Ten Years In: Finding Balance to Enjoy Twitter

Last week saw my 10th anniver­sary of being on Twit­ter (as @stopsatgreen). That’s a long time, but I’m still there and still active because I still get huge val­ue from it. I don’t want to down­play that, for some peo­ple, Twit­ter became very tox­ic and com­pelled them to leave; but for me, no oth­er net­work has come close to match­ing the expe­ri­ence it pro­vides.

Over the course of my ten years I’ve devel­oped a few rules that help me con­tin­ue get­ting the most from Twit­ter; keep­ing my time­line fresh,  inter­est­ing, and valu­able. I’ve shared them here on the off-chance that they’re use­ful to you too, dear read­er.


Twitter, Listening to Users, and Murder

I saw this car­toon gain a few thou­sand retweets on Twit­ter. In it, a Twit­ter exec­u­tive asks three col­leagues how they should grow the ser­vice. One col­league says “Algo­rithms”; anoth­er, “Moments”; a third says “Lis­ten to users”. This third response angers the exec­u­tive, to the point that he throws the man who sug­gest­ed it out of a win­dow (it’s at least a sec­ond floor win­dow, so this is pre­sum­ably mur­der).

What I infer from this car­toon is that the author believes that Twit­ter doesn’t lis­ten to its users, but should.

So what do Twit­ter users want? The #RIPTwit­ter hash­tag1 reveals a few com­mon demands.

One is that Twit­ter needs an edit but­ton. That opens the door to so much poten­tial abuse that I can’t believe it’s seri­ous­ly being pro­posed, let alone con­sid­ered:

If you get angry at peo­ple who retweet big­otry, abuse of mar­ket­ing mate­r­i­al, just imag­ine how you’ll feel when you find out you’re the one retweet­ing it.

Anoth­er is that Twit­ter needs to con­cen­trate on stop­ping abuse. Bri­an­na Wu, who has more rea­son than most to want an end to Twit­ter abuse, says that this posi­tion is non­sense:

As some­one that works with Twit­ter fre­quent­ly on harass­ment, I feel unique­ly qual­i­fied to say… [this] is bull­shit. Twitter’s harass­ment out­come is improv­ing. I have doc­u­ment­ed, sta­tis­ti­cal proof it’s improv­ing.

So giv­en that two of the most pop­u­lar user requests are rub­bish, and many vocal Twit­ter users seem to real­ly just want to pre­serve the sta­tus quo, and that Twit­ter growth con­tin­ues to stall, my take on the car­toon is that the exec­u­tive was right to get angry at the per­son who sug­gest­ed they just lis­ten to users.

Although I don’t con­done mur­der.

Yes, I write about Twit­ter quite a lot. That’s because it’s impor­tant to me, I use it fre­quent­ly, every day. I want to see it suc­ceed, and I want to see it improve. And, as M.G. Siegler notes in Tem­pest in a Tweet­pot:

Change is always scary — espe­cial­ly on the inter­net. But time goes on, we move on, and every­one is often hap­pi­er as a result.

1 Iron­i­cal­ly, Twit­ter search uses a non-lin­ear algo­rithm, and is bet­ter for it.


How farmers in Myanmar shape the future of the web

The Atlantic pub­lished The Face­book-Lov­ing Farm­ers of Myan­mar this week. It’s a fan­tas­ti­cal­ly inter­est­ing write-up by Craig Mod of ethno­graph­ic research he under­took in the South­east-Asian coun­try. What makes that mar­ket so par­tic­u­lar­ly inter­est­ing is that Myanmar’s net­work was until recent­ly arti­fi­cial­ly capped and con­trolled by the rul­ing mil­i­tary jun­ta, but a vast num­ber of peo­ple have sud­den­ly come online as the price of a SIM card plum­met­ed:

Mobile SIM cards in Myan­mar have his­tor­i­cal­ly been pro­hib­i­tive­ly expen­sive. In 2014, the cost of a SIM card dropped from about $2,000 USD to $200 USD and then once again, to $1.50 USD.

In a coun­try of 53 mil­lion peo­ple, only 12% had access to cell net­works in 2014. By this year, the government’s plan is to have 74% online. That’s a huge influx of users, and large­ly in rur­al areas. The phones of this new wave of net­worked farm­ing users are com­mon­ly import­ed from Chi­nese man­u­fac­tur­ers, often bought sec­ond-hand. The biggest imped­i­ment to using the inter­net is the cost of data:

They feel each megabyte. For about 10 U.S. cents you can pur­chase 25MB of data. If you buy in bulk (although almost nobody does) you can get 2GB of data for 11,900 Myan­mar Kyat or about $9.20 USD. Most farm­ers grab dataon their scratch cards in 1,000 or 3,000 or 5,000 Kyat chunks. How long it lasts depends on the user. For some 3,000 Kyats gets them through the month. For oth­ers, it lasts only a few days.

