free html hit counter Future of Search Archives - John Battelle's Search Blog

What Will Search Look Like In Mobile? A Visit With Jack

By - December 18, 2014

I’ve come across any number of interesting startups in my ongoing grok of the mobile world (related posts: 1, 2, 3).  And the pace has quickened as founders have begun to reach out to me to share their work. As you might expect, there’s a large group of folks building ambitious stuff – services that assume the current hegemony in mobile won’t stand for much longer. These I find fascinating – and worthy of deeper dives.

First up is Jack Mobile, a stealthy search startup founded a year or so ago by Charles Jolley, previously at Facebook and Apple, and Mike Hanson, a senior engineer at Mozilla and Cisco who early in his career wrote version 1.0 of the Sherlock search app for Apple. Jack was funded early this year by Greylock, where Mike was an EIR.

I’d link to something about Jack – but there’s pretty much nothing save a single page asking “What Is Jack?” Now that Charles and Mike have given me a peek into what Jack is in fact all about, I can report that it’s fascinating stuff, and at its heart is the problem of search in a post web world, followed quite directly by the problem of search’s UI overall. Whn you break free from the assumptions of sitting at a desk in front of a PC, what might search look like? What is search when your device is a phone, or a watch, or embedded in your clothing or the air around you?

Jack is trying to answer that question, and the team is rethinking some core user interface assumptions along the way.

Search on mobile is by almost any measure broken – the core assumptions of what makes search work on the web are absent on your device. On your phone, there are no links to index, no publicly accessible commons of web pages to crawl and analyze. Just a phalanx of isolated chiclets – disconnected apps, each focused on a particular service. But that doesn’t mean we don’t need to search in mobile, in fact, we search a lot on our phones. But the results we get ain’t that great. In the main, that’s because when we search on our phone, we get answers from…the web. But as Jolley and Hanson pointed out, answers from the web often fail when presented to us in the context of mobile.

BattelleMediaMobilevWeb

Mobile search queries are just…different. Let’s explore why:

- Context. When you search on your phone (or later, on any “liberated” device), you’re more likely than not in completely different context from when you’re “on the web.” Mobile searches tend to be service related – “How do I get to this address,” and/or location driven: “What are good nightspots nearby.”

- Query & Corpus. Because of this context, *what* we want to search is focused in a far smaller potential corpus of material. Mobile searches tend to have one exact answer – we aren’t loking for a list of links that we then want to peruse, we want an answer to a specific contextual question – mobile searches bias toward service and action as the query result. That means search’s presumptive barrier of completeness (the cost Google bears of keeping the entire Internet in RAM, for example) is not a barrier on mobile. You don’t have to have ALL the possible information “indexed” – just the right information.  And what information is that? Well, that leads us to ….

- Signal. With mobile, rich new signals are available that could (and should) inform search results (but don’t). Certainly the most robust such signal is your current location, but that’s just the start. Others include your location history (where you’ve been), the apps loaded on your phone, your usage history with those apps,  and the structure inherent in those apps to begin with. Which begs a huge possible difference in…

- User Interface. Search on mobile, for now, is identical to search on the web. It’s a command line interface, where you type in your query, and you get blue links for results. Google’s been working hard to address this, and the combination of its universal search product, which surfaces “one true answer”, with voice search, is a real step forward. But the folks at Jack showed me another potential search interface for mobile, and I found it quite compelling. That approach? Well, I’d call it “conversational.”

The Conversational Search Interface

Way back in 2004 I met with Gary Flake, then a senior technology executive at Overture, a leading search firm of the day (Yahoo! later acquired Overture, which fueled Yahoo!’s search results until the Microsoft deal in 2009.) Even way back then, before mobile was a thing, I was frustrated with search’s interface.

I asked him why we couldn’t move forward in search interface – the “ten blue links” approach was so … flat. I wanted to ask one question, get results, and then ask another. Or better yet, I wanted the service to ask me a question – “You entered ‘Jaguars’ – did you mean the football team, the car, the cats, or something else?” Gary looked at me ruefully and said something I’ve never forgotten: “If only I had just one modal dialog box…”

What he meant was that search, at that point, was a race for the best ten blue links, and anything that got in the way of that, like a modal dialog box that popped up and asked a refining question, would mean that a very large percentage of folks would abandon the search. And abandonment of the search meant loss of revenue.

But that idea – of search as a series of back and forth exchanges, a conversation if you will – has always stuck with me. So imagine my surprise when Jolley and Hanson showed me a very early prototype of Jack Mobile’s search interface and it looked like….a conversation!

Jolley and Hanson have asked me to not report the details of their prototype interface, but suffice to say, it’s quite different, and it looks and feels far more like a back and forth than anything on the web. It’s delightful, and using the service is also cool. Jack knows where you are, so if you ask it “Guardians of the Galaxy” it’ll find showtimes near where you are, and return that information as the result. If you ask it “Italian restaurants,” Jack won’t give you a list of places with the most Google+ reviews – it’ll give you highly rated eateries near where you are right now, perhaps enhanced by reviews relevant to you given the fact that you have the GrubHub or OpenTable app on your phone.

