 "digital marketing")
Only a few years ago, Google was a cool search engine looking for a revenue model. In 2002, four years after launching 'search', the company launched a pay per click (PPC) model on its Adwords platform where marketers paid for clicks instead of impressions. PPC has since been a cornerstone of Google's success and currently contributes significantly to its $50 billion annual revenues that is more than the combined revenues of Infosys, TCS, Cognizant, Wipro and HCL.
Marketers adopted PPC widely because they paid only for clicks and could sharply target audience and had better control over budget. PPC is now a norm on Bing, Yahoo and Google that account for more than 90 per cent of the global search traffic (except China). As consumers we are all familiar with text ads placed alongside search results. (WHO GETS THE CREDIT)
We are now spending more of our lives surrounded by digital media with smartphones, tablets, and PCs that are always connected to the internet. Marketers have kept pace with their audience and are spending more than 20 per cent of their ad budget on digital marketing. While the total market for advertising is growing at 3 per cent per year, digital advertising is growing by more than 10 per cent and mobile advertising by more than 50 per cent every year.
How do marketers know that their digital marketing dollars are working? In a world where a text ad is the only influence on a consumer, measuring cost effectiveness is simple: if you were selling a camera online where you made a profit of $50 per camera, and 10 per cent of people who clicked on your Adwords had bought the camera, your profit per click would be $50x10 per cent = $5. So you could bid up to $5 for each click and still turn a profit.
This math gets messy in a complex world with multiple online and offline influences on consumers. Just imagine the last time you bought a smartphone. How did you decide? Let's assume you started with two brands in mind - Samsung S4 whose ads you watched across platforms and iPhone 5S, whose launch you were waiting for. Then you might have compared these phones (by searching on Google for expert reviews or by going to Flipkart for user reviews), only to realise that your options should expand to include Nexus 5 and HTC One. You clicked on a bunch of links, spoke with a few friends whose recommendations you received on WhatsApp or Facebook, got thoroughly confused and ended up deciding in favour of, say, Nexus 5, not just because it seemed to have the best features for a reasonable price, but also because it was Android and some e-commerce store was selling it with an irresistible cashback offer!
When millions of consumers interact with each other, consume online and offline content, and make buying decisions offline or online, it is difficult for marketers to understand which media worked and the return on that investment.
Multichannel attribution is the science of determining what credit every relevant marketing channel should receive for playing a role in driving conversions. This is a critical first step in optimising marketing spend.
One clarification is in order before we proceed. Not all companies are hoping to make you buy online. Their digital spend is directed at making you 'do' something - it could be as simple as registering for a free webinar, downloading a pdf or just reading more about benefits of their cooking oil (and consequently buying it at a nearby grocery store). Each of these events is called a conversion.
Multichannel attribution is an imperfect science that requires artful interpretation. The simple (and inaccurate) ways of assigning attribution include, first, giving all credit to the last click before conversion (last click attribution), second, giving credit only to the first click that initiates conversion (first click attribution), or third, giving equal weight to all influences in the conversion path (linear attribution). (See chart: Multichannel attribution)
Companies that have developed proprietary models accurately ascribe the likely 'contribution' of each ad and the return on media investments. Good multichannel attribution models can increase relevance and effectiveness of digital marketing. Multichannel attribution, as a thumb rule, should improve conversions by 10 per cent with the same budget or reduce budget by 10 per cent for the number of conversions. That's a potential impact of $50 billion for the $500 billion per year advertising industry.
One final and important point before we conclude. So far, we have mostly discussed 'paid' media, where advertisers pay to get coverage (see chart digital media footprint). Consumers also access the 'owned' media of brands, such as their company websites and blogs. In an increasingly social world, consumers voice their opinions and influence purchase decisions of their friends through consumer reviews, blogs and social sharing - this media coverage (positive and negative) that brands receive is called 'earned media'.
While paid media may be the easiest to build an audience, it is also more expensive and often seen as an interruption by consumers. Owned media is the easiest to control and cost effective but is slow to scale audience. Earned media is almost impossible to control but is often very credible and remarkably effective when it scales (think Gangnam style and its 1.7 billion views on YouTube).
A marketer must take this holistic view of marketing, use paid media to generate awareness, create quality content to drive user engagement and social sharing to build brand and generate conversions. Digital and mobile advertising will continue to grow dramatically; it is time marketers employed state of the art tools like multichannel attribution to optimise their marketing.
Srikanth Velamakanni
co-founder & CEO, Fractal Analytics
co-founder & CEO, Fractal Analytics
