Q&A: Digg’s Andrew McLaughlin and Jake Levine Race to Replace Google Reader
For people who have grown accustomed to having their online news delivered throughout the day in neatly stacked headlines from all their favorite blogs and publishers, this next week is a tough one. It’s when Google lays to rest its eight-year-old Google Reader project — which the company decided wasn’t worth the overhead as usage declined.
The ensuing backlash Google faced from its hardcore Reader-loving following has helped spark development on alternative services. Feedly just released an all-Web version that’s a fuller replacement to Google Reader, even AOL has one of its own coming soon, and the once-great online news aggregator Digg is being reimagined as a feed reader replacement under new ownership.
While Google Reader was a long-neglected product, its strength came in part from being part of Google’s larger whole, so it’s unclear that these upstarts can extend the concept in useful enough ways to be self-sustaining.
The Digg project comes with especially high hopes, given that it’s being incubated by the New York City data-product wizards at Betaworks, with former Googler Andrew McLaughlin at the helm. It’s starting to roll out to testers now, with a Web version and an iPhone and iPad app launching this week, and Android coming before the end of July, according to McLaughlin.
As the small Digg team sprints to the start line, we chatted with McLaughlin, Digg’s CEO, and Jake Levine, its general manager, about their high hopes and realistic aspirations for Digg Reader. Here’s a lightly edited record of that conversation:
Liz Gannes: What should we expect from the Digg Reader launch?
Andrew McLaughlin: What we’re launching is a genuine “beta” in the old sense, which is to say it’s rough around the edges. There’s a whole bunch of things we know we’ve got to build, but we’re launching it now because we want to beat the Google Reader deadline with basic functionality. It’s designed to be uncluttered, clean and speedy, but it’s definitely missing features like search. None of the feed readers have search, and that’s because doing it without Google’s infrastructure is painful and expensive.
It’s funny reporting this story, because I’m completely biased. I use Google Reader, and I use the Google Reader search feature all the time.
Andrew McLaughlin: Very few users happen to be heavy users of search, but I am one of them, too. I use it all the time to dig up stuff that I thought I heard somewhere or read randomly. So yeah, I’m with you on that. I regard it as a must-have.
One thing that I don’t think we have a definitive answer on is how many people used Google Reader. How much is not a big enough audience for Google, but a potentially interesting enough audience for you guys?
McLaughlin: It’s a great question, and I’ve been trying to get one of my old Google friends to tell me.
Jake Levine: Our best bet is that, at the peak, we’re talking about 50 million users, with the low 10s of millions on an active basis.
McLaughlin: And that’s the peak — I think it’s atrophied now to be in the mid-single-digits. But as we’ve said, we’re interested in making a premium product. Because it’s not an advertising play, it doesn’t need hundreds of millions of users, and it’s part of a suite of products.
Levine: This is a crucial point. If we were just building an RSS reader, that would probably be an uninteresting thing with a crappy business model around it. What we’re trying to do is build something that has two parts to it. The overall Digg thing is trying to boil down this noisy, cacophonous Internet into smaller doses. The Digg homepage is a handful of editors looking at everything and all the signals that are coming out on the Internet, and applying editorial judgment to picking 60 to 70 items on the homepage per day. So we’re taking that theory and design aesthetic and applying it to the universe of content [for Digg Reader]. We’re trying to pull off the same effect using algorithms.
The one side is the straightforward consumption experience — you can read, view, watch, listen to the things that are coming to you — and to do so on a mobile device as well as the Web, and to do so speedily.
The second part is to use Betaworks’ data science expertise to experiment endlessly and productize different ways to rank and sort and filter those feeds so, for example, you can see the most popular things in the world right now, the most popular things in New York right now, the most popular things in your social circle. The version we’re launching now has a bunch of social and other data signals, and we apply them to the last 1,000 things in your feeds, and then we rank it so you can just let the most important stuff come up to the top. And then there are 10 to 20 ways to sort to make it comprehensible.
Do you think that social news reading is interesting? Because I’ve felt that a lot of the stuff that shows what people are reading has been really creepy.
