[English version of an original post in French published on]

The internet has been available to all for the past 20 years or so. However, the masses have only started flocking to it a decade ago thanks to ubiquitous high speed internet access, within the confines of the office, at home, from the desktops or laptops, the unmobile devices. Since then, the audience has become massive, giving advertising the eyeballs it badly needs to start spreading those banners all over the content: top of the page, skyscraper verticals, or unashamed overlays. Advertising online has become more subtle, complicated, convoluted, by ways of cookies, those breadcrumbs we leave behind us and saying so much about our digital whereabouts… we have gone far beyond the very first banner ad published on in 1994.

First ever banner ad

But then things changed, almost suddenly, with an audience more massive yet. This audience simply migrated to smaller devices, mobile ones, phones and tablets. Already last year, the user base of those tiny all-in-one computers has surpassed the “legacy” user base of those computers with a screen greater than 15” (cf. Internet Trends 2014 report from Mary Meeker @KPCB, slide 8 . This new device and usage paradigm disrupted two key components of the former advertising model: a generous screen estate, as well the cookie-friendly browser as the go to interface for most things web. Sure, the internet is growing still, but more on the mobile side than the unmobile one…

This begs the question: how will the advertising banner evolve in the mobile space? how will this artefact of the past manage to capture the user’s attention on a tiny 4” screen? It is alive and well you may say, we see it here and there, top or bottom, sticking like a sore thumb on a cramped display. Unfortunately, it does not “click” (or tap!), it converts poorly, it does not pay well (around 10x less than on the web), but also, since the eyeballs are more numerous than ever, it is gamed and soiled by fraud. Is it time to give up on this format? Is it time to kill the banner?

It is clear that despite all expectations, the mobile world has become largely app-centric. The much touted HTML5 standard has yet to impress and topple the App Stores. Indeed, on mobile, over 85% of the entire usage is done within apps those little functional units that people download and use for very specific and dedicated use cases. The combined usage of those apps is massive, spread from dawn till dusk, during commute, or stolen moments here and there, usually with haste.

Usage on mobile is highly fragmented. More than ever, users switch from task to task according to their daily activities, across both time and space, with variable attention span. It is not true multi-tasking. In fact, smartphones typically do not handle true multitasking as we know it on unmobile computers, it would consume way too much energy and deplete the battery in no time. Therefore, tasks are fragmented, sequentialised, one-app-at-a-time.

Hence, the powers that be, those very corporations that have dominated the past decade, now declare themselves as being “mobile first”, with a definite “app-centric” approach: Facebook, Twitter, Apple (they don’t have to say it; they mostly sell mobile devices now), Google (haven’t announced yet, a little late on the mobile advertising front despite Android-ownership), even IBM! The proverbial Mad Men seem a little anxious to get into the game if by judging the recent deals: Publicis/Facebook or Omnicom/Twitter contractually bound for hundreds of millions of dollars to secure budgets and spend on this medium that they do not yet fully grasp. Time to bet on the new horse, the winning horse; after all, some say it’s just a matter of time until all traditional budgets shift over to web or mobile.

Mobile technologies are second to none. It is the usage fragmentation and frequency, the utterly brutal and disloyal thumb twichting, all drowned in a sea of millions of apps to choose from, that together contribute to pushing the limits of the mobile economic model faster than anything else before. Big data is not an option, it’s a default. Traceability, attribution even multi-attribution, reporting from end-to-end, performance monetization models way past the old “display model”, accounting for app installs, app re-engagement and app re-targeting.

Criteo, one of only two French unicorns since 2000, recently IPO’ed on the NASDAQ stock exchange is leader in re-targeting technologies. Re-targeting is a technique that advertises specific e-commerce items previously visited but not transacted upon yet. This is typically rendered in a banner format of sorts, targeting the very users who did the online equivalent of window shopping hours or days before, prompting them to complete their purchase. The conversion rates on these banners are spectacular, on the web that is: clicking on such a banner simply opens another page landing straight on the item almost purchased before…

On mobile, things aren’t that simple. Indeed, the act of e-commerce transacting is lagging behind, stuck in the immobile world of the desktop browser at home or at work, anchored by the established signals of the reassuringly green SSL certificate symbol, where entering the credit card number feels somewhat safe. “Or perhaps I’ll complete a transaction on my mobile but later, because I’m between phone calls, between two subway stations… wait, but if I tap on the banner, then I’ll be taken away from my mono-usage app, how do I get back? I’ll forget… ok forget it, let’s not tap on this banner.” This rather radically different use case makes or break the flow; it is extremely hard to invite oneself into the context of a mobile user as it’s usually never the right time. Wearables and other connected devices may start to give us insights into the right moments. Facebook acquiring the Moves app is a move in the right direction adtech-wise.

Deep-linking on mobile is a strong emerging trend in 2014. Yet, we are far from massive adoption because it requires each publisher to start embracing the ramifications and the potential; they must implement a transaction oriented re-engagement and re-targeting architecture. Until now, most apps mostly focused on the raw user acquisition, caring mostly about the on-boarding just after the download, and a little about endogenous re-engagement using push notifications.

The likes of Facebook, Criteo and ourselves at Appsfire are already offering mobile deep-linking: that is the ability to bring a user directly to the right item inside a given section of the app for a given user. It will take time until each publisher can manage this level of fine-tuning. The major publishers will need to get on-board rather quickly.

Deep Linking Example

The forces at play are self-organising around protocols, some open as a consortium like or others organized by private initiatives such as the recent Applinks by Facebook. None of that is sufficient however. The building blocks are not all in place: mastering the notion of time relevancy as mentioned above, or the notion of reciprocity to come back to where the user left off in their fragmented usage. Mobile payment transactions also play a great role in the puzzle: Apple (with its 800 million credit cards on store), Google with its check-out, Facebook with its pending currency, and PayPal all vying to be the keystone.

Important WWDC 2014 update: a few weeks ago, Apple announced that with iOS 8, its iAd ad network will start offering re-targeting technologies by tagging, less laser focused than the current re-targeting champions, but nevertheless a step in that direction with a strong respect for privacy (as Apple would). Even if it’s still not strategic for them, advertising has gone a long way at Apple, frowned upon at first, relative failure in its initial ambitions with iAd and now progressing swift-ly. Other new technologies announced at WWDC will perhaps allow cross functional innovations inside and between apps: extensions, interactive notifications… Meanwhile, Apple keeps cleaning up advertising malpractice around incentivised downloads and video viewing – a very good thing. So Apple now tolerates advertising but will definitely regulate it… More than ever, the key resides in the relevance and scope of integration and targeting.

In conclusion, while waiting for HTML5 to strike back, the mobile app economy and therefore mobile advertising are organising around the following sequence: the download, the on-boarding, the re-engagement, and the re-targeting. Mastery of each link in the chain is required; the stakes are high… analysts at eMarketer say it will be a 100 billion dollar industry by 2018.