I’ve been on hiatus from the Yale Tumblr for a bit (which is why it’s been pretty quiet over there), but I’m back in the driver’s seat, part-time, starting on Monday, and I’ve been thinking once more about measuring success. How do I know if I’m doing it right, and how do I know that it’s getting better over time?
Previously, I answered the question by total number of notes (likes and reblogs) per day. I soon realized that the graph fluctuates wildly based on things out of my control. For example, old posts surge when they get discovered and reblogged by a popular blog, adding to the total tally, but that success can obscure recent poor performance.
So I’m taking a step back and asking myself which numbers would be useful in helping me understand whether anything I’m doing is moving the needle one way or another:
"On average, how many notes does each post get over the course of a week? A month?": If in January, each post got an average of 20 notes, and in March, each post got an average of 40 notes (hypothetical numbers), I know I’m doing something right. This auto-generated value would be smart enough to exclude large spikes that could skew the results.
Attrition: I keep a Google spreadsheet where I track the number of followers by hand, but I worry that new followers are masking unfollows. I would like to know, on average, how many days someone follows us before they decide to unfollow. If this number gets smaller, it might mean that people are following, quickly getting bored, and unfollowing. If it get larger, it means people are hanging out for longer, maybe because they like what I’m doing.
Anecdotal evidence: Numbers will never tell the whole story. A post might be blowing up because everyone is screaming in the reblog comments. Often, I can only tell how things are actually going by reading through reblogs and fan mail feedback. And that’s okay.
For the purposes of reports and the like, I’ll have to start keeping a folder of the nicest comments, organized by month.
Note: I know Tumblr provides an in-built analytics dashboard for advertisers. Does anyone have a screenshot? I’d love to know what Tumblr considers important to track.
I’m curious how other people evaluate success and growth (or otherwise). Say hi: equartey at gmail dot com.
About a month ago, I delivered a presentation to communications officers from across the University introducing them to the Tumblr platform, explaining how Yale approaches it, and some ideas for best practices.
The theme of the presentation was, simply, “Love your fandom.”
The SlideShare of the presentation is embedded above.
Hope this is useful. Kindly let me know what you think/point out typos at equartey [at] gmail.
Especially curious to hear whether this resonates with other higher-ed social media managers. Does this seem about right?
[UPDATE: Tumblr fixed this issue!]
Imagine if every time you posted an image to Tumblr, it automatically zoomed into a random part of the image and applied a filter over it? You would be confused and frustrated, right?
That’s how I feel about how Tumblr displays videos inside the dashboard.
Videos are a very labour-intensive content type, requiring days or even weeks of planning, scheduling and shooting. This is why it’s heartbreaking to put resources towards something, only for it to show up in the Tumblr dashboard looking entirely unrepresentative of effort that went into it.
For example, a few weeks ago, we invited people to submit questions to popular science writer Carl Zimmer via the Yale Tumblr’s Ask Box, and we turned his filmed responses into a YouTube video.
This is what that video looks like on the Yale Tumblr page:
This is what it looks like within the Tumblr dashboard on the Android app for Nexus 7:
This is what it looks like when it shows up on the Tumblr dashboard on my laptop:
It’s that last screenshot that I find frustrating. Why is it that when Tumblr previews videos within the dashboard, it ignores the specified thumbnail image? Remember that YouTube allows certain brand pages to specify a specially prepared image as the video’s thumbnail. The choice of thumbnail is an editorial decision, much like the decision about what to run as the first page story of a newspaper, or the art direction of a magazine cover.
Why does Tumblr apply a white noise filter over the preview? Even if the randomly-chosen thumbnail landed on one of the film’s most visually interesting scenes, you wouldn’t be able to tell.
One of the great things about Tumblr is that it empowers social media managers like myself to go to the powers-that-be and make the argument for investing greater resources into our content. When this happens, the entire platform benefits because the overall quality of the content on the site increases.
But what incentive do I have to make good-looking videos if they’re going to show up smudged and unrecognizable inside the dashboard, where the social action happens?
I wince a little every time I see a garbled video on my dash, and I’ve started looking for alternatives to posting straight video. I’m currently experimenting with graphics with a click-through to the video. For example, I reposted the Carl Zimmer video as the following:
Here’s another one I did for a different video - I included a YouTube play button to make it more obvious that it’s a video:
This is obviously not ideal. I would much rather do a real video post, but linked graphics are currently the option that gives me the most control over how videos appear in the dashboard, so I’m going to keep playing with different templates. Apologies for ranting, but this has been bothering me for a really long time.
Does anyone have any idea why Tumblr ignores embedded thumbnails and smudges the preview? Does this bug anyone else, or am I just crazy?
Everything is out of chronological order - it’s like the twilight zone. Is this just me or is this something affecting other people as well?
Hello, Just wanted to let you know that this is not a new feature, but rather an issue that our engineers are aware of and will fix it as soon as they can. We appreciate your patience.
"Ask X" is something I’m trying out on the Yale Tumblr. It’s an opportunity for anyone in the world to connect with people who’re doing inspiring work in the fields of business, science, politics and art.
The first two “Ask X” events were with Pentatonix and popular science writer Carl Zimmmer. Here’re a few thoughts about the “Ask X” format: