The “Real-time” of advertising

From the crisis of control to the turn of hypercontrol

One of the most interesting thesis on marketing and advertising explains how their rise spread from a reaction to an increasing state of lack of control.
In the XIX century a huge industrial push increased enormously a productive capacity that, suddenly, had to wide its selling horizon – and, as result, the terms of consumption. Marketing and advertising engaged themselves on how to match producer and consumer needs at large scale. After some centuries, we can say that marketing gained esteem and dignity regarding its original aim. Unfortunately, considerations about advertising effectiveness are less enthusiastic if we still comment it using the phrase “spray and pray”. However, diffusion and use of online media have increased very much  chances to overcome this never-ended skepticism on the ability to hook the timely needs of users with knowledge and speed.
At the beginning of this reflection we have to make a premise about what we have experimented in the meanwhile. Everywhere there is the possibility to find people or environment/events polarizing human attention, there is also the certitude to meet some kinds of advertising elaborated by marketing strategies.  As result, attention prompts by information and  entertainment needs that media as newspapers, radio and television serve was a ponderous‎ attraction for advertising investments becoming the backbone of their very survival.
The same thing happens with new media, above all after we are living by fixed and mobile devices in a spaceless and timeless world. From this new context of means and uses, hopes are rising to see realized the original promise to be able to link suddenly the user desire and its achievement.
The acquisition by Twitter of an unknown company – unknown for most people but very important  in the mobile advertising market – gives us the opportunity to unveil – remembering Luhman’s and Giddens’ teachings – some aspects of what we can call new expert systems given their ability to bind technology, technics, competence, know-how and routines in which we (misteriously) operate and (then) confide on.

The size of digital advertising
ChartOfTheDay_1188_Global_advertising_forecast_nFirst of all, we can check the economic dimension of digital advertising. The picture shows the investment shift  among the different channels. Being a fundamental driver for the internet developments, the digital world is very sensitive in catching and guiding the evolutive possibilities of advertising  based on technological enhancements. Consequently, the deep changes involving marketing techniques are common topics on the Net. In effect, it has been noted that creativity and mass targeting are loosing traction in favor of technics focusing on  analysis and matching of  the big data extracted from the many digital traces that people leave behind their online activities and presences. Thinking of ability to trace people by cellphone – from geo-location to interception based on sensories or social activities – we can say that service/info providers have an over-supply problem whose limit is only the fantasy and, at the moment, the law barriers. But (maybe the scariest thing) the MAGNAGLOBAL-Online-Ad-Forecasts-in-2013-Select-Formats-June2013(likely) user rejection knowing how deep their intrusive capacity is. However, the argument is complex. Nevertheless, we can’t negate that we are already tackling a continuous compromise among the possibilities given (almost always badly disposed) to scrutinize  our activities/ideas and the personal exploitation of free, useful and ubiquitous services.
Briefly, with data mining enhancements marketing can develop a very specific user profile based on real behaviours and real pattern of consumption. We have already developed a mental habitus related  to this accuracy as far as to show a sort of bother toward large scale advertising, considering it  as pure spamming.

“As data availability has improved the quality of potential opportunities, perpetual connectivity has increased the quantity. The proliferation of mobile devices, decreasing cost of data access, and popularity of social platforms have increased the number of avenues that brands have at their disposal. However, these factors have also made it harder for them to actually connect; today’s audience has an ever-shorter attention span, as focus is split across an increasing number of channels. Consumers are better informed than ever before; a price- or fact-check is simply a Google search away, whether users are in front of their TV screens or walking around stores.
This changing ecosystem has already resulted in a shift from transaction-oriented “product-centric marketing” to a relationship-focused “human-centric” approach. Consumer loyalty to brands is a thing of the past; brands now demonstrate loyalty to their customers. Marketing campaigns must be time-, place-, and context-aware. It’s also important to be technologically on point: cross-device, multi-channel, constantly measuring results and refining campaigns. Today’s technology has largely mitigated the Wanamaker problem, but a Goldilocks problem remains: consumers don’t want to be inundated with marketing that is too obnoxiously broad or too creepily specific. Consumers want relevance.” (DiResta, 2013).

