Nieman Foundation at Harvard
Postcards and laundromat visits: The Texas Tribune audience team experiments with IRL distribution
ABOUT                    SUBSCRIBE
Sept. 26, 2017, 12:26 p.m.

Newsonomics: Tony Haile wants to build the TSA Pre✓ for how we consume news

His new startup Scroll aims to target readers who are engaged but not willing to sign up for a dozen digital subscriptions across their favorite sites. “Publishers have to make more money from this than they would have from advertising. Which, thankfully, is increasingly easy to do.”

Tony Haile learned a lot of things about news during his seven years building Chartbeat, the analytics platform used in newsrooms worldwide. One of them: “Attempts to get this industry to work together have been slow at best.” Amen to that, one of the biggest hurdles to innovation over the last two decades.

Talk to people in the news industry about what they think of his new startup Scroll, and they hesitate. They may stumble describing its model. They’ll say it’s something they’re watching. And then they’ll tell you if Tony Haile is behind it, they expect to see something impressive. News Corp, The New York Times, and Axel Springer have all made small investment bets on Scroll, part of its $3 million seed round that now supports a staff of seven getting ready for its beta launch. (It’s one of a number of new attempts to build revenue beyond standard subscriptions, described in depth in this companion piece.)

Will Scroll, which will launch next year, succeed? Haile is the first to tell you it’s a crapshoot: “We’ll see. It’s either going to be a massive success or a massive failure.”

The publishing world certainly knows it needs a new shot of revenue, and hopes Scroll might be on to something with its monthly ad-free $5 subscription to “premium news” across many different news sites. But under the surface of Scroll resides a set of ideas about news publishing in the late 2010s. Whether Scroll ultimately succeeds or not, Haile is tackling vital questions about publishers and their readers, about experience and payment. Chief among them: Many top news organizations have done a decent job getting their most dedicated readers to pay for a digital subscription. But how can that next tier of customer — people who read and value your work, but whose news consumption goes across many different sites and who are unwilling to pay for a dozen subscriptions — be made to pay too?

He and I had a couple conversations recently about Scroll and how it fits into today’s news landscape. Here’s a lightly edited transcript.

Ken Doctor: It’s been frustrating to publishers that they can’t get beyond a couple of percent of their digital audience paying for a digital subscription.

Tony Haile: My challenge is how far beyond that you can then go. I mean, with The New York Times, you’ve got about 1.8 percent of their audience are digital-only subscribers. You’ve only got another 3 percent that even see the paywall. This is my challenge — being able to grow that access space of reader revenue, when you have a predominantly casual fan model.

This is the good thing about having all that Chartbeat data as well. I know that, for every single site that I can think of, the majority come once a week and read one story. It’s that kind of Pareto curve. That means that you can get a certain amount of money out of that top percentage, but I’m not sure how much that translates down and how far you can grow it. Maybe you can grow it to 5 percent of your audience.

So can I build a kind of high-information, low-friction product before you have to become a super fan? That’s kind of how I’m thinking about it.

Doctor: Who’s in this market?

Haile: We’ve done a fair amount of pricing studies. We know for example, that in general, if you make less than $39,000, you’re not going to buy this service. If you are older than 65, you’re also unlikely to buy this service. However, if you’re within those bounds, then we have a reasonable shot of converting you, because people are pretty pissed with the state of the open web right now. The ability to kind of make that better for people and support journalism at the same time is kind of cool.

Doctor: All-access passes across many news publications haven’t worked, because the economics just don’t work for individual publishers. If you can get a couple of hundred of dollars a year from someone buying a subscription to just your single product, why throw in with a bunch of others and just take a little out of the pot?

But Netflix came out of nowhere, essentially, and established a new habit of 50 million people in the U.S. subscribing.

Haile: Here’s actually the interesting thing. I wrote a piece in The Information that was a response to [The Information CEO] Jessica Lessin maybe a month or two ago, on why we haven’t seen an all-access pass for news, and why Amazon Prime hasn’t just said like, “We’re going to add 10 bucks to Prime and then you can have Video Prime, The Atlantic, and you just get past every paywall.” I just ran through the numbers of why the economics don’t work.

Doctor: They don’t work.

Haile: They don’t at all. You have to have generally an order-of-magnitude-larger potential audience than standing on your own, because of the metered way in which we do content and news.

