"Twitter drastically reduces the complexity of its signup process, which brings double-digit growth..Wall Street is happy, but engagement suffers, and revenue/user drops. What metric really matters now?"
Call me old-fashioned, but revenue is revenue. Slicing it up as 'revenue/user' is artificial, because unless you also marry 'revenue/user' with 'cost of user acquisition', 'revenue/user' is (almost) meaningless. Example: If Twitter suddenly got another, say, 30 million users overnight for essentially $0 in acquisition costs (perhaps just the incremental network and server costs), 'revenue/user' would indeed decline overnight. Why should they care?
Revenue and cost of generating that revenue are what really matter.
I agree with you (I'm old fashioned too), but revenue/user is a common benchmark used by Wall Street to judge a social network's performance. It's just another way of slicing the data. Overall revenue growth is still very important.
I recently signed up for a new Twitter account, and I was surprised by how difficult it was, even for someone who has had previous twitter experience. I was 'required' [1] to go through several steps, and each step required me to think heavily about what I wanted to achieve. Then, when I completed each step, the results were of questionable value.
Twitter really needs to redesign their on-boarding process. If they want to attract the middle-market, it needs to be super easy and immediately valuable. As simple as, 'Enter a few topics of which you have interest', followed by a filtered list of really interesting tweets and some recommended people to follow.
I remember how quickly Google gained the search market around 1999 - 2000. It was dead simple to use and immediately valuable.
[1]: The steps were actually optional, but I had to look really hard to find the 'skip' link.
I have seen this "what is that we don't know" argument used over and over again. The primary claim being the other side has more data and can thus make more intelligent decisions.
A big problem here is that more data somehow translated to better decisions. The second being that the twitter folks are intelligent than me (they could very well be but cannot derive at that decision. which is to say, you simply have to think that you are smarter than them).
Agree. More data does not automatically equal more or better insights. It's a well-known problem in the Market Research world that frequently, more data clouds issues and makes focusing on core problems and insights MORE difficult.
The concept of "don't give suggestions to people who have more data than you do" makes no sense to me at all.
It's very easy when you have mountains of data to spend all your time solving the problem of what the data might mean instead of solving the problem of how best to get people to sign up.
Also, when you look at a thing someone has build or a process someone is using to accomplish a thing and you say, "I really don't like the way this works. Maybe you should try something different." You are, in fact, adding to that mountain of data that twitter already has.
It's up to twitter to go back to the data and ask, "How many other people feel the same way? Did we make this decision based on usability testing? Did we make it because we wanted to make sure we were collecting certain key metrics? Are the reasons we made this decision solidly grounded still?"
I'm reminded of how difficult it was to create a new Flickr account recently. In order to sign up for Flickr, I really have to sign up for a Yahoo Account. This has been the case since ~2007, but it hasn't gotten any easier. I need to enter my first and last name, an email address, a (fake) phone number, my birthday, and my sex. All of this just to start using their 'beautifully redesigned' photo-sharing site or app.
Inversely, Instagram let me register with nothing but a username and password. No wonder Instagram is eating everyone's lunch.
Instagram has no way to sign up via the web. While I understand that the vast, vast majority of people will just sign up for an account on their phone, the last thing I want to do is give that app (and Facebook) carte blanche to my contacts list.
I had to create an account on my friend's phone because there was absolutely no way to do that via their website. This is just to be able to use their API!
I wouldn't expect to be able to sign up for a mobile app not through the mobile app. You also have the choice of granting them access to your contacts list, it's not mandatory or surreptitious.
It's not just a mobile app anymore. You can view all of the images/comments/etc on their web site. The only reason why they restrict you to signing up via the mobile app is because they want to force you through that flow.
UX 1.0: provide a shortcut for 'power users' or experienced people. For Twitter: stop making me follow random people just to get on board. Let me create my account and take it from there.
I'm having trouble seeing any point to Twitter anymore, other than as a OAuth source. I'm not sure I would sign up for it, if I didn't have an account there.
I've never been able to see the point. I signed up years ago, and abandoned my account a few weeks later. It has always struck me as a platform for self-absorbed 'celebrities' and marketers to talk to each other.
I'm surprised that when you start a new account, Twitter places so much focus on celebrities and so little focus on local users. I'm aware that connecting you with local users would require more personal data, but getting connected with a group of locals provides many more opportunities for real meaningful interaction.
So am I. I'm not interested in the thoughts of celebrities. There are better ways of getting general news but I do want local information which is where twitter could do a lot better.
In the original article, the author has a screenshot that has suggested followers. He apparently lives in NZ, and Kim DotCom (as well as other NZ handles) was recommended to him. Seems like there's some localization happening.
It just occurred to me to wonder if onboarding new users is something better handled by a real person rather than a computer program. A person can better spot where a newcomer would have problems. They can also (on average) do a much better job tailoring the onboarding experience than a computer could.
Onboarding is not an optimization problem. it is an AI problem. As in, if you're going to use a computer to do it, it's going to need decent AI to make it work.
Call me old-fashioned, but revenue is revenue. Slicing it up as 'revenue/user' is artificial, because unless you also marry 'revenue/user' with 'cost of user acquisition', 'revenue/user' is (almost) meaningless. Example: If Twitter suddenly got another, say, 30 million users overnight for essentially $0 in acquisition costs (perhaps just the incremental network and server costs), 'revenue/user' would indeed decline overnight. Why should they care?
Revenue and cost of generating that revenue are what really matter.