Takeaways

Jack is still in very early stages, but the co-founders had a number of key insights from their work so far. The first has to do with completeness – while long tail edge cases rule the roost in web search, mobile has far more distribution of “head end” search – which means you can narrow your indexing and your algorithms and still satisfy a large majority of queries.

Also, mobile search is deeply personal – there’s almost no such thing as one size fits all result. In mobile, results should be rewarded because they are the most likely answer, not because a ranking system has pre-determined them as most authoritative. Searching for “BMW 3 series” while standing at a Mercedes dealership should most certainly bring a different result than the same search from a Taco Bell on Interstate 5. While personalized search has become a mainstream attribute of Google, the truth is, it’s quite shallow – on the web, Google knows precious little about you. But your phone knows quite a bit. Unlocking all that data is still too hard, but it’s coming.

But perhaps the most interesting implication of Jack’s approach to search lies in how it might drive a new ecosystem between “publishers” and “audience members.” Web search, Hanson points out, is all about the consumer – the creator of the web page is a second-class citizen, stuck in a suckers’ deal of sorts: You have to “publish” your presence on the web, or risk irrelevance, but you are entirely at the mercy of black box forces you don’t understand when it comes to how people might find you. Hanson posits a different model for Jack’s index, one in which publishers deliver their app and content structures to Jack via a proprietary feed of discrete, small units tagged to specific query types.  If this sounds a bit like semantic search, well it is. Hanson, a veteran of open web standards fights via his work at Mozilla, told me he has “deep scar tissue” around the topic, but at the same time, he and Jolley sense that with mobile, a new kind of level playing field just might allow semantic, personalized search to truly emerge.

There are far more questions than answers hanging over Jack, but that’s why it’s interesting – here’s a small, well funded team of search, web, and mobile experts really leaning into a new way to think about a massive problem/opportunity set. It’s certainly one to watch in 2015.

  • Content Marquee

Google: The Information-First Conglomerate

By - November 21, 2014
FortunePageCover

Larry Page on the cover of Fortune, Nov. 13 2014

Last week Google CEO Larry Page got the Fortune magazine cover treatment, the latest of many such pieces attempting to quantify Google’ sprawling business. The business press is obsessed with answering the question of whether we’ve reached “Peak Google.” (Clearly Fortune’s opinion is that we have not, given they named him “Businessperson of the Year.”)

“Peak Google” is what I like to call a “contagious misconception” – it seems to make sense, and therefore is worthy of consideration. After all, we’ve seen IBM, Microsoft, and other companies hit their peaks, only to drop back as they face the innovator’s dilemma.  Search is past its prime, Google is a search company, ergo – Peak Google.

But as the Fortune piece argues (and yes, I’m quoted, for what that’s worth), Google has a lot more going on beyond search. And while it continues to milk that multi billion-dollar quarterly profit center, it’s built five additional billion-dollar businesses – some of which are directly related to its search empire, but others that are not. Google Apps/Cloud, YouTube/Play, Android, Ventures, and Adtech are already past the billion-dollar mark. Huge businesses in waiting include plays in home automation (Nest), healthcare (Calico), transportation (Chauffeur/self driving cars), and connectivity (Fiber). Beyond that group lie a dozen or so potential blockbusters in energy, robotics, AI, wearables, and the unknown moonshots behind the curtains at GoogleX.

It’s that stunning breadth of scope – what Fortune calls the company’s seemingly limitless ambition – that has caused a prolonged internal debate around Google mission statement:

“To organize the world’s information and make it universally accessible and useful.”

Page has been floating trial balloons about expanding Google’s mission statement for nearly two years. When Tony Faddell, CEO of Nest, announced Google’s acquisition to his staff in January of 2013, Page took the stage and took questions from the stunned audience. One staffer asked Page why Google had any interest in a home automation company – it seemed quite orthogonal to Google’s focus on search, apps, and mobile. According to sources at the event, Page answered by acknowledging that Google’s mission statement may not be large enough to contain his company’s ambitions.

Since that first admission, Page has been testing out the idea of an expanded mission, and with Fortune he aired his ambivalence in public, telling Miguel Helft that “it’s probably a bit too narrow.” And on first blush, that seems right – what does a thermostat have to do with organizing the world’s information, anyway?

Actually, quite a lot.

When you look at Google through the lens of what I call “information first” businesses, things start to make a lot more sense. By that measure, Google is not only an information-first company, it’s also the world’s first information-first conglomerate – starting or buying businesses in every market undergoing the transition from “matter first” to “information-first.”

We see the transportation business shifting to information first, for example. The currently maligned but nevertheless extraordinary Uber is proof of it, but so is Zip Car, Tesla, and the entire autonomous car industry. The true value of these new kind of businesses is in how they understand information flows in the transportation markets, then execute new approaches to old problems (how do I get from here to there?) using novel and/or more efficient methods based on information technologies. Uber doesn’t put cars (commodities) or drivers (means of production) first – it puts information processing first. The cars and driver then reorganize to the new information flows and – voila! – a $17 billion company is born in four years. Uber proves that if you solve difficult information processing problems in traditional markets, you can create world beating value. Airbnb, DocuSign, Lending Club, and many more are further examples of the same thesis.