It’s not binary, as it is in cable. If I’m a casual fan and I want the USA Network, then the only way I get that is by signing up for cable. Even if I only want to watch it for like one episode a month, the only way can get access to any kind of content is signing up for the bundle.

The same thing with newspapers. It was binary: You either had it or you didn’t. By having metered models, casual fans get zero value from any kind of all-access pass — because they never even see a paywall in this context.

The audience that is viable for a bundle or an all-access pass is the sliver of the New York Times audience who is also hitting the paywall on The Washington Post and also hitting the paywall on The Atlantic.

It’s a far smaller percentage of audiences who actually find value. That’s why none of these things have really worked. And the things that people have even tried have been built as separate apps. None of those things ever get to scale because they sit outside discovery paths. Yeah, that combination of the economics of bundling plus the way in which we discover content and meter content makes any kind of all-access pass basically nonviable, at least to scale.

Understanding the information paradox

Doctor: Right. Then the models that are kind of in between or run-ups to full subscription, like LaterPay.

Haile: Yeah. You’ve also got these things that are micropayments, or versions of micropayments.

But the big sites want people to get frustrated and pay and then forget about it, versus them thinking like, “Oh, I can give them 10 cents every time I go over the limit.” It’s like a gym membership.

If you’re a publication, it makes no economic sense for you to do that. If you try to do it in the Blendle pathway, which is again separated into an app, then you still have these problems.

You have an information paradox: the value of the information before you read it is unquantifiable, and the value of information after you’ve read it is zero, because you’ve already read it.

Then just the cognitive friction of it all, so you’re just replacing one form of cognitive friction with another, and that doesn’t really tend to work. That’s why I love Alex [Klöpping, CEO of Blendle]. I think he’s a great, very smart kid, but Blendle is struggling, even in its home market.

[Note: Klopping disputes that it’s struggling in the Netherlands. “We beat our own quarterly revenue record last quarter, and did the same the quarter before,” he told me last week. “We’re at 1.5 million registered accounts in the Netherlands (the Netherlands has 16 million inhabitants). Fifteen percent of those users pay with micropayments. We have 34 publishing partners in the Netherlands that together have 140 titles on the platform.”]

Doctor: So tell me more about what you’re up to.

Haile: Okay, we know that direct consumer revenue is going to be way more attractive for publishing than indirect consumer revenue, which is advertising. I also know that subscriptions do really well with that…let’s just be optimistic and say that top 5 percent, but not really beyond there. Companies start to compete with each other after a while. If I have a subscription to The New York Times, how many other ones do I get? So you have a kind of rich-get-richer scenario.

Then, the interesting thing for me was to look at ad-blocking not so much as a problem with publishing, but as more of a consumer signal. In TV, that was what you had — you had increasing ad loads, which led to the consumer signal that was TiVo, and then led to the consumer signal that is Netflix. Music had similar kinds of frustrations, leading to SiriusXM, torrents, LimeWire, and that stuff, then leading to Spotify.

In media, you’ve had increasing ad load, increasing frustration, consumer signals, and ad-blocking leading to…what? So, the question for me was: Is there an orthogonal way to get direct consumer revenue? Not from access, but from experience.

Is there a group of people who will pay in general for an ad-free experience across X number of sites, or whatever? I kind of started with that, and there’s a few problems that you have to try and solve for. Because there’s a whole ton of corpses in this particular graveyard.

Doctor: You’re making a major point here — Scroll isn’t some kind of pass across a lot of sites, as others have tried to do. It’s not about access, or all-access.

Haile: Not access at all. There is no way that you can do a bundle that matches the economics that a publisher makes from doing single-site access subscriptions.

Taking (some) of the hassle out of news reading

Doctor: So what are you offering them then?

Haile: My analogy is this: Think of it like an airport. I’m thinking about creating TSA Pre✓ for the web. Have you ever used TSA Pre✓?

Doctor: Yep, too frequently.

Haile: Remember that first time you went through it, and you didn’t have to take off your belt or your shoes or take your laptop out of your bag. TSA Pre✓ doesn’t get you into business class. That’s your New York Times subscription. It doesn’t get you past the paywall. But it makes the experience better, so it’s less frustrating.

You don’t have to X out the ad, you don’t have to wait for the preroll, whatever. That was what I was interested in. In general, when people have looked at this in the past, they’ve said, “I’m going to create this amazing app. It’s going to be a beautiful app, it’s going to be kind of Flipboard-esque — and all you have to do to use it is change everything about how you discover content.”

