The Bayes Analytic engine is a machine learning classification engine designed to predict which stock offers a have the highest probability of success. This can be applied to price movement or customer behavior. The engine can be of benefit to organizations that maintain long-term relationships with their customers. It can ensure more customers receive offers for products they will find valuable which can increase the sales per contact and cause customers pay more attention to future communications.
Major retailers such as Nordstroms could be would be a great candidates to benefit from this engine because they have long-term relationships with customers which we can mine both for purchase history and purchases relative to offers to maximize sales. I think in many cases the prediction engine eliminate 90% of the customers who would not respond favorably to a given offer such as “cosmetics” while keeping 90% of those who are most likely to respond favorably. This can increase sales per contact but even more important their customers will feel they are receiving more value from communications and will be more likely to pay attention in the future. We can reverse the same process so we can determine which products a customer is most likely to like so they can customize what is offered to the customer and reflect the customers changing behaviors in near real-time.
This engine goes way beyond simple affinity engines where if you purchased a tennis shoe I will also show you socks which is easily available. This engine can determine that since you purchased a tennis shoe on a Tuesday in Jan and a hat in Feb and have purchased more than 4,000 over a 3 year period and compare those to other customers and determine that you are also a great candidate western shirts in December and a horrible candidate for perfume especially in January.
When using the stock prediction engine for stocks and options we discovered all kinds of correlations that we would never have anticipated. For example buying any PUT option for SPY or CAT on a Monday or Tuesday is a bad idea but if you buy the same option on a Thursday you have a 82% better chance of of a win. The technology could find a very similar correlation in the Nordstrom’s data where I could identify specific days that specific sub groups of customers are more likely to buy given products. I suspect there are people who are great candidates for perfume on Tuesday and OK candidates for shirts of Saturday but you would almost never sell them perfume on Saturday. If Nordstroms knows about these patterns they can adjust their messaging and presentation to reflect this knowledge. Another example If you buy a SPY call in the 4th week (very last couple days) of the month then it doubles your chance of a win compared to the 2nd and 3rd week. Other stocks have unique characteristics just like customers. Customers are also like stocks in that their characteristics and behaviors change over time so the engine must transparently accommodate and properly weight the more recent behaviors.
I am willing to sell at my normal consulting rate for 6 months to adapt this engine for Nordstrom’s use. If Nordstrom’s agrees to pay for the full 6 months of consulting I will give them a non-exclusive license to use the engine world-wide for Nordstrom branded properties. Your upside is that once we prove the engine in a small-scale for Nordstrom they will want to adapt it for use across more their properties which could easily consume an additional 10 consultants to implement all the data plumbing. You would be the default source for these consultants.