So what markets are ripe for transition to an information first framework? Well, let’s break down what makes for a “ripe” market. I think there are two key attributes of a market ready to be radically shifted by an information-first approach. First, a market where there’s liquidity of poorly organized and processed information. In other words, there’s a ton of data, but it’s not well organized or computed. Think about the world wide web in 1998, for example. Sh*t tons of information, terribly organized and lacking a processing layer. Google came in and – voila – a multi billion dollar company was born in five short years. Secondly, look for a market currently controlled through centralized chokepoints, but with the potential to be rapidly reorganized if and when consumers gain control. Again, look at search – before Google, portals like AOL and Yahoo ruled the web. Everyone went to a chokepoint to “see what was on the Internet.” After Google, consumers took control of their own web surfing.

So…what markets have both data liquidity and are currently controlled by centralized chokepoints? Well, let’s look at mobile. Tons of data, terribly organized, controlled by the chokepoints of carriers and OS vendors. Check! Or, how about healthcare? Oh hellz yeah! Energy? Yep! Connectivity? Most certainly! Markets where there’s not yet liquidity of information, but there’s about to be – home automation, food, retail – are also ripe for reinvention.

The world is turning into information, and that information wants to be organized, accessible, and useful. I don’t think Google’s mission needs to change at all. Whether or not they knew it at the time, Google created a manifesto that I believe will prove to be dead on in the context of an economic shift to a information-first paradigm. And when the history of this era is written, I’d wager that Google will be seen as the first information-first conglomerate to both identify and exploit that shift.

The Web Will Kill Apps

By - November 17, 2014

wired web dead coverLots of the “apps are killing the web” meme going around these days, with the latest batch of casket sealant come from no greater validator of commonly agreed upon wisdom than the Wall St. Journal. “The Web Is Dying; Apps Are Killing It” argues Christopher Mims, and it’s hard to argue with him given the preponderance of current evidence.

I disagree.

I am in the midst of a long stew on the future of mobile, it’s taken me through deep links and intelligent links, to the future of search on mobile and beyond, and I’m nowhere near finished with either the reporting or the writing – so I can’t definitively counter the Journal’s argument – yet. But I feel it in my bones – apps, what I’ve disparagingly called “chiclets” – are not the model of how we will interact with information, services, or the world via mobile. The best of the web – open, low cost to entry, no gatekeepers, end-user driven, standards-based, universal namespace, etc. – will prevail.

Why am I so sure of this? Because just about every single person I’ve spoken to – some three dozen or so, to date – are convinced we’re in a secular shift from the app model to….something else, something new, something better. I had a great meeting today at the mobile search startup Jack, for example, with people who are super-qualified to have opinions on the matter (ex Facebook, Excite, Apple, et al, backed in a quiet $6mm round early this year by John Lilly and Reid Hoffman at Greylock). And they are not alone – the caliber of people I’ve encountered who share my point of view is extraordinary. Something big is brewing, and I’m deep into figuring out how to frame it. It’s  a big story, and I don’t know if I can tell it as well as it deserves to be told. But I’m going to try, and if you’re reading this, well, it’s your job to course correct my attempts.

Stay tuned. The web as we knew it ten years ago may be “dead,” but its core values and framework are alive, kicking, and poised to once again disrupt the current oligarchs of mobile.

“Peak Google”? Maybe, But Is “Native” The Reason?

By - October 23, 2014
google3-600x450

From Thompson’s “Peak Google” post.

I love Ben Thompson’s Stratechery site, so much in fact that I’m writing a response to his recent “Peak Google” post, even though these days most of us limit our bloggy commentary to the 140-character windows of Twitter.

I’m responding to Thompson’s post for a couple of reasons. First of all, the headline alone was enough to get me interested, and judging from the retweets, I was not alone. But I try not to retweet stuff I haven’t actually read (which, as Chartbeat has shown us, is not the case with most of us). I waited until this morning to read Ben’s post, which compares Google with IBM and Microsoft, each of which once could claim king of the mountain status in tech, but have since been eclipsed.

But the real reason I’m responding is Thompson’s thesis that Google is at its peak, about to be eclipsed by Facebook or Pinterest or some other advertising-based company. Thompson theorizes that native advertising will carry the day, and because Google lacks the skills necessary to win in native, the company will slowly fade into profitable irrelevance (as have IBM and Microsoft).

I think the story was picked up for the resonance of its headline meme – that Google may be at the peak of its power, poised for decline – rather that the substance of its argument around native advertising. And that’s a shame, because Thompson makes any number of interesting points about Google that bear debate. In no particular order:

– Google isn’t good at native or the more “human” side of advertising. I disagree. As Thompson acknowledges in an update to his original post, Google’s search ads are by definition native – I believe search advertising is the most scaled, profitable, and useful native platform in the world. And while Google has struggled to find its voice as a media company, it’s been near-perfect at creating a platform that lets others express their voice in advertising – millions of customers use AdWords to create authentic ads in real time. And its YouTube platform is poised to be one of the largest brand channels in the world.