Doctor: As you explain it, it makes good sense. But is it too complex?

Haile: We think of these things as quadrants. There are Scroll users, and there are New York Times subscribers. That gives us four blocks.

There is someone who isn’t a Scroll user and isn’t a Times subscriber. They get ads on 10 articles.

There is someone who is not a Scroll subscriber, but is a Times subscriber — they get ads on unlimited articles.

There is someone who is a Scroll user, but not a Times subscriber — so they get an ad-free experience on, but then after 10 articles, they hit the paywall.

Then there is someone who is a Scroll user and a New York Times subscriber, and they get unlimited articles, ad-free. Those are the four quadrants that you can possibly be.

There are two things going on there: One is access, so you get more than 10 articles. The other one is experience.

The other key thing when you’re looking at this model is the thing that got people in trouble in the past has been this desire to try and merge the two. They’ve tried to do access and experience. When you try to do that, when you try to merge The New York Times’ subscription revenues as well as their advertising revenues, that’s where the economics start to really break down. We’re avoiding all of that. The New York Times, the subscription basically exists almost in an alternate universe to our ad-free experience in that context. Does that all make sense?

Doctor: I understand it abstractly. And god knows the actual experience on too many news sites ranges from boring to insulting as ads take over pages. But as consumers, I don’t know that we think about it in discrete categories of access and experience.

Haile: This is actually the reason why we decided not to create a white-label service. We could have just started a white-label service and get the publishers to do this for themselves. The reason why we decided not to was, by having a brand in Scroll, you’re able to differentiate between the Scroll thing — which is this thing where I go around a bunch of sites and it makes my life better — and The New York Times, which is a distinct brand, which is like: “Oh, I want to read more stories.”

By doing that — having the two distinct brands — we can kind of connect messaging for those two different sets. It’s like — I like to use airline analogies — no one confuses TSA Pre✓ with business class.

You want to try and deliver an experience. The experience has to be where the consumers are, because if you’re going to ask any of them to pay, that’s like the one hard thing you can ask them, and you can’t ask any other hard things.

Then we had to look at how you do it. So no matter where they’re coming from, whether they’re coming from Facebook, email newsletters, blah blah blah, they get beautiful, fast speed. That means integrating into the sites.

Then the other side of it: How the hell do you make the money work? One problem is the actual general price itself, where at the bottom end, you’ve got to be able to distribute enough cash, at an industry level which can beat the opportunity cost of ad revenue. Publishers have to make more money from this than they would have from advertising. Which, thankfully, is increasingly easy to do.

Doctor: And you’ve cracked that nut, that arithmetic?

Haile: I looked at DCN [Digital Content Next] publishers is the kind of good proxy, yeah? The top 80 publishers, they make $7 billion in digital ad revenue across 234 million uniques. That gives me a monthly ARPU of about $2.50 a month.

Which means that if I wanted to do a 70/30 split, my minimum viable price is like $3.60 or so. That’s my minimum, and then I had to look at what consumers were willing to pay. Did a bunch of stuff around that.

Generally, the optimal price point of that acquisition is around $5. So that’s good, because it gives me a $3.50 pool — gives me like a 40 percent lift across that. [That’s the $3.50 in reader revenue compared to the $2.50 in ad revenue.]

So one of the fun things that we’ve been trying to work out — and have worked out, thank god — is basically how the hell we can say to consumers, “Here’s one flat $5 payment a month. In return for that, across the top X sites on the web, you never have to get annoyed ever again.”

What I’m trying to think about in some ways is that kind of middle of the funnel.

Doctor: Explain where you fit into that audience acquisition funnel. Publishers talk a lot about the top end, getting news samplers to their content, via Facebook and Google. And they talk about the narrow funnel at the bottom, collecting good money for a subscription.

Haile: We currently have the top of the funnel which says: “In order for us to make as much money as possible, we need to interrupt your experience as much as possible.” Especially now that we’re in a mobile world. It’s prerolls; it’s that kind of shit. The bottom of the funnel says: “We need you to be as engaged as possible so you have a strong brand relationship and we can convert you to the high value of direct consumer revenue.”

Those two things do not work in concert with each other. So the thing that I’m interested in is: “Can I take, at the network level, a kind of a broad group of people and convert them?” So, like, we get it: “You don’t know where you’re going to be reading next week, because you get your news through Facebook, you get your news through Twitter. You don’t even necessarily know the site that you’re on right now. But you know the experience that you want when you get there.”

