July 25-27, 2018 / Vancouver, BC

Re-Finding in an Algorithmic World - Pam Drouin

September 6, 2016

[responsive_vid]

Speaker - Pam Drouin

Transcription coming soon

This is the title of my talk. I wanted to pause, because otherwise, we’re not going to spend any time on it. Algorithms are everywhere, as Sara talked about at length this morning in a very important talk. We encounter them while Google, catching up on our Facebook newsfeeds and viewing Amazon recommendations, have you ever thought how algorithms might be altering the way people might be navigating online. My name is Pam as some of you know, thank you for the introduction, I really care about information architecture and how people find information online. Algorithms are affecting

How we find information, as well as our ability to find information again, over time.

When I talk about refining information, I’m referring to one of the information-seeking modes that Donna Spencer talked about some ten years ago through user research, Donna observed that people exhibit one of four search behaviors. The first is known compile seeking. I know the exact photo I want, I just have to find it. No. 2 exploratory seeking, I know I need a photo but I don’t know which one. No. 3, I don’t know what I need to know. I thought I needed a photo, but it turns out I need a map instead. Refining, I need to find a photo that I saw a few years ago.

As an industry, I feel like we do pretty well with the first three modes of information-seeking. We do a really good job of meeting those needs. Refining things, though, is trickier, especially as time goes by and algorithms are complicating this. Take a look at Netflix, a product that I use probably almost daily. I have a lot of passion for it. It turns out that everyone’s My List is sorted by an algorithm that orders the shows according to what they think you might watch based on a mix of variables. Recently right before my in-laws came to visit from Spain I added a bunch of Spanish language movies to my list, but when we were ready to watch a movie, I couldn’t find any of them in my list. They were scattered throughout, not organized by how I had added them. It turns out you can turn this setting off, but it takes some research and while I was going about my research, I discovered that many Netflix customers are also many confused about why their lists have to be ordered on their behalf.

Another example, I’ve been thinking about is media, another thing that we probably use almost every day. Recently I wanted to find a medium post that really stuck with me that I had read a few years ago, but I couldn’t locate a list of ones that I liked. Medium calls them recommended posts. I tried using search, but it turns out their results are order of magnitude by a relevance algorithm that you can’t sort chronologically.

I eventually found the post I was looking for when a friend of mine told me that you can look on your public profile to see a list of things that you recommended over the years. That’s not somewhere I thought to look. So it turns out when you’re trying to refined a post on medium, you really have to hope that you liked it back then. The last example: Another product I use multiple times a day. Instagram.

Just last month they rolled out an algorithm that puzzled a lot of people once it rolled out. Many users, me included once we woke up and opened up our Instagram, we thought it wasn’t refreshing the content because we saw yesterday’s photos. All of a sudden, you know, we’re seeing old posts first, not new ones, what’s going on? It’s disorienting to see things out of place all of a sudden. Instagram says they’ll determine what we care about, in order to show us those moments first, but I don’t think they’ve quite gained our trust yet, because you can still see users — Instagram users on Twitter complaining about the user experience.

Algorithms are altering the wayfinding that helps us find and refined information over time. According to Christina Waktke … Wayfinding is grounded literally in proximity. We tend to think of proximity in terms of nearness in space or distance, but it also means nearness in time. We’re taking time for granted and confusing our users in the process.

Algorithms are powerful tools. They affect people’s lives every day, but we need to consider proximity and making that explicit in all senses of the word to users, in order to help them refined information over time.

Architect not just for now, but for tomorrow, and for years to come. Thank you.