About relevance, I come back to the “real-time” of title to face the cutting-edge of digital advertising, the Real Time Bidding (RTB), a new way of delivery having a general application but that wants to be specific for the mobile fruition by smartphones and tablets. Following RTB techniques we can scan the deep articulation that previous considerations have sketched.

The turn of “real-time”
Starting from the news, Twitter paid 350  Million $ for the acquisition of MoPub, a company runs a  platform to exchange  mobile advertising. More specifically, MoPub is the biggest RTB platform in the mobile market – it permits to people to sell “attention dedicated to mobile devices” and does it “daily, billions of times”. It seems that Twitter did well because the acquisition tackles two general trends such as the preference of people to move activities on mobile, and the shift of advertising selling processes toward automization. Advertising automation circuits are a recent novelty of web technologies and their proliferation follows the availability of timely information about user behaviours and needs, most of them recovered above all from social networks.
Briefly, RTB system runs to immediately match advertising demand and supply, maximizing price and targeting, a win-win process for all participants. As we said, there are many sources of data enabling these mechanisms. The browser is the most common one. Indeed, it is often damned for its devious effects of tracking (thinking of cookie, a piece of information dynamically exchanged between browser and web sites – original aim was to have memory of our passages to avoid the repetition of the same actions in terms of data insertion, navigation paths, etc.). As result, many initiatives are going to easily limit the functionality, see the “do not track” request to instruct automatically web site and ad servers – from the browser – to disable it. But we must not be ingenuous. As a recent study shows, it’s possible individuate an internet user – with 94% of precision – by the unique specificities of our digital footprints derived from combination of hw/sw device characteristics (Acar et al, 2013).
(However, cookie tracking is simple and diffuse and alternative modes require new investments and a policy of armonization  among software systems; this explain why  advertiser lobbies still contrast the “do not track” function).
However, the browser can tell advertiser what and where it has been seen some type of item (e.g. a camera). The most valid information regards the interest of people while the place of information, also important, is a easier recoverable one. However, social networks have marked a radical shift in term of the quality of information gathering. Twitter knows us very well: followers and following people, what we write, while Twitter’s bottons to log in through are distributed everywhere.  These data have a unique value at  real-time web level and their utility could be valid for a long time. If the intent window regarding the camera could last 1 o 2 weeks, the fact being a fan of somebody indicates a demographic category that does not change for years. Only Twitter or Facebook know this information because linked to your ID and not to cookie, that can change during the time or being eliminated intentionally.

Targeting and attribution
MoPub platform works to maximize advertising compliant to the actual need of advertisers. The publisher that owns a such system into our business network is more confident to be able to increase its revenues. But there are more advantages considering the potentiality deriving from the control of all user experience.

“It’s not knowing your age, gender, and rough whereabouts. That has some value in the ads space, but it’s pretty minimal… No. It’s about knowing that the same person who is browsing for a pair of shoes on Zappos.com on their work computer is the same person who’s whiling away a few minutes on their iPhone later that day while waiting for a friend. And also knowing it’s that same person who’s checking Twitter on their iPad when they get home that night. Threading together all those experiences (one desktop computer + two mobile devices) is something nobody actually does these days. There’s a tiny minority of companies who can even do this, but Twitter is one of them (Google, Apple, Amazon, and Facebook are the others)” (Garcia, 2013).

Since identified by an account and associated to the own digital footprints furnished by the various access devices, our user experiences are traceable all the time. Now, the magic can be exploited by MoPub in terms of targeting and attribution.

“If the person who is browsing your shoes site can’t be found online again (via a desktop ad exchange, for example), there is nothing you can do with that data.  … However, if Twitter can find that person for you while they’re playing Candy Crush Saga, that changes everything. Suddenly that Candy Crush Saga ad impression goes from some bargain basement $0.20 per thousand ad impressions to $20, you have your lead again, and the entire mobile ads world changes. ” (Garcia, 2013).