You don’t want to have to click away popups, you don’t want to have to deal with Outbrain or Taboola shit.

For $5 a month, Scroll can offer a more engaging experience on the site. It means the users are more engaged, especially those further down the funnel. And then, because I already have registration information and payment credentials, at that point where they hit a paywall for whomever, The New York Times, the Star Tribune, The Washington Post, it can then be one click to get them through and get them signed up.

These are the things that I’m kind of interested in and playing around with right now. What’s the group of people who will pay for experience in the way that people will pay for premium Spotify, pay for ad-free Pandora and Twitch and Hulu and YouTube Red and so forth? Is there a viable network to be created there?

How does that one make more money on a per-user basis for each publisher, so that they know they’re not losing any money, and how does it help improve that kind of subscription down to like the access part?

Beyond the super fans

Doctor: So, the target customer, you said, was middle-of-the-funnel? As you said, at the top, you can maybe now get 2 percent — maybe 5 at some more distant point — of people to pay if you just ask them, essentially. Is this the next 2 percent, the next 5 percent? What percentage do you think you can get to pay something?

Haile: Well, the good thing for us is we don’t have to make anyone a super fan of a particular site. That’s the challenge with the 2 percent.

We can be casual fans across a network. That means there’s a much, much broader audience as we go. If we were to apply standard kind of premium models, which is like 10, 15 percent of an audience converts — when there’s 10, 15 percent of a network audience, you can get quite large. So that’s how I think about market size. The vast majority of people are still going to be ad-supported, still going to be going down the free options. But can you get 15, 20 percent of people to pay for ad-free?

To give you the Spotify example, about 50 percent of the people who convert to pay for Spotify do so just to get rid of ads. What, they’ve got like 40, 50 million people now? [Spotify reached 60 million subscribers in July, up 10 million from March.]

The whole point here has been get beyond the super fans and be able to find some way to directly monetize casual fans. Because there are a lot more casual fans than there are super fans.

Doctor: Have you put a number on the casual fans?

Haile: We’ve looked at various things. Say you can look at the 236 million [using desktop ad blockers in North America and Europe] and say, “Let’s take 20 percent of those as a market size.” Or you could be even more bold and you could say you could look at all content consumers and say you want to get 15, 20 percent of those. But to be honest, we’re trying to be fairly cautious in the first year or two. We’re looking at growth up to 2 million or so in terms of consumers.

Doctor: If I step back from this, the major attraction for me as a reader is that this is ad-free. That’s the consumer pitch. You will still hit the paywall if there is one. That may be confusing to “subscribers.”

Haile: You basically get whatever you would have gotten if you weren’t a Scroll user. So, if you’re not a New York Times subscriber, you get 10 ad-free articles. If you are a New York Times subscriber, then you get as many ad-free articles as you want. But it’s completely orthogonal to access. We know that we can’t handle the economics of access. That’s an entirely different thing and it should be handled at the single-site level.

On ad-free, the interesting thing for me, once we go beyond that, is what can you build when you no longer have to make a real-time call to the ad server to show someone something. It makes audio versions of articles super easy. It makes offline really easy. It does all kinds of fun things that you can do.

But to begin with, the very first thing it should do is remove pain. Right now, you can do that by removing just the friction of people’s lives.

I don’t actually think people have a problem with ads. They have a problem with friction. If we can get the friction out of the way so that someone, when they go to the web, it’s fast, it’s beautiful, they’re not distracted, they just get to read the things they wanted to read — if we can do that while also making more money for publishers, then that’s kind of a web worth fighting for, I think.

Doctor: Another inevitable question in this time of outsized platform power. How does Scroll relate to Facebook?

Haile: We work within Facebook. We have to be able do that; that’s where most consumption happens. It’s interesting for us: It’s where most consumption happens, but no ad-blockers work there. It’s the reason why mobile ad-blocking is so low. It’s not because people don’t want to block ads on mobile. The problem is that because it’s happening in apps, their ad blocker doesn’t work. Scroll does because we’re in partnership with the site. So now, when people are in Facebook and they go to a story, it can be clean, fast and beautiful. That’s something that really helps us.

Doctor: So this is a fully licensed model, and you’re getting format contracts from publishers. What kind of integration do publishers need to do with Scroll?

