue sharing more and more with you over time. The more data you have for each customer, the better your data quality, your analytics, and your ability to predict customer behavior.

But, what good is having all this first party data if you have nowhere to store it and no way to act on it? That is where a Customer Data Platform comes in to play. A CDP can help you harness all of your first party data from every customer touch point into one place, and utilize the resulting insights to market to your customer when, where, and how they want to be reached. A few use cases you might want to consider when evaluating a CDPs capabilities include:

  • Outbound Marketing - Customer data platforms allow you to segment customers based on their value, behavior, or other attributes that drive your outbound marketing strategy. This has really been the bread and butter of CDPs for the last few years, but marketers are finding it increasingly more important to have more robust functions other than just segmenting.
  • Digital Advertising - Thanks to capabilities such as Facebook custom audiences and Google custom match, companies can now serve up extremely targeted online advertisements to their customers. Example: utilize your CDP to pull a list of the customers most likely to convert in the next 30 days and begin serving them online ads for their preferred product(s). Simple, yet extremely effective. Additionally, you can produce look-a-like audiences of your highest value customers to acquire new customers who are more likely to purchase.
  • Customer Experience - Sure, you can use your current analytic platform to pull a list of customers who have purchased in the last year. But, do you know what they did in between those purchases? What they browsed online? How many times they browsed? If they went in store to purchase something? Being able to track every customer touch point not only allows for better targeting, it also gives marketers a better understanding of the whole customer journey and each customer's potential value.
  • Analytics and AI/Machine Learning - his goes beyond your standard reporting, and is more about a CDPs ability to use your data to determine predictive models. Think less reactionary analytics and more proactive analytics. This AI and machine learning functionality in respect to CDP analytics is where we are really going to dive in.

AI and Machine Learning - Not Just Buzzwords

In order for AI and machine learning functionalities to work, you need data and you need it to be accurate. A CDP allows you to quickly input and de-dupe thousands of customer data points, ensuring the machine learning algorithms are working off of your most current and up-to-date customer information, and subsequently, your analytics are providing the proper insights for growth and success.

The Five Stages of Machine Learning Analytics for Marketers

1. KPIs

Regardless of industry, competitors, or marketing strategy, it will be key to ensure your KPIs are customer centered, and that you establish benchmarks and goals to measure success. At the end of each analytics cycle, these KPIs might change depending on your findings.

“Every transaction and interaction drives to a view of the world from your customer's perspective, and this is extremely important for retailers.” - Omer Artun

2. Measuring, Testing, and Optimizing

As marketers, we are more than familiar with the above terms. From A/B to multivariate testing, we know the importance of a good test and optimization. However, according to David Raab, founder of the Customer Data Platform Institute, marketers aren't yet trained in testing for long-term value.

Instead, we are focused on metrics with short-term value such as click rates, CPC etc. By having access to more in-depth and layered analytics, marketers will be able to shift their focus to more long-term tests cycles, such customer lifetime value or marketing attribution reports. Without the layered and in-depth data about our customers, testing and optimizing against these types of KPIs can be extremely difficult and unreliable.

3. Segments

Once the data is in place, you can begin segmenting customers based on specific data sets and values. While most marketers are already doing this to some extent, it goes without saying that the more data you have to work with, the more precise and effective your segmentation will get.

4. Machine Learning

The machine learning stage of the analytics cycle takes your segmentation and makes it much more sophisticated. Machine learning mimics the human ability to solve problems regardless of noise and impartial data, and it will help marketers better understand a customers true intent thanks to algorithms that can analyze huge amounts of customer data and patterns. Machine learning algorithms are able to produce usable outputs for marketers so they can target customers based on their likelihood to perform an action.

Machine learning in a customer data platform will also apply outlier treatment to your data and help you analyze what actions might be out of the norm, as well as utilize feature scaling that determines hundreds of customer attributes, but tells you which are most important to hone in on for ROI.

5. ROI planning

Once you have reached this stage, you will need to assemble costs and budgets to determine the ROI of your marketing campaigns. Once you evaluate the outcomes, it is time to start back at stage 1 and revisit your marketing KPIs and start the cycle over again.

So, what are the key takeaway from this webinar?

  • You should be using a Customer Data Platform to help you shift from a third party data reliance to harnessing the power of your first party data.
  • If you are evaluating a CDP, evaluate features that will meet your use cases, such as outbound marketing and customer experience.
  • Consider what analytics capabilities you need your CDP to perform for your marketing strategy to ultimately be successful.
Watch the Webinar