During a walk up Camelback Mountain in Phoenix one recent morning with my colleague Joe Kamenar, Steve Harris, senior director web analytics, web digital group at Capital One and David Millrod, founder of visualization platform Insight Rocket, we talked about data integration. We talked about analytics staffing. We talked about multichannel attribution models. We talked about “huddles.” Huddles? These practitioner-led discussion groups are the heart of XChange, the digital analytics conference we were attending. Now in its eighth year, XChange huddles reflect what’s on the mind of digital analytics program leaders who work at both global and national brand organizations.
Here's a taste of some of the topics that came up this year:
1. Analytics Platforms
No one really talks about analytics tools anymore. Is it because there’s no longer a choice?
During the first few years of XChange, there were endless discussions about web analytics vendors, but at that point there were at least five viable options. Now there are really only two. But is lack of competition the reason no one talks about analytics platforms? I don’t think so. Here’s why:
- Organizations are generally committed at this point to either Google Analytics or Adobe Analytics, so there isn’t a whole lot of controversy.
- On the other hand, there are lots of analytics platforms for social media, video, mobile, content and so forth, but in many global and national organizations, the digital team has grown so big that there are now silos for all of the digital channels and a different analytics tool is used by all of them, so it’s much harder for the digital analytics program to control all analytics. This isn’t a positive development.
2. Testing: Everyone Talks About It, But Few Do It Well
There was a lot of talk at XChange this year about testing: how to set up programs, develop tests, create a testing culture and adoption. Yet it seems like there are only a few companies that do this well and really have committed to integrating testing into their overall digital development process. One company spoke of having a member of the testing team work closely with project managers to discuss testing before a campaign is launched or new microsite is developed. This approach resonated with me because analytics has faced -- and still faces -- the same challenge for inclusion in the digital development lifecycle. And like analytics, the successful testing programs all had executive sponsorship from either the CMO or CFO.
3. Personalization: Everyone Talks About It and Has a Spin On How to Do It
There are few companies I talk to where personalization does not come up in the conversation. It goes like this:
Company: Can you get us to a place where we can do personalization?
Me: What’s the basic quality of your digital analytics today? What type of visitor segmentation do you do?
Company: We have lots of questions about the accuracy of our data and no one is really using the reports that are available. What did you say about segmentation?
Me: So I think that you need to get your analytics dependability sorted out, develop insights from the data that are going to help people in your organization get smart enough to know how to tune the personalization approach, and creating segmentation models based on visitor behavior and characteristics.
Company: Really? That sounds like a lot of work.
Me: Yes, personalization is at the top of the analytics maturity curve. Right now you’re just at the beginning.
Company: Hmmm ….
In our huddle discussion about personalization, a common thread that came up was that when people talk about the topic with marketers, e-commerce and content teams, the requirement is more or less described as “let’s be like Amazon or Netflix.” But from an implementation perspective, what does that mean?
Implementing personalization can take different paths, for example, using retargeting based on what was in a shopping cart, using multivariate testing to determine color and selection refinements, developing rules based on behavior (such as most popular products or recently clicked products), setting rules based on propensity analyses to determine what would be the likely action a visitor segment would take, developing personalization rules from direct knowledge about customer characteristics based on surveys, customer boards or field team feedback.
Thinking about going down the personalization road? We determined a path that aligned with gradually more discrete levels of segmentation, starting from the basic segmentation:
- Segmentation: behavioral, demographic, firmagraphic, geolocation
- Micro-segmentation: possible combination of initial segmentation clusters plus user identifier
- Targeted rules: developing rules for serving content to micro-segments
- Personalization: developing rules for content serving on increasingly more granular data points about customers.
4. Digital Analytics - It’s Not Just for Marketing
In my last few articles, we covered how your organization can affect real digital transformation by listening to your customers and focusing on what they want from your organization. If you do, then the digital channel becomes a vehicle for changing how you do business with your customers. It moves beyond marketing, to a comprehensive channel that encompasses product development, supply chain, marketing, customer service and internal operations.
At XChange this year, there was more evidence of companies starting to use analytics in a way that broke out of the strictly marketing or site optimization mode. One firm analyzes customer satisfaction in relation to page load time to determine optimal performance metrics. Another analyzes online buying patterns to plan new merchandise design direction.
We’ll start seeing more of this trend in the coming year as firms in sectors that have traditionally done most of their business offline move into multi-digital channels.