There is no denying that predicting outcomes, and attributing those outcomes to marketing channels, media, or touch points, remains the biggest challenge faced by marketers, even today. However, while a lot has been done over the years to address this challenge, no standard attribution model could be established.
Now, with social and mobile becoming ‘the way of life’ for digital age buyers, the B2B Buyer’s journey has evolved a lot – expanding the role of marketing to cover even those aspects of the customer lifecycle which traditionally fell under the purview of Sales. While previously, Marketing was responsible only for generating demand, and handing over qualified leads to Sales, today it has a major role to play in deal acceleration, revenue growth, and customer retention. This transfer of responsibility from Sales to Marketing has made it all the more important for marketers to introduce predictability within their marketing engines.
A recent survey report, by Forrester, suggests that companies using predictive marketing analytics significantly outperform others across business metrics. Doesn’t that sound logical, anyways? In a multi-channel, multi-device marketing model, if marketers can predict and attribute outcomes, it will make their job of optimizing channel effectiveness, allocating budgets, and defining the media mix, much simpler. This will directly translate into better returns on marketing investments, while also impacting the overall profitability of the business.
However, having acknowledged the importance of building predictability within our marketing models, doesn’t make this task any easier. Practically, the famous quote by John Wanamaker – “Half the money I spend on advertising is wasted; the trouble is I don’t know which half”, still holds true to a great extent. Moreover, in most cases it’s applicable not just to the advertising spends but to the overall marketing spends of organizations.
Further, the extremely fast paced evolution of the marketing technology landscape adds to the complexity of the situation. I was overwhelmed looking at the year on year growth of this space, charted by Scott Brinker in his Marketing Technology landscape super graphic. With intervention of technology across the marketing value chain, and a plethora of choices at their disposal, marketing leaders today need to make difficult choices.
On the one hand, they need to be competitively well-equipped and make the best use of available technologies. On the other hand, they can’t afford to give in to the fear of missing out (FOMO) latest solutions that their competitors are using, and end up stacking up costly technologies that hardly promise any value.
Coming back to predictability, we need to understand that predicting marketing outcomes is very different from predicting the weather – since we have a lot more control on our marketing strategy than we have on the factors that are responsible for the weather.
It’s not like throwing a fair dice and waiting to get a ‘six’. It’s more like playing snooker – focus, precision, and tact matter. To our advantage data science has evolved a lot over the years to help us make smarter marketing decisions. The key is to focus on the right set of prospects, instead of wasting resources on mass outreach. In a B2C context, this approach is fairly simplistic – since it’s about targeting, engaging, and retaining individual consumers.
But in the B2B context, marketers need to understand that although it’s individuals who plan, evaluate, and finally make purchase decisions, to establish a long term relationship it’s about engaging the account, not just an individual. The key is to begin with a contact-based approach and gradually expand with an account-based strategy.
Try deep learning using MATLAB