McLaughlin: I hated that, I thought that was horrendous. Do you remember the Washington Post Social Reader? That I thought was a catastrophe. Because what it did was provoke this huge creepiness feeling, and literally make people anxious to go to the Washington Post. I can’t imagine anything worse.
I do think there’s one social action that’s an obvious one, which is sharing with an established group. Now that Instapaper is in Betaworks, we’ve been trying to look closely to see some overlap, and it might be an open reading list that you can share.
Levine: I think the active share never really bothers anyone, because you know where it’s going. The passive stuff never works at the individual level, but the aggregate data around what are people talking about on the Internet is valuable, and can be applied in an anonymized way.
Speaking of Instapaper, what do you think is the special sauce or DNA you guys will be able to add to this effort to create a reader experience? There are already other people calling themselves Google Reader alternatives, who I think have a better claim to that title, given that they’ve been working on it for a while. Not to say you can’t catch up, but you’re just starting.
McLaughlin: What we think we’re bringing is a design approach, an expertise with data — big data, real-time, and with social in particular. That’s what Betaworks is good at, and it’s at least what we flatter ourselves with what our core competency is. We’ve also got a good general product development operation here.
The trick for us is to create a package that’s compelling enough, useful enough, that people will actually want to pay money for our service.
When are you going to start charging people money, and for what?
McLaughlin: Later in the year. Our plan is to do a freemium product, and it will be things that go beyond what we’re launching.
We haven’t decided on any particular features, like search, but once we get this working, we’re going to get to work on some premium features, some of which will be enterprise workflow things, for hardcore readers who want to turn their reading into writing.
How fast are you guys going to be at launch?
McLaughlin: It should be really speedy. You’ll have keyboard shortcuts. One of the things we’ve optimized for is making it as fast as possible to zip up and down your feeds.
But will you be pulling content from the feeds fast, as well?
McLaughlin: We’re not going to be able to crawl every feed every minute, but we’re using math and algorithms to try to understand which feeds post how often, and then crawl them at the right time interval. And that will get better as we get more data.
It’s interesting to me that this is still a hard problem.
McLaughlin: This is a solved problem — we actually found a bunch of great academic papers that laid out good formulas. This is an area of optimization, not an area of innovation.
Yeah, hopefully it’s not some insurmountable advantage that only a Google could do, since they’re not going to do it anymore.
McLaughlin: I think it’s a little too early for us to declare victory on that. We don’t have Google’s crawling infrastructure, we just don’t. We’re trying to build our own approximation with five engineers and a relatively tiny budget and no data center. We got a lotta heart.
What’s your prediction about how this might play out? There’s a lot of attention on what you’re doing now; there will probably be some letdown because you can’t do everything at launch. But is this an important enough product that people will stick with you? I was reading a couple of profiles of you guys, and they’re written with the gravitas that you give to a product that everyone’s really anticipating, but at the same time, it’s about a product that somebody’s canceling because nobody cares about it.
McLaughlin: Those pieces just make me unbelievably anxious. I grew up Catholic, expecting that any good thing that happens has to be met with a deeply humiliating and horrific failure, so I’m sure that that’s true. Some people will be disappointed because what we’re launching is a beta, and they will want a fully featured version one, and it’s just not there. But I keep trying to tell people: We’re not just trying to build an RSS reader. We’re not trying to build a better Google Reader. We’re trying to build something that, over the next five years, millions of people will want to use, because it’s a great way to organize their incoming floods of stuff on the Internet.
That’s just a growing problem. It’s not just that more people are getting online and have smartphones, it’s that more people are signing up for Twitter, it’s that more people want to find ways to read blogs and newspapers and more, it’s more, more, more. In that context, a really useful tool for getting less, less, less, but much more valuable, there’s just no shortage of demand for that.
For us, we’re not aiming to get more users than Feedly this year, that’s not our goal. Our goal is to get some users that help us get to scale, and that like what we’re doing, and find it useful, and then keep building on that foundation — not toward the goal of a better RSS reader, but toward the goal of a comprehensive and very useful tool for managing the floods of content that pervade the Internet. That’s the idea, anyway.