Time Segmentation: Building Y(our) History on the Social Web

The series on time segmentation continues! So far we’ve covered time segmentation by-the-hour for streaming music, day and night cycles for Twitter, then speculated a bit about its possible application in election cycles as part of the community development process. Next we’re going to take the macro perspective all the way: how time segmentation can be used to make history — literally!

I’ve taken a look, but stonehenge.com doesn’t exist yet. They may be having trouble finding an admin maybe?

Given the forward-looking nature of the tech industry, most software companies tend to be fairly hesitant of looking “backwards” in time, believing that doing so will impede progress and development toward a better future. (Or better profits, at least!) Scientifically speaking, however, time doesn’t move in a straight line from a beginning to an end — all we’re really doing is counting the number of times Earth goes around the Sun (years), and the number of times it happens to spin around while doing so (days). Being that time flows circularly, it’s no surprise that our past never quite leaves us…and history does in fact repeat itself, for better or worse.

The most notable exception to the tech industry’s history-agnosticism is the Internet Archive’s Wayback Machine, which has attempted to create a historical archive of web’s entire history, all in one place. Comparing how the the web was way-back then to now gives us a pretty clear picture of how much things have changed, which helps to give us some perspective on where things have been, and where things may be headed.

Mcdonalds.com from 1996. In a lot of ways this site was way ahead of its time, paving way to the minimalist, primary-color aesthetic that’s commonly seen in today’s more contemporary designs.

While the Wayback Machine is an incredible achievement in its own right, in recent years it’s been having trouble adapting to Web 2.0+ environments, mostly because the Machine was designed with webpage views as the user’s primary point of contact. The Internet of today is a lot more “social”, however, especially as web traffic continues to move towards mobile as people’s medium of choice. Data from social media sources are much more difficult to archive, especially when a lot of it is private, disintermediated, inconsistent, or legally ambiguous.

As a former library technician, however, I do believe that there will always be a need for historical record — there’s always the grand vision of archiving humanity’s knowledge for posterity of course, but there’s also a lot of direct benefits that people can gain from these practices, even in their day-to-day lives. And many of these ideas can be built into software platforms as features and methods of increasing engagement and discovery.

The next big question for the World Wide Web: Can history be made social?

Personal vs. Collective History

If you’ve been on Facebook lately, you might have noticed that they’ve recently introduced a “memories” feature that brings up older posts and updates from the past — while still in its early stages, it marks an interesting turn for the company since in many ways it’s the first time a social media platform became explicitly interested in “socializing” users’ historical data.

There are some issues, however. Mainly, this:

Not all memories are good memories.

Now there are a lot of people who genuinely appreciate the “memories” that the platform recalls for them, but I think that Facebook has greatly underestimated how risky this endeavor really is, for a number of reasons:

1. It pops up right at the top of your feed, whether you want it or not. (This isn’t such a big deal since it can be easily fixed, though.)
2. Bad or embarrassing memories…well, make people feel bad. Machine learning algorithms help to filter out unwanted content, but they aren’t known to be perfect, and probably won’t be any time soon.
3. Neutral/inconsequential memories (which is the majority of its output) come across as being spammy, making people take the feature less seriously.
4. Even good memories (ex. above) can trigger painful emotions in the user. Even when a memory is undeniably good in every way, many people will be reluctant to relive their past glories since it makes them look like they can’t let go of the past. (An implied social taboo.)
5. Because people tend to view personal data as private (even if it isn’t, legally/technically), some users might find the memories feature very intrusive — worse case scenario, it may even lead to a lawsuit if someone gets offended enough.

The problem with Facebook’s “memories” is that all it takes is one bad experience with the feature for people to be turned off from it entirely. One embarrassing/painful memory is enough to outweigh any sort of benefit that you might get from being able to share a cool thing you did a few years ago. So it’s not really a question of how high of a statistical % you can get on the good/bad ratio — there’s actually very little room for error, if at all. Facebook is playing with fire here, and in a way, the feature is just a time bomb waiting to go off at any moment.

Collective History Solves Most of the Problems Related to Personal History

So the header above pretty much says it all: group histories are much more useful and engaging than personal histories, especially for social media platforms. Here’s why:

1. Group histories give people a sense of collective purpose and belonging, while also providing the necessary material to create overarching narratives.
2. “Bad” memories — the struggles that a group of people goes through — can be turned into narratives of redemption and perseverance. Stories of triumph, struggle, and glory are much more powerful when you have others to share the experience with.
3. You don’t have to worry about privacy issues nearly as much because group activities are generally considered public expressions.
4. Group histories by default are much more readily legitimized than personal ones, given that the latter requires a level of celebrity/stature for it to be considered interesting. (i.e. much more scalable)
5. Group histories don’t have the personal and psychological baggage that comes with “digging into your past”. Historical problems are seen as problems to be overcome, lessons to be learned — rather than something that’s seen as debilitating or an hindrance to an healthy lifestyle.

Sports teams are actually the easiest example to use as an illustration of all of the points above: A true fan sticks with their team through thick and thin, through the good times and bad. Reasons for the team doing well are points of conversation, but when a team isn’t doing well, those issues are also brought to light. You’re always competing against other teams of course, but at least within the team you experience a kind of cohesion between the players, fans, and staff members that you don’t generally get anywhere else. The draw of sports as both a player and spectator is indeed very strong — for at least a brief few moments, it allows people to set aside their differences and just focus solely on the game.

Now there’s an interesting thing about sports that often goes overlooked: it’s a system that runs on time, mainly that of seasons. (Something I was working on at my previous job, incidentally.) The Golden State Warriors of 2016 is an entirely different entity than that of the Warriors back in 1996, with each season becoming an entity onto itself. Online, group entities are generally treated as singular objects (including Facebook) but any sports fan would know that the story behind a sports team varies wildly from year to year just because the roster and staff are always in shift.

Now this idea can be applied more broadly: the exact years you spent in school, were part of a club/association, served in the military, played for a musical ensemble/band, worked at a company, were part of a political movement, were part of an expedition, spent volunteering for an organization…all play a critical role in identifying the whos, whats, whens, and wheres of your experience at the time. This puts the segmentation in “time segmentation”, which allows you to segment social entities into specific enough categories to the point where you might be able to connect to others on an individual level.

Facebook is probably the platform most compatible with social time-segmentation right now — these features have yet to be built, though.

Being that Facebook built its original networks upon college mailing lists, they’re actually well-positioned to take advantage of this methodology since much of its user-base are actually used to seeing these types of structures applied in social settings. If the company was willing to put make their platform more group-oriented (rather than individual-oriented) they could make significant headway in this area, solidifying their position as the most popular social media platform on the web.

If you played the violin in the top orchestra while you were college, for example, you might list that as part of your activities or “projects” as part of your college experience. Is anyone really going to care after you’ve left school itself, though? If you segment this information by time…say, that you attended schools during the years of 2009–2013, then the scope of the search becomes much more manageable.

In my personal research, I know that there’s a lot of people out there looking to reconnect to people they used to know in this way (especially military veterans), but don’t have the time to go through the time consuming process of going through alumni or veteran organizations to find everyone one by one. Here, you’ve managed to solve an age-old problem just by adding a little bit of metadata to an object that normally sees no usage anyway. It gives people the incentive and a compelling reason to use it, while greatly reducing the friction involved in trying to do it in the real world.

That’s it for this section of the time segmentation series, hope you’ve been enjoying it so far! Always interested in hearing comments and feedback, if you have any.