– Search isn’t a brand platform, and native is. I also disagree – that search isn’t a brand platform, anyway. Yes, search ads tend to be “down funnel” and direct response driven. But any brand who isn’t playing in search, which includes creation of search-friendly content, is missing a huge part of the role brands must now play as creators of great content.

– No one has yet “won” in native. Totally agree. We’re in the very early innings, though Facebook has been a clear winner to date.

– Search ads will be eclipsed by native. I sort of agree – but I prefer to think that neither native nor search ads will dominate going forward. Instead, we’ll see a blending of the two – what I’ve called programmatic native. And there’s no greater example of programmatic native than search. It’s not clear Google will win here, but it doesn’t hurt to own both search AND YouTube, not to mention Android and Google Maps/Local/Earth/Now, as you work on figuring it out.

Again, I think there’s a reason that “Peak Google” resonates so strongly in this industry – we’ve seen the movie with IBM and Microsoft, and everyone loves a story that fits a common narrative. But I’m not sure native advertising is the reason Google will fade to irrelevance over time.

The Next Stage of Mobile Quickening: Links Get Intelligent

By - October 05, 2014
HowItWorks

How Branch Metrics works…click to enlarge.

Early in a conversation with Alex Austin, CEO of mobile startup Branch Metrics, I had to interrupt and ask what seemed like a really dumb question. “So, wait, Alex, you’re telling me that the essence of your company’s solution is that it….makes sure a link works?”

Alex had heard the question before. But yes, in truth, what his company specializes in is making sure that a link works in a very particular kind of mobile use case. And doing so is a lot harder than it might seem, he added. Branch Metrics, a three-year old startup that began as a way to create and share photo albums from your iPhone, is now devoted entirely to solving what should be a dead easy problem, but thanks to the way the mobile ecosystem has played out, it’s just not. (Alex has written up a great overview of his journey at Branch, worth reading here).

A month or so I wrote Early Lessons From My Mobile Deep Dive: The Quickening Is Nigh, an overview of my initial learnings as I explored today’s mobile landscape. A major conclusion: the emergence of deep linking is leading to entirely new opportunities in mobile, and the mobile marketing machine is a key place to explore if you want to understand the implications.

Since then, I’ve spent more time talking to folks like Alex, and I’ve come to another conclusion: the next step in the mobile quickening will be intelligent links.

Now, before you go Googling “intelligent links” – I’ll admit there is no clear nomenclature per se, because in the past we’ve not had a need for such a distinction. After all, on the open web, all links can be intelligent, because they can pass information from site to site via cookies, redirects, and various increasingly sophisticated hacks.

Not so in mobile.

In his wonderful post outlining Branch’s initial failures and eventual pivot, Alex notes: “The biggest growth issue we faced in our mobile app was the fact that Apple doesn’t let you track users and pass context through the install process. …To break down this barrier would mean making the mobile app ecosystem more like the functionality we’re used to on the web.”

So that’s what Branch set out to do – in essence, to make mobile work more like the web. Branch’s initial photo book product may have failed for any number of reasons, but what stood out for Alex was how hard it was for the product to self-replicate across a customer base. A customer would create a cool photo book, and then want to share it with a friend. Of course, the best way to share is via a link to the photo book – that’s the viral calling card. But when a friend clicks on the link, Branch ran into the limits of mobile apps. It gets kind of convoluted, so let me break it down in steps:

1. Customer downloads Branch and uses it to create a cool photo book.

2. Customer wants to share the photo book with her friends, which she does using Branch’s internal sharing features.

3. Branch’s sharing features generate a deep link that is sent via email (or a Tweet, or Facebook, etc).

4. Friend receives invitation via email to check out a cool photo book.

5. Friend clicks on Branch’s deep link.

6. Friend does NOT have Branch’s app installed, so is linked to the Branch app download landing page in the iTunes store.

**THIS IS FRICTION POINT #1. In an ideal world, a potential customer should not have to go through the Apple app store just to view a cool media object that’s been shared (this wouldn’t happen on the web). **

7. Friend decides to download the app, tells Apple OK, accepts the app’s terms and services, fires up the app, and….

8. Sees the generic welcome screen that the app brings up for every new user. Now he has to create a new account, set a password, etc. Confused, he wonders whatever happened to the photo book he was looking for.

**THIS IS FRICTION POINT #2. The friend just wanted to check out the cool photo book, but the information of the original URL, which pointed to the actual media object, has been lost.**

9. Friend is confused as how to actually use the Branch app to see his friend’s cool photo book. He pokes around a bit, but quickly loses interest when he sees a new notification from SnapChat, or Facebook, or whatever.

10. Friend never becomes a new customer of Branch, nor ever actually sees the photo book.

This is a deeply lame experience, and one that seriously limits any app developer’s business. “You can’t have someone have to type their password in, and go through a long install and configuration to start using the app,” Alex told me.

So Branch pivoted, and created a lightweight SDK (software development kit) that, when installed by the app maker, allows the media object in question to appear once the app is installed.

