Hello,
Here are some things I’ve done over the last seven years.* Please enjoy.
Tag Savage
* Mostly working at Tumblr.
Hello,
Here are some things I’ve done over the last seven years.* Please enjoy.
Tag Savage
* Mostly working at Tumblr.
Hello,
HEY WAIT
Wait, is this a Google Slides presentation?
Yes.
Not, like, a real website?
No.
Why?
Well, it’s really easy to keep up-to-date and change on the fly,
and also the challenge of a limited toolset is extremely appealing.
Okay. It looks super shitty on mobile, though.
I know. I’m sorry. It’s a little better if you rotate your phone.
Okay, let’s begin.
Tumblr Brand Guidelines.
Yep, we wrote some brand guidelines. Our marketing department did some research, and those of us on the
Creative team turned that research into actionable
instructions for anyone working/speaking/designing/whatever on behalf of the brand.
Here’s a smattering of pages about how to handle copywriting.
I wrote some pangrams for our type specimens. They’re vaguely Tumblr-themed.
Tumblrbot.
Over the course of 2016, we built an experimental chatbot on top of Tumblr’s messenging platform. Its functionality was limited. It could recommend posts and blogs based on ones you already liked, and...not much more than that.
We were determined to avoid the pitfalls of most quasi-intelligent chatbots, which tend to be overly clever or totally boring, and neither of those responses is welcome when the failure rate is so high. So we wrote Tumblrbot to be very earnest and slightly baffled. The theory was that users would endeared to this equal-parts helpful and helpless being in the same way they’re endeared to, say, a younger sibling.
Amazingly, it worked.
Tumblrbot’s brain is currently one giant spreadsheet of potential inputs and the range of responses they might elicit. It’s not a brilliant system, but it gets the job done.
Can you believe the responses? So nice.
Tumblr Decides.
Okay, so, a lot of brands do things for April Fool’s now, and most of it is groanworthy, and for a long time Tumblr didn’t participate because we didn’t want to be part of the marketing frenzy that April Fool’s has turned into. But at some point, we figured out that April Fool’s didn’t have to be a look-at-us jokey-joke. It could instead be a big art project for the whole community to take part in.
And so we started doing April Fool’s jokes.
In 2016, we held an election.
Throughout the day, we ran a series of videos about happenings in the election.
Partially due to budget constraints, and partially to create an anything-goes atmosphere, we worked in an extremely lo-fi aesthetic.
People made art.
Lots of it! All told, there were 750,000 more posts made on April Fools than on an average weekday.
Live Video Launch.
It was 2016 and everyone was launching a live video product and so was Tumblr. We were a bit late to the game, actually, so we needed to differentiate ourselves a bit.
We assumed that everyone already knew what live video was (since they almost definitely did) and rather than bore people with an explanation of the product we emphasized how Tumblr’s unique community would use it in fittingly unique ways.
Outside of Tumblr’s owned channels, we ran these simple banner ads. “This should be interesting” proved to be an effective wink to viewers.
Within Tumblr’s owned channels, we ran a multiple-hour live stream of compelling nonsense. People could tune in and out as they wished.
Tumblr Executive Suite.
Another April Fool’s joke that became an art happening. Ha ha, wouldn’t it be funny if Tumblr was a suite of business software? Sure, yes. We could have made a video goofin’ about that. But that wouldn’t be very participatory, and we wanted participation. So we made a bunch of little business widgets people could use to make posts (spreadsheets, note pads, clocks, stock tickers, etc.), and added a little virtual assistant called Coppy.
Coppy was a joke on Microsoft’s Clippy, which is a pretty obvious joke to make, but Coppy was such a helpless idiot that he quickly endeared himself to users rather than annoying them. At the end of the day, he got sick and died. A hero and a martyr, our Coppy.
Worth noting: Everything we do is a learning exercise, and Coppy’s can’t-hate-him combination of sweetness, dumbness, and eagerness formed the underlying personality of Tumblrbot two years later.
An introductory video
Left: Various UI elements of Tumblr ExecutiveSuite 2016.
Below: Posts created using various widgets in the “product.”
Coppy was so popular that people started making fan art about him.
Toward the end of the day,
he got very sick, but his commitment to helping users never waned. He was literally working himself to death.
