Ask any VP of Ecommerce or Online Merchandising Manager what they are doing to increase conversion rate and revenue on their site and they are likely to tell you they are using the latest recommendation engine to generate a highly personalized shopping experience. I would say they are leaving money on the table.
Click on a pair of shoes and it will show you three other popular choices, a belt and a handbag to go along with it marioloncarek.com . Recommendation engines can be simple or quite sophisticated but the objective is the same; increase conversion rate and AOV (average order value).
So with all of this technology available to dynamically personalize the shopping experience for online shoppers, what place does a “Milk and Eggs” approach have in a cutting edge Ecommerce operation? Better yet, what is a “Milk and Eggs” approach?
This term is familiar to anyone with traditional merchandising experience. The concept is pervasive in the retail world and has stood the test of time for one simple reason; it works. You put the Milk and Eggs in the back of the store because you know the customer will find it regardless of location. Other products are strategically placed in route to and from the milk and eggs to increase exposure a.k.a selling opportunities which in-turn increases AOV.
The same principle applies to online merchandising but many people are missing the boat. How many sites to you see put the “Top Sellers” or “Most Popular” items at the top of the page or category? This is a self fulfilling prophecy. Your Top Sellers are in the top slots and they remain your Top Sellers as you would naturally expect. It’s like putting your Milk and Eggs at the check-out.
Imagine the following Ecommerce scenario in which 12 products fit on your category page. Product A is your most popular product so you put it in slot 1. It sells $100 in a given day and Product B which is less popular is in Slot 16 (page 2) and it sells $25 for a total of $125 in sales.
An optimization algorithm following the Milk and Eggs approach would put Product B in Slot 1 and Product A in Slot 8. Product A would sell $90 but Product B sells $60 for a total of $150 in sales which is a 20% increase in sales just by changing your slotting order.
The evolution of the slot machine industry from computerised machines to video slots and the increasing popularity of online casinos and mobile gambling open a world of opportunities for game designers to develop new types of game routines. In this new setting the potential of increased, even unlimited, player interaction arises. Game designers must embrace the change and develop innovative games that make the most of it. Attractive game ideas are needed that will catch the interest of the players by offering them new and exciting possibilities, yet not too new and different that the players are alienated. Succeeding in attracting player attention and interest is important because of the size of the industry. In the UK, for example, gambling makes a significant contribution to the economy with an estimated expenditure of £8,875 million (0.8% of UK GDP (Gambling Act, 2005)), of which £1.74 billion is made in costumer losses playing slot machines (Gaming Board, 2005).
Slot machine games need a corresponding mathematical model to make sure the game is profitable and to accurately calculate the minimum house edge. In those games that the player has an input, the player’s optimal strategy must be calculated in order to calculate the minimum house edge. Probability, Operational Research techniques and Stochastic Processes are used to build these models. Good programming skills are also necessary. If video slots are to allow games to be more interactive these games could become more sophisticated and, consequently, more difficult to model. This is a specialised job that only well trained professionals can do.