Sounds super simple, but according to Alex, it was quite complicated, not least because getting app makers to install SDKs is non-trivial. However, Branch is finding traction with scores of app makers because the company solves a major marketing problem in mobile – how to create more fluid conversion and engagement paths which ultimately lead to more customers.

This is the evolution of the intelligent mobile link – something that’s sorely needed in the mobile ecosystem. It all starts with the ability to pass data through a link – something that Apple has not allowed in the past. But Branch’s elegant hack around Apple’s shortsighted policy is one more important step toward creating a truly mobile web, one that combines the richness and device-specific capabilities of an app with the universality of an open web architecture.

“It’s like 1995″ in mobile apps, Alex concluded. “We are just figuring out how to turn on the Internet on the phone.”

When I start to think about where this goes from here, I start to get very excited – intelligent links are the beginning of a whole new mobile experience. The next step is to break down the hegemony of the app store itself – why should we have to go through an authentication, download, and configuration process just to see what’s behind a link? We shouldn’t, and soon, I imagine we won’t. Of course this has serious implications for the hegemonies of Apple and Google’s app store choke points, but in the end, both companies are all about creating great experiences for their users, right?

Take it one step beyond erasing the app store friction, and we can imagine a world where apps work like always on-call services, at the ready to execute their portion of a fluid user experience. Explaining that experience will be the subject of a future post. But for now,  amen for folks like Alex and companies like Branch Metrics. Keep up the good work.

Early Lessons From My Mobile Deep Dive: The Quickening Is Nigh

By - September 06, 2014
chiclets

Do you really want to eat them one at a time? Me, I prefer mashing ‘em up.

Recently I began a walkabout of sorts, with a goal of ameliorating my rather thin understanding of the mobile marketplace. If you read me closely, you know I’ve been more than frustrated with what I call the “chicletized world” of disconnected mobile apps. It’s rise was so counter to everything I loved about the Internet, I’m afraid as a result I underestimated its impact on that very world.

My corrective starting point – the metaphorical bit of yarn upon which I felt compelled to tug  – was the impact of “deep linking” on the overall ecosystem. The phrase has something of a  “dark pool” feel to it, but it’s actually a rather mundane concept: Developers tag their mobile apps and – if relevant – their complementary websites – with a linking structure that allows others to link directly into various points of entry into their applications. This is why, for example, you can jump from a Google search for “Tycho” on your phone to the “Tycho” page inside your Spotify app.

So far, I’ve had more than a dozen or so meetings and phone calls on the subject, and I’ve begun to formulate some working theses about what’s happening out there. While my education continues, here are some initial findings:

1. Deep linking is indeed a Very Big Deal. Nearly all the folks I spoke with believed deep linking in mobile was the beginning of something important, something I’ve started to call…

2. The Quickening… which I believe is nearly upon us. Mobile app developers are humans driven by business goals. If the business opportunity is large, but proscribed by narrow rules, they will follow those rules to gain the initial opportunity. For example, when the convener of a new market (Apple) imposes strict rules about how data is shared, and how apps must behave with regard to each other, app builders will initially conform, and behaviors will fall narrowly in line for a cycle or two (in this case, about five years). However, once those rules prove burdensome, businesses will look for ways around them. This is happening in mobile, for reasons that come down to new competitive players (primarily Android) and to a maturation in distribution, revenue, and engagement models (more on that below). The end result: The market is about to enter a phase of “quickening” – a rapid increase in linking between apps and web-like backends, harkening a new ecosystem in which both foreseeable and unforseen “life” will be created.

2. App Installs Rule. Till They Don’t. The market for mobile apps is – predictably – driven by app installs. And unless you’re the teen viral sensation of the moment, the only reliable way to get app installs is to buy them – almost exclusively via advertising on mobile devices. Facebook figured this out, and holy cow, did the market love that. But app makers are now realizing they have to do more than get their app installed. It’s actually just as critical to get their current installed base to actually engage with their app – lest it be forever relegated to the dustbin that is our current (deeply crappy) mobile desktop metaphor.  Hence the rise of  “re-engagement advertising,” which is serving as something akin to search-engine marketing (SEM) in the desktop web.  Several folks I spoke to told me that 80% of the money in mobile advertising is in app installs, but they quickly cautioned that installs are a house of cards which will not be sustained absent the rise of re-engagement advertising.

3. We’ve Seen This Movie. Which got me thinking. Jeez, have we ever seen this movie before. It’s called publishing. You can buy crappy circulation, crappy audiences, and crappy one-time visitors, and you can also buy great audiences, but the true gauge of a publication, a service, or an app is whether folks keep coming back. And even if you have a great app/service/publication, you need to remind them of your existence more than a few times before they are hooked (this is why classic magazine circulation has three phases – marketing, sampling, and conversion). The link-economy of the open web allowed this process to happen rather naturally, but there is no such economy in mobile, at least not yet. Thanks to early decision made by the conveners of the mobile ecosystem, mobile is deeply shitty at providing business owners with a way of reminding consumers about the value of their proposition, which is why they are frantic for some kind of channel for doing just that. This leads me to hypothesize that…

