When a person is required to provide an answer, there are two ways they can retrieve their response. Memory recall is a function in which we sort through all of the information we have stored, looking for the best answer without much help (e.g. What is your password?). Memory recognition is a more "fuel-efficient" function in which we simply identify the most probable answer from a set of solutions given by a system or source (e.g. A classic multiple-choice question).
When directing users to recall information that could end in a mistake with no simple remedy, users began creating habitual behavior around inefficient workflows. By adjusting the workflow to increase initial recognition, we saw an adjustment in behavior, faster completion times, and fewer mistakes.
Being a social media planning platform with unlimited storage, handling high-quality media was one of our most important functions, and it's process looked something this:
The problem was that they were faced with a decision point every time they wanted to upload media.
This was all made worse if they decide to view the library, but the media isn't there. There was no easy way to upload media from the library or return to the decision point without needing to close everything and start again. That experience of receiving more effort from a wrong decision causes habits, and for us, it created a bad habit.
Since there was a chance of failing and having to do the process twice if you decided to look in the library and weren't able to find your media, there was simply less risk of extra effort by just using the upload media flow every time you needed something. Every. Single. Time...
And with us trying to be helpful by storing every file you upload in your library automatically (without checking for duplicates), it caused ridiculous numbers of duplicated media over time, adding unnecessary server costs.
Improve the experience of using media, and reduce the chance of duplicated media.
We wanted an experience to go from something like this:
To something a bit more enjoyable like this:
Users will reduce the amount of duplicated media if they're first allowed to scan their library without risk of extra effort.
Initial sketches with the wider team came down to 4 goals of improvement.
1. Remove the decision point all together.
2. Include a way to upload media from within the library.
3. Clean up and upgrade the old early-days design to our new design system.
4. Add a search bar.
Quick wireframes gave us a general idea of layouts, which was reduced down to this.
And the combination of wireframes + a little detail gave us a good enough prototype to test out with users.
As always, a couple of last-minute tweaks to add 5 columns of media was completed to show more media at a glance.
The final step was to simply translate it all into code! But as we did this and came upon a stage to begin testing our a rough version, we discovered one final problem to solve.
When a user decides to upload new media from this view, should it:
a) Be placed into this library where they can then select to use it (the way it currently functions)? Or..
b) Be added directly into the post they're working on and close out of this library screen (leading towards the assumption that they only do a single upload of one or more media at a time vs. multiple uploads of single media)?
Quick survey launched to hear from our users while we clued everything else up and a few days later we decided to close the library screen and add everything directly into the post to help speed them up even further.
Positive feedback from both surveys and customer success, partnered with reduced duplicate media numbers, were our signs that we did succeded. It may have been a small improvement, but it's speeding users up with faster post creation times, and over time, it will also have a noticeable savings of server space, leaving more money for the company.
That's a win!