But, as matter of fact, even attribution becomes easier because Twitter can be addressed as the origin site where somebody RTB Growthclicked and bought following advertising message. As result, site spaces can be sold better. Google is a pioneer of advertising automation with platform as Google Display Network, DoubleClick Ad Exchange and Invite Media, followed from Microsoft that has had a minor role given the diverse take on searching services. Yet, many new players are populating ad networks.  For example, Facebook introduced its own ad exchange (FBX) in 2012 and this kind of business becomes usual – see the recent announcement by mobile social network Foursquare (Foursquare is a social application to share the physical positions of people among contacts). The below picture shows RTB evolution and forecasting adoptions until 2016. The blue line is the digital ad percentage of ad total managed by RTB techniques (28% in the 2016) while the black bars represent the million $ revenues. Figures refer US market – the red line registers the RTB percentual increments, with a big initial growth rate that slows once becoming a consolidated commercial practice.

 

The back-end
Finally, we’ll try to describe the RTB mechanisms that work silently “everyday, for billion of times” behind our quite browsing. To sum: RTB platform permits to auction (in real-time) our editorial spaces and advertisers to propose a price for those specific impressions, well-defined in terms of target but also more expensive – it’s named advertising premium. howdspssspsandadexchangeswork_opt-1The process is completely managed by IT platforms – ad server, SSP (Sell Side Platform), DSP (Demand Side Platform), RTB Exchange. The DSP systems aggregate and manage advertising demand – buyers – and respond each time to the bids launched by SSP systems – sellers – that close the deals choosing the best offer, and then inserting advertising in the relative slot. In effect, web pages are an instantaneous assemblage of contents arriving from disparate and  various sources. Generally, web publisher manages personally a part (more or less 50% ) of total advertising with its own ad server while the circuits of ad networks  fill the remainder.   An example of ad process involving a download of a web page is described by Wikipedia.

“A certain user visits a website with one displayed advertising. A call is then made by the exchange server (database web server) supporting Real Time Bidding (RTB) to the Demand Side Platforms (DSP) or Ad Networks (Ad Exchange) to determine which advertiser gets to serve the ad. Each user has an associated set of attributes (cookies), which is transferred from exchange server to the DSP and determines whether the user has the desired attributes (cookies) that the advertiser wants to target. Based on the perceived value of this user, a bid is placed on this ad impression by relevant advertisers and the highest bidding advertiser gets the placement.
Let’s explain with a simple real time example: a user heads to a page on a website (the “publisher”), causing it to start loading. In the same instant the website (“publisher”) sends out a “bid request” to thousands of potential advertisers saying, “We’ve got this user who is 30, Indian, male and based in New Jersey, US, and recently searched for return air tickets to Delhi, opening a page on our site. How much are you willing to bid for being the only ad on this page?” Within about 100 milliseconds the publisher receives bids from different advertisers, which in turn analyzes the group to determine both the highest bidder and the brands being advertised. The winner is alerted by the publisher and allowed to place its ad on the page. The remarkable thing about this entire process is how fast and how often it takes place. The entire series of to-and-from communication between publisher and advertisers takes place in 300-500 milliseconds, causing no visible delay to the user. This process is repeated for every ad slot on a page.”.

A final consideration. Often we get the proofs that industries more related to the desiring body, such as porn and adverting, have a better feeling with new media, so much that they can immediately exploit their smartness for  their own advantages. Another demonstration of how – recalling the lesson of media historian  John Durham Peters – of those bodies the media are a strong emanation.

 

References

Acar et al., 2013, FPDetective: Dusting the Web for Fingerprinters

Beniger, J., 1989, The Control Revolution: Technological and Economic Origins of the Information Society, Cambridge, Harvard University Press.

DiResta, R., 2013, “Demographics are dead: the new, technical face of marketing“, radar.oreilly.com, 3/11.

Foursquare Opens Up Its Self-Serve Ad Platform“, allthingsd.com, 14/10/2003.

Global Ad Market to Grow 3% in 2013“, Statista.com, 17/06/2013.

Garcia, A., 2013, “Why Twitter Buying MoPub Is a Very Big Deal“, Medium.com, 27/10.

Peters, J. D., 1999, Speaking Into the Air:  A History of the Idea of Communication, Chicago, Chicago Press.