4. The App Store’s Days Are Limited. Remember when Yahoo! owned Web 1.0, because it had the entire Web in its directory? Or when Google owned Web 2.0, because it put the entire web in RAM? Yep, both those models created massive companies, along with massive ecosystems, but neither hegemony lasted forever. Apple’s App Store (and Google’s) are subject to the same forces. The model may be dominant, but it’s not going to last. As one senior executive in mobile media put it: “The app store is a weigh station, not an end point.” What might replace the App Store as a model for distribution? That’s a fine question, and one I don’t have a strong opinion about, at least not yet. But I sense the Quickening will lay the groundwork for new vectors of app adoption and engagement, similar – but not identical  – to the link economy of the web. Which is why I believe…

5. Re-engagement ads open the door to new topologies (and economics) across mobile. A pretty obvious point, if you’ve managed to stay with me to this point, but one I think is worth restatement and elaboration. Re-engagement advertising is driven by a fundamental business (and consumer) need, and Facebook, Twitter, Apple, Yahoo!, and Google are all responding with deep linking topologies that enable re-engagement. This is a relatively new development, and it’s hard to predict where it might go. But one thing’s for sure – deep linking is good for both the developer and the consumer. It’s just a better experience to go directly into the exact right place inside an app that’s already on your phone. And for marketers, deep linking enables far superior “landing pages” inside their apps, driving a conversion path that is measurable and repeatable. It’s not hard to imagine that re-engagement is the beginning of a more robust economic model for mobile, one that will re-integrate much of the goodness we created when the Web broke wide open ten or more years ago. And that makes me wonder if….

6. The home screen of “chiclets” is mutable. Broadly established consumer engagement models don’t shift rapidly, and the colorful, 16×16 sudoku model of App World isn’t going away anytime soon.  But do we really believe we’ll be poking at squares representing apps forever? I don’t. A more fluid experience based on declared and modeled intent makes a lot more sense – one in which we flow seamlessly from need to need, serviced in each state by a particular application without having to pull back, chose a new app, and then dive back in. I’ve not yet spoken to many UX/UI folks, but I sense this is coming, and deep linking is a first step in enabling it. Somehow, I sense that…

7. Search is key to all of this. Hey, this is Searchblog, after all. It strikes me that search on mobile is pretty broken, because it forces the entirety of the web onto a model that has far more specific – and useful – parameters to work with. The signals emanating from a mobile phone give search entirely new use cases, but so far, we’ve got precious little to show for it. This can’t stand for long.

I’ve got a lot more thinking going on, but it’s too nascent to be of much use at the moment. Topics I’m also thinking about include mapping the dependencies of the mobile ecosystem, grokking the concept of “agency” and how it relates to search and mobile data,  the role of programmatic in mobile, and understanding the flow of money between the big platforms and the little guys.

As you can probably tell, my comprehension of this space is still very limited, but I hope this update sparks some of your own thinking, and that you might share those insights with me in comments or via email or other forms of media. I will continue my walkabout in coming weeks, and I’ll keep writing about it here. Thanks for reading.

And thanks to the many folks I spoke with so far, many of whom are working on stealth projects or agreed to our conversations on background. Hence, I’ve not quoted anyone directly, but again, thanks, and you know who you are!

It’s Time For Twitter To Filter Our Feeds. But How?

By - July 27, 2014

article-2678561-1F58F0DC00000578-816_634x394(image)

“We don’t put an algorithm between you and your feed.” – Twitter exec Adam Bain, March 2013

“Please do.” Me, today

Twitter has always appealed to tinkerers, to makers, to the people who first took up blogging, who championed RSS and HTML in the early days – you know, the people who created the open web. And because of that, Twitter has always had a strong dose of egalitarianism in its DNA. Twitter expresses that DNA in a particular way: it never decides what you might see in your feed. Whenever you come to the service, you are presented with everything. It’s up to you to figure out what’s valuable.

Compare that to Google, which decides what content you see based on your search query or, more recently, your location (and tons of other data), or Facebook, whose impassive algorithms sift through a sea of friends’ updates and determine what the service, in its ineffable wisdom, decides you will see. Both of these giant companies have, at their core, the idea of editorial judgement – they decide what you see, and for the most part, you have no idea how they made that decision, or why.

Twitter makes no such distinction. And this, of course, has always been both its declared strength and its obvious Achilles heel.

For it is in making editorial judgements that the edges of a media product emerge – and to most of us, Twitter is  a media product (it’s certainly an advertising product, which to my mind makes it a media product as well).

In the coming months, I expect Twitter will finally execute a major shift in its approach to our feeds, and roll out an algorithm, not unlike Facebook’s EdgeRank, which consumes the raw material of our feeds and process them into a series of media products that redefine our experience with the service. Doing so will solve for three of Twitter’s most critical business problems/opportunities: Its vexing “I don’t get Twitter” issue, its slowing user growth and engagement, and Wall Street’s ongoing uncertainty around how far the company’s current advertising model can scale (IE, whether it can grow to Facebook or Google level revenues, currently orders of magnitude larger).

