Where Are the Family Plans for Web Apps?
The Internet treats us like we’re loners, but the fact is, we’re not. We have partners and families and kids. That’s very normal.
The gap between the individual and the household is perhaps most evident and annoying on personalized recommendation sites, like Netflix and Amazon. The person who loves “Yo Gabba Gabba” is not always the one who watches every Jennifer Aniston movie or the one who can’t be budged from the Criterion Collection. But then, we watch movies and TV together, or sometimes we’re just in the mood for something different from our usual taste.
Netflix is asked to address this personalization problem all the time by analysts and users. On Netflix’s most recent earnings call, CEO Reed Hastings said to expect “multi-account options” within the next year. The company has talked about making this a premium feature, where customers pay extra to maintain multiple user accounts on one bill.
Single-user accounts are certainly more straightforward to manage for both companies and users, but they ignore the reality that we live together and depend on each other. If a personalization system doesn’t even understand the difference between two people, then it’s not very good, is it?
A very interesting Netflix developer post from this weekend about the company’s recommendation system noted that the company already implicitly tries to acknowledge different users in a household. “It is important to keep in mind that Netflix’ personalization is intended to handle a household that is likely to have different people with different tastes. … To achieve this, in many parts of our system we are not only optimizing for accuracy, but also for diversity.”
Because of this focus on diversity, the “Top 10” recommendations on each user’s personalized Netflix homepage are really for each household, wrote Netflix personalization engineers Xavier Amatriain and Justin Basilico.
Really? Because when I log into Netflix on my joint account with my husband, I see “Top 10 for Michael.” It doesn’t exactly make me feel acknowledged.
But Netflix isn’t the only one. A household or “inner circle” setting makes sense for any online content seller, from the New York Times to iTunes. For instance, the household finance manager, Mint.com, has no household account option. Two people who use the service have to associate the account with one of their emails and a joint password. I don’t have high expectations of the wide range of banking and utility bill sites, but they suffer from the same problem.
Or how about counting multiple devices as members of a family? Family mobile data plans are finally supposed to arrive in the U.S. this year.
I’d propose that these sites and services create a level between billing account and user account. They should understand and design around the idea of a household and allow users to switch between profiles — like players on gaming consoles — without logging in and out each time.
PandoDaily writer Hamish McKenzie seems to be thinking along some of the same lines in a post today suggesting that Pair, the new app that helps romantic couples share their days with each other, could become a platform for joint management of calendars, travel and bank accounts. I think he’s getting at a similar concept — but this is probably a setting that would be better done by each service individually, rather than an outside authenticator.
These joint household accounts would present challenges — and opportunities. It’s not as simple as designing for just one person or the general public.
For one thing: privacy. Are you as a household member allowed to see everything your family watches, reads, listens to, pays for, etc? I’d say generally it makes sense to be open within a household, but there are reasons to build in secret or incognito activity. For instance, I don’t necessarily want my husband to be able to anticipate what I’m giving him for his birthday by seeing the purchase pop up on our joint Mint account.
There could also be interesting design around the Venn diagrams of multiple users. I was chatting about this topic with investor Manu Kumar of K9 Ventures, who suggested a Netflix interface where people sit down to watch a movie and tell the service who is in the room — kind of like a check-in. Then, Netflix would be able to recommend something for the group, and also capture better data about how to make recommendations for each person in the future.
I’d be interested to know what you think, and whether you’ve seen examples of people doing this well.