A lot of programmes, wrongly in my opinion, will look to AI as the answer for how to personalise. AI is not the answer for how to personalise, it’s the answer for how to scale personalisation.
If you were to feed an AI no customer information and no relevant behaviour, you’ll do an incredibly poor job of personalisation.
On the other hand, if you pick a couple of well-defined customers, and solve real problems for them with well-crafted experiences, you’ll do a great job of personalising (even if it is just a couple of segments). So, the lever to personalisation is closer to profile data and understanding problems, rather than “letting the AI handle it”.
There are two aspects of personalising based on user data, which we explore below.
Webtrends Optimize-collected data
The first aspect (and a key part of the equation) is the data you observe and can make sense of, based solely on what’s captured by Webtrends Optimize.
This is typically browsing behaviour, collected passively, or inferred meaning from where people go.
If someone clicks through a targeted paid-campaign based on OCEAN personality attributes (a test conducted to understand individual differences in personality and behaviour) and you know they rank high on the N for Neuroticism. They then land on a PDP and we know the brand and model, category and subcategories, price, colour, etc. They then browse around, clicking through categories and products, so we understand the journey and direction they’re heading in.
We can interpret this data to then predict what their needs are. For example, if they are searching for turquoise Nike products, then we know the most accurate information and products with that make up to be showing them.
If you choose to remember all of this data gathered, which Webtrends Optimize can do for you, it’s not hard to then ideate on how to improve an experience for those users.
Customer profile data
The other is looking at data that we don’t have but could and should import.
Datalayers are part of this equation – if you have Google Tag Manager, Tealium, Bloomreach etc., it’s very normal to find and ingest data from these sources for targeting and segmentation.
The other side, and by far the most powerful piece of all of this, is to look towards your CDPs and Single Customer Views.
Tools like Snowflake and Bigquery are very normal for housing all of your customer profile data. If you can calculate which segments you’d like to target in these systems, you can ship these lists to Webtrends Optimize to segment and personalise with.
Importing profiles into Webtrends Optimize
The process to import profiles from any Data Warehouse into Webtrends Optimize is to:
- Set up scheduled queries. Whether directly as with Snowflake, or in a roundabout way as with Bigquery, you will be able to ship your profile data into our containers in Azure Blob Storage
- Webtrends Optimize will link up User IDs. Your data will have User Identifiers, we need to find these on the page. These could be in your GA cookie or somewhere else on the page. We will need to hook these up to Webtrends Optimize.
- Imports process on a regular basis, and data is made available.
- Target directly in the Webtrends Optimize segment builder, or access the values with Javascript to dynamically build experiences.
All in all, we expect this to take no more than 1-2 days to set up (although it could take 1 hour if everyone is coordinated).
The net result – personalisation
There are many ways to unlock personalisation.
Whether it’s capturing a few extra bits of Custom Data in your test to segment by and interpret results on a cohort basis, or continuing paid media campaigns by targeting users based on their UTM parameters, or importing customer profile data to target users based on their history.
There's a ton of very simple things you can do to start personalising and work towards more relevant digital experiences. Get in touch if you want to learn more about how you can leverage your data and elevate your personalisation strategy.
