I just spent some time with Nicole Dingess, vice president/product, design, and UX at Gannett to get an update on the work she and her team are doing around the UX for personalisation.
They saw stronger audience and engagement numbers over the past three months due to personalised experiences such as the “For Subscriber” module, which saw daily engagement from non-subscribers jump from 2.5k per day to as high as 12.5k, highlighting the app’s ability to increase engagement through personalisation of both what content users see and how they interact with it.
So it’s no wonder they wanted to double down and figure out how they can get more people choosing their preferences.
Before we start, hats off to Nicole and anyone at Gannett working on centralised features because this encompasses 200+ titles and, as we know, you can never please everyone. However, I think she and the team will come pretty close because of it. Not only does she have internal stakeholders throwing questions and comments, but she had a thoughtful answer to everything I threw at her (and I love asking questions).
We focused the conversation on two parts:
- Onboarding to get people to select what they are interested in at the moment of subscription.
- User prompts afterwards further along the journey.
Although we know active personalisation increases engagement, getting readers to give their preferences in the first place is tough.
In an ideal world, you will already know a little about your customer: What was the article that tipped them over to subscribe? What topics have they been interested in to date? But sometimes you won’t or it’s simply not a priority to build this into v1 of your onboarding. For the sake of this post, I’m going to shelve that thought for a later discussion.
There are three forms of follows within active personalisation Gannett is looking at:
Having two to three categories for users to pick from is fairly standard (look at examples from Medium, Substack, Spotify, and more here).
Of course, not every user is going to go through this flow when they subscribe. Some just want to get to the content. And they want it now so it’s important to give them the option to skip for now. This doesn’t mean you can’t come back to it, but you have to be mindful of what the user wants at any given time. And if they have just paid for access to something specific, their mindset may not be in the right place to start handing over information about their preferences.
Onboarding in this way is likely to be tailored to where they are signing up from: Are they in app or on the desktop? Gannet started their personalisation in-app only. Either way the experience will be slightly different. Make sure you think about both.
Gannett is building this out right now, and Nicole told me the first thing they will be looking for in their tests is where the friction is: What do users skip? What’s tripping them up? It’s hard to know the right number of steps for your users until you actually try it. So how do we get around this?
If a reader doesn’t go through the full onboarding at that time, the prompts can be broken up into different visits. There are a number of ways of slicing and dicing this. For example, you could prompt people to follow an author when they have read a full article or two or more articles by the same person. Ditto with topics. Or you can add subtle cues such as: “This article was highlighted in our morning newsletter. Would you like to sign up?”
Another way of looking at this is to highlight things users may not otherwise know they have access to (or don’t know they exist). Gannett is mostly made up of local titles so that could be high school sports scores or local events through to a new podcast. These are additional benefits which Gannett highlights as “Did you know?” questions to readers.
The beauty of in-app prompts is that you can ask a little each time to build up a profile. When done well, users want to add information as they see an additional value and it doesn’t interrupt their experience. Rather than 10 questions in one session, they may get these questions over 15 visits in one month. It feels more manageable.
For example, the FT adds related topics when you do a search. It’s a simple yet effective way of nudging people to add interests and preferences.
It’s not just in-app or on Web prompts that can be used. Gannett is doing some experimentation with other formats — particularly using “Did you know” type questions to highlight products and services within newsletters people are already receiving.
The real thing to understand in all this is that we know active personalisation can increase engagement dramatically. Therefore, if you decide to go down this path, you’ll need to allocate time and investment into onboarding readers through a variety of means.
If you’d like to subscribe to my bi-weekly newsletter, INMA members can do so here.