Welcome back! I’m glad you’re back for Part Deux of my data layers discourse (if you missed Part The First, just click here). Let’s jump right in where we left off.
Data layers in the real world
Today, I know there are plenty of digital analytics professionals out there who have aided others in ensuring the data of interest can be found quickly and easily…and that both comforts and excites me. The progression has not happened quickly enough, though, with data layers relegated mostly to being an “analytics thing,” and not quite as broadly adopted as I’d have hoped by now. As a hybrid digital analytics-slash-digital marketing guy, that makes me sad…I need all the pumping up I can get, so I poked around the Web for a few minutes to see what kind of data layers I could turn up in the real world. When you’re done reading, you’ll find the results at the bottom of this post.
Most folks I come across are still working towards a collaborative and cross-functional approach to data. I credit Google Analytics (er, Universal Analytics), who through Google Tag Manager has made great strides in popularizing the ever-cryptic concept of a digital marketing data layer. When the W3C’s Customer Experience Digital Data Layer standard definition was in progress, I was very happy to see folks coming together to define a standard. This meant that businesses of all shapes and sizes would know, as long as they published their data in a specific way, that a multitude of digital marketing technologies would be able to handle it. Right? Well…maybe. Just as data layers have been adopted slowly, the W3C specification hasn’t exactly taken off as a widely adopted standard, though I do think that it, too, has done a lot to popularize the idea of a data layer.
BlueConic customers can rest assured that whether that spec is in place, or a proprietary data layer has been instituted, we can collect the data as you intended it to be collected.
So, where to begin?
Item one: have a really good understanding of the pain points. You’re probably going to need it.
Tech savvy organizations, keen on marketing automation and eliminating technical debt, are likely to be on board with the idea of a data layer. Still, asking for a data layer can sound like a net new effort request. It’s not. You’re fixing and optimizing your digital marketing data practices – a noble cause that will save time and money. If you can apply numbers to the potential time savings and missed opportunities over the last 12+ months, your case will be a strong one.
Whether you’re in digital marketing, analytics, or supporting cross-functional endeavors from the same: if a data layer hasn’t been put in place, I would start at the bottom. The very bottom. Ask a few questions of yourself!
- Has your organization suffered from tagging efforts eating up development time?
- Is there a lot of back and forth required in order to define data collection requirements and then implement appropriate tagging for those requirements?
- Are there asterisks/flat-lines in your analytics in the past 12+ months as a result of missed data collection opportunities or disrupted collection (e.g. when a page/site changes)?
If you’re answering yes to any of these, get started by thinking small!
If indeed you are starting from scratch, a walk-first mentality will probably save everyone time and reduce any potential shock from an effort perspective. Arm yourself with some analysis on how a data layer will help, and ask to jointly define some very basic requirements. For example, if you can align on how to make available the following data points, you will be in great shape to get started.
1. A user ID
2. Another key attribute of the user (e.g. tenure or a lifetime value indicator)
3. Product or content categorization
By defining a standard place for that type of data to live moving forward, you’re creating that common language and helping to ensure the success of efforts that rely on that data – and laying the ground work for your data layer.
A quick look at a shining example
Before I go, I had promised to take a look at some data layers. Nothing like the real thing to really show the value for all things related to digital marketing. Here is one for now, and I’ll take a closer look next week when we show you how to read/write to data layers in BlueConic!
We’re fond of Drizly both as a customer and as an alcohol delivery service (truly: we need a second refrigerator in the office to accommodate their goods). I chose them as a shining example of data layer adoption because transaction data is surfaced directly to a data layer embedded within the order confirmation page. Various marketing technologies are able to make use of the information that includes product and transaction details:
Full data layers can be a bit ugly to the untrained eye, but if you’re a digital marketing aficionado, you might be excited to see things like the city/state the order was delivered to, the promo code used in conjunction with the order, and the products ordered. All of this rich data is made available for the various marketing systems that rely on it for relevant interactions.
Consider what capabilities this enables from an advertising and behavioral data perspective, especially with a system like BlueConic in place to distribute the data along with other collected behavioral and demographic data points! Advertising platforms can stop presenting ads related to this promo code and advance to another stage of the customer lifecycle.
Email service providers can target the user with content related to product categories there has been interest in (BEER!), and not just from this single order, but from every view and click leading up to the order.
Early next week, I’ll be back in part three of this discourse to show some examples from BlueConic listeners and tag management connections that allow us to read/write data layer information!