The Blog

Algorithm Tinkering - Twitter and Facebook Creating More Problems Than They're Solving?

The problem with this is that rather than solving a problem that exists, Twitter seems to be identifying one which doesn't, then creating another which didn't.

Late last year, speculation grew that 2015 will see Twitter tinker with the algorithm which dictates the feed a user sees - moving from a system where tweets come in chronological order (aside from the promoted tweets), to one which is more aligned to what an algorithm thinks 'you' - the user - will find most interesting right here, right now.

This has in part been fuelled by indications that user figures are beginning to plateau according to the company's recent financial announcements - the latest figures show the lowest growth on record, (and despite forecast-beating revenues). The decision to change the algorithm could therefore be interpreted as an unnecessary knee jerk reaction, (especially given iOS8 was cited as a source of lost users who simply didn't re-install the app); if your user growth is not as high as expected, yet revenues are, shouldn't it be more sensible to watch developments rather than immediately react and change the basics overnight?

Either way, the results seem to have spurred Twitter to continue the roll out of a shift in algorithm - one which will disrupt the user experience hugely. The problem with this is that rather than solving a problem that exists, Twitter seems to be identifying one which doesn't, then creating another which didn't.

Facebook has suffered similarly - earlier this year it changed how users saw information in their own feeds, again moving away from the chronological route and moving to a more contextual and interpretative algorithm. This hasn't been wholly positively welcomed, (I personally hate it), and it's strange that Twitter should seek to go down the same route - especially given the platform's core attractions are based on time-specific discussion, sharing and engagement. Having days-old conversations appear in timelines because an algorithm decides they're now relevant will simply confuse the whole premise of the service. (The difference between the two models was well summed up by the BBC College of Journalism's Marc Blank-Settle who pointed out to me that while Twitter was covering the Ferguson riots, Facebook was still awash (pardon the pun) with videos of Ice Bucket Challenges).

This is particularly intriguing given Facebook's recent surge of advertising. It's not had to promote itself for years due to its huge word-of-mouth (and word-of-social) explosion, yet now the Underground is awash with tickboxes against the words 'Friends', and TV ads are being equally heavily pushed. It begs the question 'why?' - has the platform simply plateaued, or has user engagement dropped as people get jaded or simply don't like the shift in feed prioritisation that's been forced upon them? Twitter could learn plenty if an an algorithmic tinker has now caused Facebook to lose users, and thus revert to advertising to plug itself.

Discussing it with Sue Llewllyn, an ex-journalist and broadcaster and now specialist social media trainer, she suggested that while Twitter should absolutely continue to find new ways to engage and retain new users, relevance must remain key. "We've all chosen who to follow and personally curated our Twitter experience and that's the fun of it," she said. On the proposed revision of the timeline, she suggests Twitter could have its way and retain users - on a proviso; "an opt out button would solve the problem of upsetting loyal users who don't want an algorithm serving up stuff we neither like nor want," she adds.

With this opt in / opt out element in mind, it would be intriguing to see how developers react to users' responses if (or rather 'when') it arrives; there could be an obvious opportunity for someone smart to develop a plug-in which reverts the platform to the traditional time-centric feed. Looking back at how Windows 8 was received when it came out, with its massively revised user interface and baffling menu systems, there was a groundswell of users finding out how to revert back to the original desktop.

The fact remains however, that innovation is all well and good if it's well received, but if it immediately alienates users it becomes counter intuitive. In listening to its investors' demands about what they perceive as user dissatisfaction, rather than boldly responding that a plateau based on a single-quarter forecast is not basis for a shift in fundamental user-experience, Twitter could be risking forcing the plateau effect further. It's putting the cart before the horse - why initiate a shift in user experience without asking users? And why do so without looking at the problems a major rival, Facebook, has in having done the exact same?