Soon, the memorials started pouring in.
Net Neutrality.
Tumblr has a history of stirring up political actions in the name of internet freedom, and when net neutrality was threatened back in 2015, we gave our users a way to fight for it. A dashboard call-to-action popped open a tool for calling your representative and offered a loose script for what you might say. This has since become pretty common (5 Calls is doing an incredible job of reducing the friction in this flow) but it was pretty novel at the time.
We also ginned up some decent scare ads. Have a look.
An introductory video
Oh yeah, I got to talk about all this on MSNBC, which was fun.
Tumblr 4.3.1 Release Notes
Yeah, okay. This is another thing I have mixed feelings about. Obviously a lot of brands get silly in their release notes, and sometimes that feels good, and sometimes it feels self-indulgent. We usually try to 1) keep it light and quick 2) keep it informative but entertaining 3) not make shitty or obvious jokes. For whatever reason, though, we went way, way, over the top with 4.3.1. It got a pretty great reaction from most people, and then a bunch of backlash from the backlash-y channels, and then settled into an overall net-positive experience.
My main regret is not ending the notes with David saying “It includes bug fixes and performance improvements.” These things should be informative, after all.
Tumblr Terms of Service Rewrite
We wanted people to actually read our Terms of Service, so we worked to keep them as digestible as possible while retaining all the necessary legal specifics. THEN, on top of that, we interspersed it with plain-English translations of all the key points. We also peppered it with little jokes. Everyone needs little jokes.
Tumblr Presents a New Video Player
One of the most-requested features in Tumblr’s history was to simply make the video player more reliable. After dragging our feet for reasons I can’t recall, we finally did so, and added a few niceties in the process.
Given the long wait, it seemed like a good opportunity to swap out our normally laid-back approach to announcing new features with some deliberately over-the-top anti-marketing. Inspired by the unchecked pomp of movie theater pre-roll PSAs, we made...this thing.
An introductory video
Finally, a Search Page
When I first started working at Tumblr, we didn’t have a real search function. Can you believe it? All the search field did was return the most recent posts that were tagged with the term you searched for. You could say “architecture” 100 times in a post, forget to tag it “#architecture,” and no one who was searching for “architecture” would ever find it.
Moreover, the search results page was the same as the dashboard view—a long scroll of full-sized posts. Unskimmable, unsortable.
Once our Search team had the technology in place, I got to take the first crack at a web interface for it.
Heavily inspired by an appreciation for Google Image Search and a mild disdain for masonry grids, here’s what we came up with.
This was a web-only project and didn’t fully anticipate mobile views, but it was responsive across a considerable spectrum of widths. See the next slide...
The Tumblr Logo Page
Everyone needs a Tumblr logo! As part of helping redesign Tumblr’s logo, I also got to design the page where it was available for download, and write the guidelines for how it should be used.
Knowing that most do’s-and-don’ts around logo usage are either completely obvious or totally ignored, our guidelines anticipated how designers might want to actually use our logo and tacitly presented creative suggestions around that.
We wanted to be enablers, not enforcers.
(Hovering over one of the logo variations cycles through examples of the kinds of situations it’s intended for.)
Byliner
One of the first and most notably successful publisher of Kindle Singles, Byliner was making a play to develop their own device-independent reading experience on the web. I helped them extend their (pretty bold) visual brand language into a (handsome, unobtrusive) set of pages for reading longform pieces on any screen.
Memberly
Memberly was part of the circa-2010 “Kickstarter for x” craze, and in this case the x was “subscription services.” Anticipating the success of things like Blue Apron, Winc, Me Undies, and all those other podcast-sponsoring companies, it was both a white-label backend for running subscriptions and a marketplace for finding subscriptions.
For a number of reasons, it didn’t succeed. But the design was—for 2012, anyway—pretty thoughtful.
The consumer-
facing side of Memberly.
The Memberly provider dashboard.
Small Publishers
Real quick, I did some designs for a few mid-sized blogs back in 2010. This was before mobile-first thinking, so in functional terms they didn’t age well, but they were quite striking at the time. In any case, my pride for them is compact and true.
Okay, goodbye.
Email me at tag@tagsavage.com if you need anything.