- How do you think about the difference between personalization and one-to-one marketing?
I think that we’re getting to a point where every single touch-point with a customer should be personalized, but not every single interaction needs to be one-to-one marketing.
When you have a known or anonymized user identity, you can use a richer set of data to personalize to that very specific customer.
For example, if you want to do broad-based targeting and messaging, those experiences can be personalized, whether through content on the website or through advertising.
Then, there’s different channels where you really have more of a user identity that you can go after e-mail or social.
These channels allow for a more native place to do one-to-one marketing, where you are personalizing to that specific customer based on the PII (personally identifiable information) information and a full user portrait specific to that customer.
For example, if I’m out there browsing my cooking blog, and all of a sudden I get an ad from a specific airline saying, “Hey, Kiki: I don’t think you like your particular seat,” that’s a little invasive, and I may not be receptive.
On the other hand, if I get an email saying, “Hey, your upcoming flight to JFK has you sitting in a middle seat.” I will see that and appreciate the reminder.
That was a completely different type of engagement. It’s the same data about me with a different medium of talking to me, so it feels more appropriate and timely.
Editor’s note: See this blog post for further explanation on the differences between personalization and one-to-one marketing.
- Are there specific best-practices around measurement or testing for personalized campaigns?
What we find with a lot of our customers is there are certain attribution metrics, tests, and models which can be done within the DMP by the marketer to improve personalization.
But if our clients have data scientists and analysts who have been crunching traditional data sets for years, we think the best practice is to export the raw data from the DMP to analyst teams who will link the raw data to anonymized data from a CRM platform for measurement, modeling, or propensity scoring.
It’s really important to have a multi-directional data flow. Data is re-ingested back into the DMP for analysis by the data science teams, but is also sent back out to be used in campaigns by the marketing teams.
- When you think about the different data types: Is there a certain crawl, walk, or run for data types to use?
We always recommend our customers start with their own first-party data.
Within the customer data, pull all customer loyalty and subscriber information together.
Then anonymize and tie the first-party data to your onsite marketing and e-mail activity. This will reveal where the gaps and holes are; whether it’s specific types of customers or specific data attributes.
Then go out and fill those gaps with second- or third-party data.
You’d be surprised at how much data natively lives within your organization. There’s no better source of data that will tell you the most about your customer beyond the data that you already have.
- What is a key problem your customers are experiencing that you feel you can help with?
One problem we see over and over is data centralization.
We often see several different groups within an organization working with a lot of different platforms that all categorize and utilize data differently.
We can make sure data sets are correlated and include anonymized first-party and CRM data that can be onboarded with a partner like LiveRamp.
It can include second and third-party data as well as analytics and information around impression exposures, display inventory, and search inventory. Consolidating that information enables a marketer to make sure that everybody is working off one singular template.
- Where do you think personalization is going?
It will be a lot easier for the end user to be able to harness all these different data sets independently.
If you think of a marketer today, they have a DMP to consolidate all the different data sets but all these different data sources are varied and need to be heavily managed, so there’s different onboarding processes for each set.
It takes somebody with a very high aptitude to be able to understand all of that data and pull it in.
I think that in the future the workflow is going to be easier for an end user.
I also think that today, personalization is tied to relatively traditional engagement channels, but with email, social, and display our ability to reach customers will grow exponentially as there’s more and more mediums to interact and personalize your customer’s experience.
Learn everything you need to know about 1:1 marketing. Download our Definitive Guide.