Three years ago I wrote Twitter and the Ultimate Algorithm: Signal Over Noise (With Major Business Model Implications). My main argument was that Twitter has to figure out how to make my feed valuable to me – a point I’ve been talking about for years. It would take a lot of math, a lot of algorithms, and a lot of trial and error, but ultimately, I wanted Twitter to surprise and delight me each time I came back, and there’s no way a raw feed could do that. In short, I argued that it was time for Twitter to create algorithmically-driven editorial voice, one that presents me media product(s) that extract maximum value out of the feeds I followed.

It’s fair to say that three years later, Twitter hasn’t done what I wished for. Back then, Twitter wasn’t a public company, and its ad business was in its early stages. But today Twitter is a $24 billion public company with strong advertising revenues tracking at more than a billion dollars a year. So what do I know?

Well, I know that the problem still exists, and there’s no way Twitter can grow into (and beyond) its current valuation, much less compete with Facebook and Google, if it doesn’t tack into the waters of editorial judgement. This means Twitter has to stare down its existential DNA problem – it has to be willing to put itself between us and our  feeds.

And I think there’s all sorts of opportunity in doing so. I think nearly everyone wants Twitter to try, and while I have no inside information, I’m pretty sure that Twitter is working hard on doing just that. Ever since the company made it clear it didn’t want developers creating consumer facing applications that built new interfaces for the consumption of tweets, the responsibility for creating that value lies squarely with Twitter.

But even as the product and engineering folks at Twitter labor to create these new interfaces, there’s no need for the company to abandon its core philosophy of showing us everything – that should be a mainstay (and differentiating) feature of the service. We just want media products on top of those feeds that mine the best stuff and present it to us in a way that keeps us engaged, provides us significant value, and thereby keeps us coming back. This of course would solve for quite a few other pesky problems – user growth and engagement chief amongst them. Oh, and it’d create the kind of media product that’s rife with signals of user intent  – exactly the place where new Twitter ad products can thrive.

Earlier this year I argued that Twitter might encourage a class of “super curators,” a kind of crowd sourced approach to solving the problem, but that’s not enough. For Twitter to grow at Facebook or Google like rates, it has to build a media product that is automated, but feels uniquely “Twitter-y.” And to me, that means making something that exposes its inner workings to its users, and lets those users customize their consumption in ways that can be shared, celebrated, and even commercialized.  In Who Owns The Right to Filter Your Feed?, I wrote “No one company can boil the ocean, but together an ecosystem can certainly simmer the sea.”

It’s my hope that Twitter lets its tinkerers, makers, and users help make it better and better. The company’s roots are as a user-driven service. Users came up with hashtags, retweets, and other core Twitter features. One of its most valuable assets is its open DNA – and it needn’t abandon that to create an algorithmically edited version of its main product. In fact, given all the suspicions both Facebook and Google have fostered because of their black box algorithms, a more open approach could be a great strength for any new Twitter product. Show us why your algorithm created a particular media product, and let us play around with making it better. I’d bet that plenty of folks would love to do just that. I know I would.

Google’s Year End Zeitgeist: Once Again, Insights Lacking

By - December 17, 2013
Zeitgeist13

Great photo, but not one we searched for….

It’s become something of a ritual – every year Google publishes its year-end summary of what the world wants, and every year I complain about how shallow it is, given what Google *really* knows about what the world is up to.

At least this year Google did a good job of turning its data into a pretty media experience. There are endless scrolling visual charts, there’s a emotional, highly produced video, and there’s a ton of lists to explore once you drill down. But there’s also a Google+ integration that frankly, was utterly confusing. Called #my2013 Gallery (sorry, there’s no link for it), it showed photos from a bunch of people I didn’t know, then invited me to add my own. Not sure what that was about. The “Search Trends Globe” shows top search terms by location, but you can’t click through to see results. Odd.

So kudos to Google for giving us a lot of eye candy – there are top ten lists for all manner of categories, from dog breeds to NFL teams to memes – all by geography. But the search capacity is, well – confusing. Once you search inside what you think is the year end Zeitgeist, you end up getting Google Trends data, and you’re kind of lost, not sure if you’re in the year-end special anymore. Bummer.

And while there are far more lists than I’ve seen before, there’s still no … insight. Even the “What is…” function, which was an interesting, if limited feature from last year’s Zeitgeist, is gone this year, most likely a victim of political correctness. (For why, see my post about last year’s Zeitgeist).

I sure wish Google would surprise us with Zeitgeist, but once again, no dice.

Traffic of Good Intent: We Beat Fraud By Working Together

By - December 06, 2013

TOGIscreen

Earlier this year I wrote a post titled It’s Time To Call Out Fraud In The Adtech Ecosystem. The overwhelming response to that riposte led to a lunch at this year’s IAB annual meeting, which then led to the formation of the Traffic of Good Intent task force (TOGI), an IAB-sanctioned working group composed of leaders from nearly every major player in the media and adtech industry. We’ve made a lot of progress since our first informal luncheon meeting nine months ago – I think the issue of fraud is now a top priority in our industry, and we continue work on best practices, solutions, and education. Today marks a milestone for our industry, the release of two white papers. Both are clearly written and intended to catalyze our progress to date.

