Monetate – A Multivariate Platform for Ecommerce
One of the great parts of a trade show is being able to quickly see trends. You can see what companies are going after similar markets. You can see what companies are just getting starting in meeting a customer’s need and who is well along the development cycle. Internet Retailer-2012 gave me another chance to see an industry developing. Monetate is a single company represents new trends multivariate
Monetate is a single company represents new trends multivariate ecommerce platform optimization.
At IRCE 2012 (Internet Retailer Conference and Exhibition) at Chicago’s McCormick Place this week, I was able to how the cost of computing continually dropping is affecting how we merchandise (1st on line, but you know the wisdom will be carried to bricks and mortar stores as well).
One of the booths I got to see was Monetate. What stood out for me was how much computing is being done in real time and in an anticipatory fashion. In essence,
they are taking your shopping cart system (which may in all likelihood is probably a very large collection of SKU’s, Descriptions, media assets) and creating variations of it for all sorts of testing. They can test – by what city you come from, or if it is raining today. They can test or present – based on your past history with you, or how close you are to a Walmart. They can even handle compound testing: if you are a repeat customer on a rainy day, in a media market that is inexpensive, and there is a festival this weekend in town.
Watching the monetate’s demo and remembering their presentations, I am reminded of when I used to run scenario’s and forget to limit the sample size to ‘reasonable
number’. I would get my manager bringing my report the next day, He would show me the chargeback cost (in the mainframe days) that exceeded both my and my manager’s cost for a year. Monetate’s system lets me create these scenario’s on an ongoing basis either as test, or production – in real time.
What I see here is the magic of being able to move as much computing power before the web customer shows up. It also is able to quickly identify what the identity of each visitor is. This allows responding much more like a human store clerk would – making many decisions with ease because once you know the ‘rules’ they are easy and ‘obvious’. A store clerk seeing a visitor show up on a long skateboard is probably more interested in snowboarding then ski’s or toboggans. A customer wearing a $4,000 watch is probably not looking at the costume jewelry in the corner of the store. These snap judgements are what allow us to determine which of the thousands of cars we pass on our way to work will be a problem, and which are just going along in our lane that can be ignored safely.
We have been trying to get computers to have this same ‘artificial intelligence’ for decades. And we are still a long way from AI. But increasingly, we are able to in small domains, define experiments that let us test and tweak. Testing what is the best way to present information or experiences for website visitors that meet their needs on their terms. Monetate is helping that become incredibly easier for online shopping.
One of the other keys I see in Monetate (and similar tools), is the ability embrace Lean Startup methodology. The concept of iteration. The key is understanding that now that your cost of experimentation is lower than analysis, the best approach is lowest cost in revenue optimizing is different. When each computer batch ran on a mainframe cost thousands of dollars and days of time, it was far more effective to spend more time ‘bench testing’ ideas. We would sit around and using human power to determine the best approach. But, when a computer can now not only check your theory, but create a method to retest against different data (be it a Texas visitor over an Oregon visitor, or this month’s customers over last quarter’s customers), it is poor use of resources (especially human) to think about what the results are, and ‘just do it’. Let the system or computer or software do the testing and provide the results.
The challenge will be not in saving the last CPU cycles to run one more test, but how to let our creative juices loose again to see how to continually improve the solutions we are trying to build. One of the big challenges for many in ‘corporate’ is to understand that website visitors are individuals. The market of potential customers need to (as much as possible) be treated as individuals. As we learn to look at our visitors as having different mindsets, often determined by factors we can identify, we will be able to create better experiences for our customers.
There is a whole other dimension of visitor privacy that enters into this, but I am not going to tackle that one today.
As usual, monetate’s solutions are an incredible toolset that lifts one burden from the ecommerce manager allowing a much larger responsibility to rise to the top – looking at the world from the customer’s perspective. It was never – not the top priority, but now, there are a many fewer excuses as to why we cannot focus on the customer’s view point.
No comments yet.