Haile: Basically, the way it works is this. When someone who’s signed up for Scroll comes to a site with a Scroll cookie — let’s say they hit The Washington Post — the Washington Post CMS can check for our cookie. Once they check for our cookie and don’t find it, they then say: “Okay. Not a Scroll user.” That’s fine — it loads exactly as normal.”

But if they are a Scroll user, it says: Don’t load any of the ads. Don’t load any of the Outbrain or Taboola. Then what we do is we track the behavior of that user on the page, and that helps us to distribute the revenue to The Washington Post, in this context, that would make them more money than they would have made from advertisements.

Doctor: Any prerolls in video?

Haile: No prerolls. One of the most annoying ads.

Doctor: If a publisher gets very few or “too many” Scroll views, how do you handle that?

Haile: We’re effectively, in some ways, acting as … almost like a programmatic network would in that when a user comes to the page, that user is identified, and then we deliver that money. So even if we only have one person signing up for Scroll, the publisher should make more money than they would have from advertising.

Doctor: You’re saying each person basically proves out the model. So, you’ve got your one user, they’re paying five dollars a month, right?

Haile: Yeah. My cofounder [Sachin Doshi] was the guy who’d run the business model of Spotify for the last six years. He ran all the content licensing, that kind of thing. He’s the guy that kind of owns the model with us, thank god.

So the way that we’re doing it is that there’s a 70 percent split. Fifty percent of the gross is done on share of attention. So the more engaging your content, the more money you make. I didn’t want to do any kind of model that would prioritize or encourage clickbait or any kind of crap. By doing it on a kind of share of time spent, you have, generally, the kind of more premium content catching more engagement — and also video is more expensive on the ad side.

Doctor: What’s your timeline this fall?

Haile: We’ve been closing deals at a rate of about one a week since mid-August. We’re moving into long-form contracts, going back to the people who’ve said they’re good with the short-term. We’re looking at the next two or three months’ time for testing the experience. This is a slow process. Daniel Ek took three years to get the publishers together for Spotify. I’m hoping like hell that I don’t take that long with this, because my wife will kill me.

But yeah, I’ve got to try and persuade enough publishers that the model works and that they should be innovating in this kind of way to make it real. I’ve been super lucky thus far, and we’ve had a tremendous amount of support and help, but it’s still the big challenge. With all of these things, in general, attempts to get this industry to work together have been slow at best.

Doctor: We haven’t talked about membership, that voluntary public radio–like payment that’s much in use at nonprofits and, of course, big time with The Guardian.

Haile: That’s a wonderfully British and passive-aggressive campaign.

Doctor: Yeah, exactly. That’s a good way to describe it, yeah.

Haile: It’s only fair. “Give us your money.” Yeah, it’s brilliant. It’s just the proper thing to do, you know? It’s sort of: “We’re not giving you a reason, we’re not giving you the hard sell.” You know the American way of saying it would be like: “Here are all the amazing things you could get if you just become a Guardian member.” No. “We’re going to give you absolutely nothing, apart from a feeling of the general well-being” — that is very English as an approach.

Yeah, I’m kind of fascinated by it, though, because it does show something, which is like basically with that top 2 percent or so, if you just give them a reason to give you money — or an opportunity even, not even a reason, an opportunity to give them money — you can probably get money from them.”

Doctor: Getting beyond that 2 percent is the name of the game, isn’t it? Is it too dramatic to say this is a do-or-die time for much of the press?

Haile: I’m trying to save the web in some way. We’ll see! It’s either going to be a massive success or a massive failure.

Doctor: Why be a medium-sized struggling little thing, right? You don’t want to do that.

Haile: You don’t want to be medium, no.

Photo of security line at Atlanta’s airport by Josh Hallett used under a Creative Commons license.

POSTED     Sept. 26, 2017, 12:26 p.m.
Show tags
Join the 60,000 who get the freshest future-of-journalism news in our daily email.
Postcards and laundromat visits: The Texas Tribune audience team experiments with IRL distribution
As social platforms falter for news, a number of nonprofit outlets are rethinking distribution for impact and in-person engagement.
Radio Ambulante launches its own record label as a home for its podcast’s original music
“So much of podcast music is background, feels like filler sometimes, but with our composers, it never is.”
How uncritical news coverage feeds the AI hype machine
“The coverage tends to be led by industry sources and often takes claims about what the technology can and can’t do, and might be able to do in the future, at face value in ways that contribute to the hype cycle.”