5 things about email and customer data platforms

Email and customer data have gone hand-in-hand for quite some time. The benefits are clear.

Some organizations haven't gotten there yet. Others see overlap in features sets between different marketing technologies and wonder what to do.

One of these technologies with overlap is the good old customer data platform, so here are five things about CDPs that might be of interest if email is your thing!

Thing 1: Email is not a core capability of customer data platforms

There are a wide range of email capabilities within customer data platforms, as there are a wide range of data collection capabilities, but being an ESP is not a core function of a customer data platform like collecting data is. There are quite a handful of systems out there that originated as ESPs or marketing automation systems, however, and now find themselves sitting in the customer data platform market while offering a full email solution to boot.

[Photo credit: www.reviewconnection.co.uk/]

It is easy to point to these ESP-CDPs and say "that's a system that does email." The idea being that such a system can not possibly be as good at customer profiles, anonymous data collection, etc. And this kind of dismissal happens, especially by purpose-built CDPs.

Being an ESP can't exclude a vendor from the category or conversation. At the same time, I agree that these are marketing automation systems with a database unless there is coverage to use multiple other ESPs to:

  1. personalize and send or suppress batch emails based on the customer profile,
  2. personalize transactional emails based on the customer profile, and
  3. send personalized triggered emails based on the customer profile

Small-enough businesses might be able to get away with an attached database to the ESP, but at a certain point this data needs to be centralized to be relevant and efficient.

Thing 2: Segmented and 1:1 emailing should not be a major production

A CDP/ESP integration may take hours, days, or weeks. Once the integration is complete, it ought to be stupid simple to create emails that use data collected from websites and mobile apps, as well as other systems (POS, CRM, etc.).

The myth that ESP-CDPs push is that this pair of systems needs to be too tightly integrated for them to be separate. It's just not true -- plenty of vendors that offer standalone ESPs are doing a great job with APIs to support personalization across batch, transactional, and triggered email sends. The best CDPs are able to connect to more than a few of these.

Thing 3: A customer data roadmap will help

The best email marketers drive growth by not stopping. Their continuous approach needs to be supported by data collection and segmentation focused on email recipients. With this level of focus and partnership, email marketers should be able to both get a seat at the data collection table and produce personalized emails with ease.

A customer data roadmap will help beyond email. With email especially, it is easy to keep your head down, always focused on the next one or seven. Taking what's given. In those situations, it might not even be clear that metrics are suffering.

Email is an area of tremendous opportunity because of the level of intent already shown by these customers or prospects. A lack of data can mean a lack of ideas, a lack of campaigns. Once the data is flowing, like past purchase information or real-time behaviors, ideas and campaigns also flow. There needs to be useful data in the ESP, and the data need to be in ready-to-use formats for your use cases. This is where planning really helps. What is needed now? What will be needed to support future initiatives? Keep email in the data collection conversation, and then data collection and analysis/segmentation can be organized around 1:1 email desires, for example.

Thing 4: Initial identification, identification for anonymized personalization, and identification for personalization

flickr/natashamayers

Identification is perhaps the toughest thing about CDP to plan for, and email is right there in the thick of it. Identification is what makes it easy for CRM+ and ESP-CDPs to dismiss some purpose-built CDPs as insufficient. As with all other things CDP, blanket statements must be picked apart. Actual use cases must be considered.

Thanks to the existence of an identity (email address!), any activity from the ESP can be combined in a CDP to round out the email side of customer profiles with ease. On the other side is where things are a bit more complicated, especially with personalization after a click. If someone clicks through on an email that was sent to recipient ABCDEF, are they really ABCDEF? Or is ABCDEF a distribution list? Or did ABCDEF forward the email to someone? Or did ABCDEF copy the link and share it on Facebook? Pay close attention to how customer data platforms handle identity given specific oddball but not infrequent scenarios.

There are a lot more feature weeds in the areas of identity resolution and multi-leveled identities to consider, but at a high level I want to look at these three subthings. CDP tend to be flexible enough, or built directly to handle these three separate valuable areas of identification:

A. Driving initial identification - converting anonymous to known. The anonymous might be visitors of a website, app, or physical location. Customer data platforms help with this in various ways, depending on the platform. Two good examples include asking for the identity with a timely personalized approach and adding supplemental data to the record when the new identity stored.

B. Identification for anonymized personalization - collecting identifiers from "anonymous" visitors that ties to existing data about a customer. These are unauthenticated, or sometimes probabilistic matches to identities. Not good enough for full-on personalization, but definitely good enough for targeting based on the assumed profile. This is the easiest of the three because it requires no incentives for the customer.

C. Driving identification for personalization - driving anonymous or presumed visitors to become authenticated. When data from actual customer records might be exposed to visitors during the experience, or when data is being permanently attached to a customer record, there are steps that have to be taken differently than when this is not happening.

To further explain the difference between subthings B and C here, customer data platforms help with some amazing marketing shortcuts. The same amazing shortcuts might conflict with other use cases, so striking a balance is something the CDP will have to help with. For example, loose identification of a website visitor based on an email click is an amazing marketing shortcut. Someone clicks through on an email for the first time on a mobile device, and bam! you are placing individualized messaging, content or product recommendations, etc. It's probably the right person, but you are not yet ready to trust that this really is the person you are targeting.

Thing 5: Not all email data collection is the same

Without good connectivity to the ESP, identification programs will be lacking. A CDP-friendly ESP will:

  1. Accept new members via API or non-iframed code embedded on a website
  2. Pass link clickers through a redirect, collecting and sharing important details on the way through. One super helpful detail is a probability score to inform the CDP on how confident to be that the clicker was the recipient.
  3. Make it easy to tag links. Especially if redirects are not an option, it best be easy to tag links with information from the ESP view of the profile. Also, it should be easy to tag links to your sites only.

Volatility continues, clarity awaits

The ESP market is volatile on its own, and to make matters more interesting many organizations work with more than one. Customer data platforms support the volatility as well as provide flexibility to use the same customer profile across multiple ESPs without much hassle.

I hope this has been helpful to folks considering the connection between email and customer data platforms! It would be great fun to take a question or two in the comments, or by email/LinkedIn.

Leave a comment