by John Judge, Business Consultant at NavigationArts, now an EPAM company
Amazon.com is responsible not only for making digital personalization a mainstream concept, but for doing it so well that it is no longer a privilege, but an expectation. As such a widely used eCommerce giant, their “People also purchased…” capability is coveted by everyone from start-ups to market leaders.
Amazon’s approach to personalization started simple – they made it easy to understand, and to the user it didn’t seem to overly intrusive or “creepy”. That traditional presentation of recommendations still exists, however it’s quite clear that the model has evolved and the logic has been refined over the years. Amazon has an incredible amount of data in years of purchase history, and they’ve made no secret of putting that data to work. They have been the golden example of how to do personalization for as long as I can remember.
Recently while discussing the personalization end-goals for retail client, I began to use a familiar analogy but stopped myself mid-Amazon in saying “Once we get through all of this they will be personalizing their products like Amaz…”, replacing it with Netflix.
Now that was a first! This wasn’t necessarily because I now believe that Netflix personalizes better than Amazon, I’m sure an argument could be made either way, but instead I was referring to how Netflix successfully uses data to personalize user experience.
This is a trend here that extends well beyond eCommerce. User Experience (UX) used to be about interpreting web properties through the theoretical lens of intended audiences, but now with so much data available to us about who that audience is, we see (as in the case of Netflix) how valuable this can be not just for a company, but ultimately for the end user.
A lot of users balk at the thought of their data being made available. The thought of someone you don’t know having information about you is a little unsettling, I’ll admit. But did you ever think about how that data can actually make your experience as a user on websites and applications (and dare I say, life) a whole lot easier? I can’t believe there are many people out there who have not found a new song or artist they love because iTunes’ Genius recommended it. Ultimately, understanding how unique each one of your customers is makes their digital experience with you more memorable, and in return they become more engaged in your content, making them more likely to be happy with your product or service and company as a whole (aka success).
Personalization Strategy In Action
I’ve recently been working with a travel company to outline how personalization could make their site a leader in the industry. Like many organizations, they have a single website that serves both perspective and existing customers, and on top of that they have a great deal of business intelligence based on trends they’ve seen over their 40 years in business, which is applicable to both audiences.
Below I’ve laid out the basic types of digital personalization, and I’ve applied the travel company theme to provide context around each. These are not specific to the travel industry, and any company can take advantage of these if you have either or both user types.
1. We started with what information they had (e.g. analytics, user profiles, ecommerce data, consumer trends)
2. We then worked with them to understand what could be learned as a user moved through the site (e.g. preference for wine-tours)
3. Then we explored how these could be converted into logic that informs an experience
4. Finally, we defined additional information that would be useful and explored the technical feasibility of obtaining that data (e.g. geo-tracking of a user’s general location)
Implied personalization does not require a user to have an account on your site or be logged in/authenticated in any way. Users instead self-identify by clicking on their interests, and based on those interests you can suggest or show things to support those details in conjunction with the main content they’re exploring. The more a user browses, the better you can personalize.
Example: if our analysis of their customer trends showed that travelers who visit Yosemite also frequently book an excursion to the Grand Canyon, we will unobtrusively display content around the Grand Canyon or other related trips when an individual is looking at Yosemite.
How to inform: Look through purchase histories and call center logs for trends. Customize analytics to test these trends and continually refine.
This is all the information you get from a user once they login, or through a lookup of their IP. Things like their location, purchase history, email opens/clicks, visit frequency and any other stored or collected information.
Example: an individual accessing the site who has shown price-conscious history (e.g. only purchases cheap or discounted vacations) with the travel company may be shown sales and discounts along their path through the site. A database, CRM and/or email marketing platform would be updated accordingly so that the newsletter and other outreach (online and offline) can also be personalized around sales and discounts as well.
How to inform: Integration of digital assets with customer database/CRM/email marketing platform, implementation of geo-IP lookup, interviews with existing customers, surveys and feedback post-trip
Implicit and Explicit
Browsing history to date will have been stored with each unknown user, and once they authenticate (create an account or make a purchase), that history comes along with them into the customer database. This allows you to continually and collectively personalize the website experience from the very first visit. New registrants come with implied interests and can be personalized to right off the bat.
Example: An anonymous user spends a lot of time looking for vineyard tours in the South of France, and later ends up creating an account when purchasing a ticket to Sugarloaf. This known user will now receive information on Napa Valley vineyard outings as an ‘upsell’ during their trip to Sugarloaf.
How to inform: cookies, CRM and analytics integration
Stay on top of your game
Finally personalization is highly specific by organization. Even when you hammer down your organization’s specific tactics and trends, they will likely perpetually change. Testing and iteration of your personalization logic and it’s performance is extremely important for this very reason.