What is Digital Analytics? Why is it the next frontier for marketing? We discuss these questions and more with Anametrix co-founder and CTO, Anders Olsson, who has pioneered completely new methodologies for collecting and visualizing data. Anders previously led technical design and strategy at WebSideStory as their System Architect.
We find out about the past and future of marketing data analysis in our interview:
Why did you start Anametrix? How did your WebSideStory web analytics experience contribute to the development of the Anametrix digital analytics platform?
WebSideStory was very successful at solving a tangible marketing problem: measuring and improving website effectiveness. But people kept asking us for more. Yes, we had something agile and worthwhile, yet our clients still had all this other data sitting elsewhere that was also very interesting to them, in many cases much more interesting. In fact, for most companies, web data is valuable, but it’s not the core of the business. More important are their POS data, supply-chain data, data from various advertising channels, and most recently, data from social media. They wanted to see THAT data come alive, just as we made their web data come alive. That challenge suited my tendencies as a developer; I tend to build things that are generic and open-ended so that people can use them creatively.
Anametrix is the next step after WebSideStory. Here we take the same data analysis concepts that worked so well for websites, and apply them across all channels to data that’s even more valuable to our clients.
How is Anametrix different from a business intelligence (BI) application?
Most business intelligence applications are both too hard to use and too expensive for our target user – MARKETING!
What is your vision for the Anametrix digital analytics platform going forward?
I’ve structured the development of the Anametrix platform in four phases. We move up the stack of data refinement and deployment through each of these phases:
Phase 1 is data collection. We start with collecting any marketing data you want, getting the connections to all relevant systems inside and outside the organization to pull in the data.
Phase 2 is visualization. We visualize the marketing data and provide enough extraction capability to be able to do something worthwhile with the data. We have built more sophisticated capabilities, such as Excel integration, scheduling, predictive alerting, custom data tables, etc., for additional flexibility as well as ease of extraction and manipulation.
Phase 3 is predictive data analysis. We only got to the second phase with WebSideStory – we only gathered the online marketing data and presented it to the users. With the Anametrix digital analytics platform, we’re actually working on understanding the data. Once we have all this data from the various sources in one place, we can take the data, build a regression model and produce a classifying algorithm to predict customer behavior.
These powerful models pick up nuances based on all the information patterns we have. For example, if I have a data set and I know a lot about historical behavior (by demographics, geographies, psychographics, etc.), we can do some data analysis to see how that correlates with spending patterns. We could tell you whether someone browsing the website right now is likely to spend money, and if they were accurately targeted by your ad spend to begin with. We have access to hundreds of data points. The more we know about the customers, from their various channels of engagement – i.e. the richer we can get the data – the more powerful these predictive capabilities become.
Phase 4 is automation. As marketers begin to understand the data and can predict behavior, they can be like the quants on Wall Street, with automated trading algorithms. They will make the determinations, buy the media and automate all the actions. Now, that’s the nirvana of digital marketing!
How do you see the future of digital analytics evolving?
In one word: REFINEMENT. Just imagine analytics as the refinement of commodities. For example, big data in its raw form (data mining) – it is just like a mine, you get iron ore. There’s some value, but pretty low. Then you refine the iron ore and turn it into steel. Still a commodity, but more valuable now.
Similarly, we take the source data – clickstream, advertising, social, or POS, etc. – and do some data analysis to make sense out of it. It’s more valuable now, but since we are still talking about a single data source in a silo, it is still not as valuable as it can be. If you combine the steel with glass, plastic, carbon fiber, you can now build interesting stuff like a car. That’s the final stage of refinement for digital analyics as well – very high value. Marketers start gaining truly new insights – discover new truths – when they look at all the data sources together and conduct more holistic data analysis. The final step also includes social collaboration to facilitate the dissemination of these new insights to all the business users. The result: better business decision making, based on a 360 degree view of customer and market related data.