The digital revolution, offering unparalleled opportunities within internet retailing, shows no sign of abating. Knowledge of key internet technologies will allow brand, values and personality to be maintained across every channel, while ensuring a seamless customer experience. Marketers who use these new technologies in combination with data analysis and appropriate targeting will drive better results. At Essence, we believe there are three digital trends that are set to make a huge impact on internet retailing in 2012. These are big data and attribution, the evolution of display advertising technology and finally f-commerce. All of these trends offer retailers the chance to interact in a more informed, personal and immediate way with their customers.
Big Data & Attribution
A study by BIA Kelsey in March 2011 in the US suggested that, on average, consumers look at 7.9 different sources for product information. There are multiple touchpoints and multiple opportunities for retailers to influence consumers, which makes it critical to understand the complete journey and attribute influence where it is due. “Attribution” allocates actual sales across the marketing touchpoints that have influenced them so that retailers can make informed decisions based on a rounded view of ROI.
The idea of “Attribution modelling” has been around for some time but methods were generally very approximate; useful perhaps directionally or, in practice, not at all. The challenge with attribution has always been the same: the sheer volume of data involved. Web-scale: many billions of impressions, clicks, rich media events and, ultimately, sales creating Teraor even Peta bytes of data. The problem was that you simply couldn’t store and analyse all of the interactions for users who bought and also for those who didn’t (it is essential to have both sides for modelling probability) or all of the associated meta-data to properly assess the impact of individual marketing ‘recipes’.
The Big Data revolution has made the challenge of genuine attribution addressable, in a robust way, for savvy retailers. The revolution is two-fold: Cloud Computing, offering the accessibility of large-scale computing resource without vast capital commitment; and secondly the invention and subsequent democratisation of technologies to query such vast volumes of data (Hadoop, must rank as the best known of these).
Big Data tools make a new approach possible: you can measure influence, not just predict it based on estimates and averages. Efficient sampling and mathematical decision-making techniques make this do-able and deliver truly accurate and actionable results. Big Data tools are also fantastic at finding specific answers with huge real-world benefit to retailers: how much lift do homepage takeovers give my other activity? To what extent is re-marketing cannibalising existing sales? How do TV spots affect my search conversion? … and so on.
Finding questions is clearly not the challenge, but at least Big Data is starting to deliver real, specific and actionable answers.
The Evolution of Display Advertising Technology
The economic environment is making advertisers pay even closer attention to what works for them. For years, the growth in online marketing has been driven predominantly by paid Search. However, according to research from the IAB, display is now the fastest growing online driver in Europe, a trend that looks set to continue into 2012. So what has changed? Display has undergone a technological revolution.
Targeting has moved from a site-based approach to a system that identifies the audience; and retargeting (showing adverts to people who have previously been to your site) has moved display into the ‘lower funnel’ space occupied by search. Simultaneously, Display Exchanges have emerged. Websites place available ad space on an Exchange, which advertisers then bid for. This bid can be set based on any information the advertiser has about the impression. It could be: contextual (this impression is on a Men’s Lifestyle site), the advertiser’s own data (this person has been looking at shirts on my site) or purchased third party data (this person has shown shirt purchasing behaviours elsewhere online). The theory is that advertisers can make much better targeting decisions, and can buy display in a more cost-effective manner. In practice, scalable, effective display buying through this method requires using more than one exchange, which can become difficult.
Hence the next development: a Demand Side Platform (DSP). DSP is a technology platform that enables advertisers to bid simultaneously across multiple exchanges. It is a single point of management for data, audience segments, bids and budgets. Using a DSP also allows use of rules like ‘Do not bid if this person has already seen three of my ads today.’ Anyone who has been disturbed by over-zealous retargeting will know why that is desirable.
For example, using a DSP I can identify available impressions of people who have recently browsed designer shirts on my site but not yet bought. I can bid on these impressions across multiple buying points. I can vary the bid based on factors like the number of my adverts the person has already seen today, and the time of day. For every impression available, the DSP decides in real-time what bid (if any) to place for me based on the rules I’ve given it. For an internet retailer with detailed audience data, these developments are potentially very powerful.
F-commerce
Drawing a line between the traditional web and the social web is increasingly irrelevant. The new web is the social web. It seems certain that in 2012f-commerce will become a significant part of online activity and engagement. Some 800 million users have proven Facebook is indeed no fad, rather it’s now a utility, and progressively becoming the new web browser. As such, commercial offers and transactions on the platform are simply a given. For me, it’s simply a matter of ‘when’ rather than ‘if’. ‘How’ also plays a huge part – right now there’s too much friction in the payment process, but those pioneering in the space will resolve this, just as they did in eCommerce.
It’s imperative to understand that f-commerce is as much about marketing as it is about revenue – those seeking to compare eCommerce revenue with f-commerce are simply missing the point. For example, take a premier league football club: within a minute of the final whistle, the goal scorer’s shirt can be posted to the Facebook wall with a contextual offer, “Epic Win! To celebrate Stan Vercy’s goal, we’re offering 10% off his shirt”. This message goes out to millions of the club’s fans, and then all interactions around the post are shared and seen by millions more people associated to each fan’s network. This is pure awareness and engagement – it’s got nothing to do with revenue. Now, assume I buy one of these shirts; my purchase can be shared with my entire network (and beyond). This generates even more awareness, but also a direct recommendation and endorsement from me. That’s awareness, advocacy, recommendation and revenue, all in one hit.
It’s still early days and f-commerce is currently reserved for pioneers, just like eCommerce once was. Those who set up shop first will most likely get the best plots.
And so, as the media report daily on economic gloom, a fall in consumer spending and massive profit losses, retail is undeniably facing a tough 2012. If internet retailing marketers wish to weather the storm, I believe they should embrace new technologies. With an overwhelming amount of data available at your fingertips, brands will be able to maintain a consistent voice across multiple channels to interact with customers and increase sales.
Source: Internet Retailing
Comments