What the research dis­cov­ers is that every­one is on Facebook—and often noth­ing else, except per­haps a mes­sag­ing app. The rea­son has lit­tle to do with the social graph, and more with that data restric­tion:

Face­book has a com­pelling advan­tage over oth­er news apps or even Twit­ter: The con­tent of many posts and news items live inside Face­book itself. There are exter­nal links, but most of the arti­cle sum­maries and pho­tos are self con­tained. As Face­book con­tin­ues to ramp up their Instant Articles—special ver­sions of web arti­cles that are lean­er, load more quick­ly, and are Face­book optimized—the amount of con­tent that lives in Face­book will only increase. For those who are data sen­si­tive, this is a clear virtue.

I wrote recent­ly about Twitter’s planned move to allow 10,000 char­ac­ters in tweets, quot­ing Will Oremus’s arti­cle Twit­ter Isn’t Rais­ing the Char­ac­ter Lim­it. It’s Becom­ing a Walled Gar­den. and its con­tention that:

If I’m right about what’s real­ly going on here, this move will not fun­da­men­tal­ly alter how Twit­ter looks or feels, nor how peo­ple use it. Rather, it will change where online con­tent is host­ed, who con­trols it, and who is in a posi­tion to mon­e­tize it.

I’m not naïve enough to think there’s def­i­nite­ly not an ele­ment of con­trol involved, but my orig­i­nal response was “it’s prob­a­bly more com­pli­cat­ed than that”, and I still think that way. To me this is not nec­es­sar­i­ly about monetising—not yet, at least—but more about being rel­e­vant. Twit­ter are tak­ing a look at the emerg­ing mar­ket and see­ing them­selves with no place in it. Right now, it’s not even a com­pe­ti­tion.

The oth­er com­pa­ny that stands to lose big from this is Google—the open web pow­ers pret­ty much every­thing it does, so los­ing out to Face­book on the next bil­lion online users would be dev­as­tat­ing to its busi­ness. Hence they’re push­ing Accel­er­at­ed Mobile Pages (AMP), a strict and opti­mised sub­set of HTML aimed at pub­lish­ing sto­ries with­out lega­cy adver­tis­ing and track­ing cruft. AMP is set to launch next month, with Twit­ter on board as a part­ner.

Why employ a strict sub­set of HTML rather than get­ting pub­lish­ers to clean up their acts and pro­duce bet­ter-opti­mised pages? Time. Although per­for­mance is a hot top­ic in web devel­op­ment cir­cles, web pages con­tin­ue to get heav­ier. It will take more time to cor­rect that, and the rapid growth of mar­kets like Myan­mar shows that time is an ill-afford­ed lux­u­ry.


Twitter, Good Faith, Skepticism, and Open-Mindedness

Yes, this is anoth­er piece about Twit­ter. But it’s also not real­ly about Twit­ter.

Jack Dorsey all but con­firmed a report that Twit­ter are going to allow up to 10,000 char­ac­ters in tweets (like­ly as a kind of media item attach­ment):

We’ve spent a lot of time observ­ing what peo­ple are doing on Twit­ter, and we see them tak­ing screen­shots instead of text and tweet­ing it. Instead, what if that text… was actu­al­ly text.Text that could be searched. Text that could be high­light­ed. That’s more util­i­ty and pow­er.

I wrote a piece say­ing I thought it was a pret­ty good idea. Sur­pris­ing­ly, not every­one agreed with me; for exam­ple, @chiller tweet­ed this:

Things no @twitter user wants:
— Moments
— Non chrono­log­i­cal tweets
— Tweets from peo­ple we don’t fol­low
— Pro­mot­ed bs.
— More than 140 chars

I have a few quib­bles with that, not least that I do want more than 140 chars (iron­i­cal­ly, the author could have used the full word ‘char­ac­ters’ if more were avail­able). But that tweet has 6,800 retweets at the time of writ­ing, so per­haps I’m in the minor­i­ty.

Will Ore­mus wrote a piece for Slate say­ing that Twit­ter are mak­ing this change just because they want to keep peo­ple on the plat­form in a ‘walled gar­den’:

What’s real­ly chang­ing here, then, is not the length of the tweet. It’s where that link at the bot­tom takes you when you click on it—or, rather, where it doesn’t take you. Instead of fun­nel­ing traf­fic to blogs, news sites, and oth­er sites around the Web, the “read more” but­ton will keep you play­ing in Twitter’s own gar­den.

Who should we believe? Jack, when he says this is just stan­dar­d­is­ing user behav­iour? Will, when he says that Twit­ter are not doing this for users but as a land grab? Or @chiller, when they say that users don’t want this at all?

I don’t know. I sus­pect there might be a lit­tle bit of truth in all of those posi­tions.

Per­haps that’s fence-sit­ting. How­ev­er, there are two things that I think are impor­tant to remem­ber when we’re talk­ing about sit­u­a­tions like this: Cen­ny­dd Bowles’ plea for good faith, It’s Not What You Think; and Carl Sagan’s expla­na­tion of good sci­ence, The Bur­den of Skep­ti­cism:

It seems to me what is called for is an exquis­ite bal­ance between two con­flict­ing needs: the most skep­ti­cal scruti­ny of all hypothe­ses that are served up to us and at the same time a great open­ness to new ideas. Obvi­ous­ly those two modes of thought are in some ten­sion. But if you are able to exer­cise only one of these modes, which ever one it is, you’re in deep trou­ble.