Personalization in Episerver Part 2: One hub to rule them all
The question is no longer why we should personalize – but rather to what extent. Data aggregation across all channels and predictive algorithms will be significant factors.
Although still in Beta release as of writing this, the new Forms (obsoleting XForms) looks very promising. Besides the obvious benefits of more accessible markup output and a blocks-based approach to form building, the new Forms also helps site owners gather personalization data.
Each form can contain a number of hidden profiling fields, such as
- geographical data (e.g. user’s IP, location)
- profile data (name, email, company etc, if user is logged in)
- user agent/mobile device
- Visitor Groups criteria met by the user
These are automatically gathered upon form submission, and while they currently don’t contribute directly to a personalized experience in real-time, the stats can certainly be used to learn more about your consumer base and tweak your personalization strategy to better suit their needs.
Obviously, when silently gathering data like this - potentially without the consumer noticing (or consenting) - site owners must treat such data responsibly and respectfully – to avoid crossing over from helpful to creepy.
Another really promising feature demoed briefly at the Ascend 2015 keynote was the new Profile Store. Although it’s still a work in progress, the demo by Jeff Wallace (from 35:50) suggested some interesting capabilities.
While Episerver has several tools (Visitor Groups, CMO, Analytics integration, etc) that separately track personalization data and monitor inbound traffic from various channels and email/ad campaigns, there hasn’t really yet been a consolidating feature to tie it all together.
That’s about to change with the Profile Store:
- In a single dashboard, site owners will be able to see all the aggregated profile data available about their visitors, whether they are identified as registered users/customers, or still anonymous.
- Each individual visitor’s interaction history is tracked, from which campaign/channel led them to your site, right down to concrete page visits, downloads or shopping cart checkout status.
- Profile filters can be saved as segments, which in turn can be used as a Visitor Groups criteria when personalizing content.
I see the Profile Store as a natural progression of the Digital Experience Hub (DXH), the bridge between personalized content and the various marketing/CRM connectors already available. It will be interesting to see how far Episerver will take it, and whether it will eventually enable seamless data flow between all these connected systems using Episerver as a hub.
Machine learning and predictive analytics
Whether based on real-time data (Visitor Groups) or aggregated profile history (marketing automation connectors), today’s personalization techniques in Episerver are mostly rules-based. In the next wave, built-in smart analytics will be able to predict consumer intent more accurately, and automate personalization based on their past and present interactions.
This concept centers around machine learning, in which the system collects data gathered from every touchpoint the consumer has with a brand. This not only applies to web sites, but also social channels, email marketing, online ads – every place we can identify and track an individual consumer.
When data related to a unique consumer is consolidated in one place, we are able to build more elaborate profiles and more accurately determine how to target them – which channel to approach them in, in what context, at what time, and to what messaging they respond best.
The ultimate goal is that the system is self-regulatory – think artificial intelligence – so site owners don’t have to manually manage a very complex set of rules and schemes.
We’re not quite there yet, but consumers are becoming more used to – even expecting – that every interaction or information they share with brands contributes to a more fluent user experience.