Remember the last time you visited a website or tried out a new app and decided that it was not for you? Maybe for a different audience with different needs, but definitely not worth your time? How long did that decision take you? 30 seconds? 10 seconds? 1 second?
A study from Princeton University suggests, that people tend to make up their mind about a previously unfamiliar person, in as little as 100ms following a face-to-face encounter.
“When exposure time increased from 500 to 1,000 ms, trait judgments and response times did not change significantly, but confidence increased for some of the judgments; this result suggests that additional time may simply boost confidence in judgments.” (1)
Without a doubt, humans are far better equipped to read and interpret other people’s facial impressions, than judging the value-added of an app simply by glancing at it’s content, layout and UI for 100ms.
One is an instinctual trait which has been developed over a period of two million years and became crucial for survival. The other, not so much. The first website was launched in 1992. It’s easy to forget how young this phenomenon is and how fast-paced its development has been ever since.
Whatever the minimum exposure time for deriving judgements in the digital landscape, it does not allow for a lot of time. As in real life, first experiences are crucial. So how do you still create a meaningful, first experience?
At Record Bird, we have the goal of instantly informing a new user about — at the minimum — 1 upcoming record by an artist the user strongly cares about, but would have missed otherwise.
It’s this feeling of “S**t, I’d have totally missed that…”, which makes a user remember Record Bird and come back again.
This exemplifies an important insight: After signup, a user could follow dozens of artists, but not a single one of them might release a new record at the given moment. Unlikely, but possible. Until recently, the user would then sign up to a new service and read something like this:
This is the equivalent of a barkeeper asking you: “Are you sure you’re in the right place, man?”. No matter how unlikely, we’re here to avoid such negative experiences. Here’s how:
1. The Power of Sourcing Taste
Most digital experiences work best when personalised. This could mean using one’s location, connecting one’s friends or — in the case of Record Bird — knowing from the start which artists a user likes already. Fortunately, this information exists: Facebook knows which artists we like, Spotify & Last.FM know which ones we actually listen to and iTunes & Discogs know what we actually were willing to pay for.
The power lies in giving the user the tools for sourcing and importing such information with a single click, therefore preventing the need to painfully add artist by artist manually. While we currently focus on integrations for Facebook (built) and Spotify (in development) we know how crucial all of these are for a user’s first experience. Still, focus is key and user research suggests that Spotify and Last.FM will be key assets in our on-boarding experience.
Note: Apple Music is still taking its time to release their API to the public.
2. Personalised Recommendations
Yet, the steps described above still don’t provide any value, if none of those sourced artists release a record at the given moment. Obviously, the more artists the smaller the chance, but still there’s the possibility of a terrible experience. Personalised recommendations can change that.
In general, a user should always be able to grow their artist spectrum as easily as possible. We believe this leaves us with two prolific moments for offering recommendations:
- During the last step of the On-Boarding experience, where we ideally have already gained some information on the user’s taste of music.
- Directly on top of the user’s My Favourites section, where she can effortlessly follow artists without leaving the comfort of their feed.
The crucial aspect of recommendations though is not primarily their positioning but their relevance. Our logic for recommended artists is based on the technique of collaborative filtering. The upside of this approach is that it evokes highly relevant recommendations, as long as — here comes the downside — sufficient data for filtering exists. In short: If only few users on Record Bird track “Swedish Hardcore Death Metal Bands”, we might not yet be able to recommend highly relevant artists in the domain — but we’re currently working on a hack around this issue! :)
The beauty of it is though, that while our service grows, recommendations will automatically become more and more relevant, even to the most unique users out there!
3. Showing What The Community Loves
Here’s an assumption: People who are part of a community are likely to share a more similar taste compared to people who are not. To test this assumption and — if true — directly draw benefits from it, we now feature a trending records section on top of our All Records feed. In a nutshell, this is what most people on Record Bird think you shouldn’t miss. So if you can’t be bothered to go through the entire feed, here’s where to look. The algorithm is currently based on aggregated user follows, but we’ll integrate a social-media based popularity index in the future, so fast-rising artists rank higher.
4. Cutting Through The Clutter
Information is only valuable if digestible and relevant. In today’s digital landscape, a news feed is one of the dominant mediums for providing users with a quick overview of recent updates and developments. In the case of Record Bird, our All Records feed provides users with a quick overview on what’s being released on any given day. This allows users to learn of new releases by artists which they like but haven’t followed yet. The main focus is overview & discovery. Yet, with more than 250 releases / day this turned out to be quite inefficient. We now have all releases broken down on a day-to-day basis, and ranked according to user engagement. The user is now faced only with the Top-8 releases per day, with the option to view all others if desired.
With the music industry’s recent decision of setting Friday as Global Record Release Day (http://bit.ly/1TUX8ms), we’ll experiment with structuring the feed on a weekly basis as we currently monitor a high volatility in the relevance of the Top-8 records on a given day — e.g. Sunday vs Friday.
5. Sugarcoat A Sour Pill
Great copywriting won’t make a terrible experience great, just as a nice barkeeper won’t make up for an empty glass. Still, both have the power to make a customer smile and prevent her from storming our angry and upset. I am not saying we mastered this skill yet, but here are some examples:
Ouch, your feed isn’t looking to healthy.
You currently have no records in your feed.
Looks like all your artists are sipping margaritas in the Caribbean right now.
We’ll let you know when they’re back in the studio…
Your currently following 0 Artists
Even Kanye West follows Kanye West on Record Bird.
Follow some artists to get the best experience…
Unfortunately, there are no guarantees what works and what doesn’t so testing is key — but comparing an empty feed with Justin Bieber’s friend list can go a long way with the right target audience! The important thing is to always combine informative copy with instructional messages. What should a user do next? How can she improve the experience most efficiently? Why does this not work?
You should know, so tell them!
In case you have any further thoughts or feedback, please get in touch via Twitter (@mahringer_a).
If you’ve enjoyed the writing above, please click “Recommend” below. You can also share it on Twitter and Facebook. This will help get the opinion of others.