We're living in the era of the digital data deluge. Think about how often you check your bank balance online versus going to a teller, how many emails you receive a day, the number of hours you spend on your smartphone not making phone calls.
Consumers leave contrails of data as a result of their digital interactions, and this behavioral data creates opportunities to drive customer acquisition, reduce operating expenses, and make faster, better decisions.
But there are also some fundamental problems.
75% of all digital data is created by individuals. The advent of ubiquitous computing is changing the nature of data – there are 5.3 billion mobile subscribers in the world today; 25% of Internet users in the U.S. are mobile-only; and 1 out of every 13 people on Earth use Facebook.
The sheer size and quantity of real-time digital data can't be processed by existing solutions and can overwhelm marketers. Even worse, data can be incomplete, duplicative or siloed to the point that it actually slows the speed of decision-making. Dealing with this deluge and translating it into an enterprise asset will be the defining challenge for marketers. The good news is that advances in automated decisioning technology have redefined how we manage data and make decisions.
The evolution of automated decision making
Automated decision making has historically evolved along three dimensions:
- Business impact
- Decision rate
- Freshness of data
In the 1990s, the emphasis was on automating relatively few decisions of high impact (for instance, simulating the effects of major changes in the supply chain). In the past several years, with more current data available, businesses have accelerated the rate at which decisions are being made and automated a greater variety of decisions of lesser business impact. Think of the online speed of approving consumer credit (which used to take up to a week).
We now have the technology for using enormous amounts of real-time data to make instantaneous decisions about a variety of low-impact business decisions whose combined effect is large. High-frequency trading is a great example of this, as is programmatic buying across online exchanges. In the latter case, automated decisioning systems can determine milliseconds whether or not to bid on an individual ad impression and how much to pay for it based on the performance of previous impressions.
How do we start the real-time revolution?
No real-time revolution will happen until marketers commit to putting their data to work. Here are some important considerations as you get started:
- Treat the web as a medium for continuous research and experimentation – a giant sensor with billions of inputs. Traditional market research techniques like focus groups and surveys have centered on what people say they want, not what they actually do. On the Internet a company can experiment with product features and pricing and measure responses in real time. It only takes about 1000 consumer actions to reveal a pattern worth betting on.
- Come to terms with the new reality of advertising. It's a data and analytics game. Over the next five years, corporate marketing budgets are expected to decline, and the challenge is to make these reduced budgets work harder for you by understanding and profitably fulfilling dynamic demand.
- Operate in real time. Waiting for static reports to describe what worked in the past your increase your competitive edge. Your digital marketing system should be able to turn insights from consumer behavior into immediate action and continually learn and reset based on the freshest information.
- CMOs and CIOs should work together. CMOs are struggling with how to utilize all the marketing data available and are in need of guidance. CIOs have traditionally focused on supporting other departments, but today, corporate profits can be increased by focusing on creating efficiencies with automated marketing platforms.
- Treat your analysts and developers like rock stars. McKinsey Global Institute predicts that the U.S. will see a shortage of 1.5 million Big Data-literate managers and analysts in the next six years. We have the technologies to transform the process and reduce the cost of customer acquisition, but we need smart developers to put the technologies to use.
Finally, beware of being "comquested." If you don't service the needs of your customers, your competitors certainly will. They may already be using a digital marketing management platform to deliver a more relevant brand promise to better satisfy your customers - and might not even sense that it's happening.
Julie Ginches is the Vice President of Marketing for DataXu.
No hay comentarios:
Publicar un comentario