Understanding Online Traffic Fraud gives a broad overview of the problem, laying out definitions of non-human traffic, and lays out half a dozen reasons you should give a sh*t. For me, the money quote is this: “Failing to root out traffic fraud funds criminal activity and supports organized crime.” Because as an advocate for publishers, that’s what fraud is: it’s stealing. It’s taking money and value out of the pockets of publishers, and putting it into the pockets of criminals. Along the way, any number of intermediaries also make money, and in the short term, they may be incented to continue to do so. We have to change that.

The second document, Traffic Fraud: Best Practices for Reducing Risks to Exposure, details actions all player in this ecosystem – brands, agencies, trading desks, technology providers, exchanges, publishers and more – can do to clean up this pervasive problem.

If you buy, sell, or traffic in online advertising, please read these documents, and help us move the needle even further. Fraud is not a problem that can be solved by pointing fingers or blaming one side or the other. We have to work together – and these documents are living proof that we can.

Apple+Topsy: It’s Not About Twitter (And Twitter Is Probably Cool With That)

By - December 03, 2013

TopsyApple

I’m going out on a limb, but a fairly stout one: Like Azeem, I think Apple bought Topsy for its search chops. But Azeem, who I admire greatly, says Topsy could become the search engine “for iOS… to index both the social Web, but also the best bits of the Web that power Siri and Apple Maps, [and] reduce the reliance on Google and reduce the flow of advertising dollars to the big G.” Certainly possible, but I don’t think Apple bought Topsy for its ability to search the web, or even for its trove of Twitter data. That might be a nice bonus, but I don’t think it’s the bogey.* Others have written that Topsy might be used to improve Apple’s iTunes/app search, but again, I think that’s not thinking big enough.

No, Apple most likely bought Topsy because Topsy has the infrastructure to address one of Apple’s biggest problems: the iOS interface. Let’s face it, iOS (and the app-based interface in general) is slowly becoming awful. It’s like the web before good search showed up.  To move to the next level, Apple needs a way to improve how its customers interact with iOS. Topsy will help them get there. Also, I think Twitter is happy that Apple bought Topsy – but more on that later.

Let me explain. First, my statement that iOS is “becoming awful.” Faithful readers know I’m not a fan of iOS. I switched to Android almost two years ago, and I’ve never looked back. But it’s not as if the Android interface is much better – I just like its chances of developing into something more powerful down the line. In the past few years, I’ve written several posts about the kind of interface I believe needs to emerge across mobile (which until last year, Apple pretty much dominated). Given my  obsession with the topic, it’s probably no secret that I view mobile’s biggest problem boils down to one of search.

In  Apple Won’t Build a (Web) Search Engine and Of Course Apple Is Going to Do Search, I argued that Apple must get into the “app search” game. Just as web search became the coin of the web realm, app search will be next. It won’t look like web search, I argued, but at its core, it’s quite similar.

That was three years ago, right after Apple bought Siri, launched iAds, and was relentlessly touting the growth of its app ecosystem. I was certain Apple was going to figure out a way to create value above the level of a particular app, using all that tasty data it had within its restrictive walled garden to build the next generation iOS interface.

But so far, Apple has failed to innovate inside its own ecosystem (unless you count minimalist icons and bright base colors as innovation). Three years later, we’re still stuck in a user interface of app-filled screens, most of which we never use, each disconnected  from the other save for the fact they happen to reside on your phone, possibly right next to each other, but otherwise unaware of the value they might reap should they magically start sharing links and data with each other. (You know, the way the web works.)

This has to change.

Google knows it, which is why I find Google Now so fascinating. Apple knows it too – the days of home screens littered with app icons are numbered. What will replace it?

My guess is some kind of intelligent, search-driven interface that “understands” you, based on the intent you signal through your use of all kinds of apps – including browser apps, of course, as well as true search apps like Siri (or Google Now). This new kind of interface responds to your voice as well as your location, your history, and anything else you might willingly (or unwittingly) feed it. It will strive to always put the very thing you need at your fingertips – something that simply isn’t possible without understanding your interactions as the equivalent of …. well, a personal interest graph.

And to do that, Apple needs a powerful engine, the kind of engine that, say, has been hard at work understanding a massive corpus of interest data for, say, six or so years. Something like Topsy.

My prediction: Apple doesn’t really care about Twitter data. The more I think about it, the more I’d wager that Twitter most likely blessed Apple’s purchase – and why not. With its newfound post-IPO billions, Twitter could have easily forced Topsy’s price well past $200 million. But Twitter is probably thrilled that Apple bought Topsy – Apple just took out a company that Twitter eventually would have had to either buy or kill. Now, Twitter is free to build enterprise value on top of its own data, as well it should, and Apple has a team of engineers who I’m guessing can’t wait to get their hands on a new kind of tweet stream – all that structured data captured, but not leveraged, off your mobile phone. It’s a win win win – if I’m right. Apple gets the tech and talent to build the guts of its next interface, consumers get a better OS, and Twitter gets to keep its cash and eliminate a potential competitor to boot.

Smart move, Apple. I hope I’m right.

*For the record, I spoke to no one at Twitter or Apple before I wrote this. It’s all my own brand of pure speculation.