Complex analytics are geared toward complex decision making for unstructured problems or opportunities. But what makes decisions complex? The answer lies in the roots of any decision support system: data. This may sound easy at first, but data issues for complex decisions are indeed complex because the data we need is initially non-existent. Over the last decade, we’ve been building data warehouses and our rule of thumb has been: Our solution is only as good as the data. Now, we want to build a solution without the data. This may sound a bit too ambiguous at first. The data we need is non-existent at the time we’re faced with a complex decision. Think about the decision to hire the next senior executive or open a new store, expand in a new territory, set the inventory levels accurately or identify the optimum price of our products which would maximize our earnings or increase our market share. Or, it might just be the decision to buy a house.
These are all unstructured decisions because:
- There are too many components of the equation,
- There are too many alternatives, and
- There is very little data.
These three factors are what we face as soon as we start looking for a solution to our complex problem or opportunity. What do we usually do? We start analyzing what we have. We conduct meetings; look at our competitors and their activity, and maybe talk to some of the competitors to get a shortcut into the decision process. Oftentimes, the decision that we make comes from a gut feeling or a dominant player in the team. We simply try to make the decision with what little data we have. This is exactly where the new concept of complex analytics comes into play.
What is the first step? It is being proactive about the data. When buying a house, the buyer and the seller go through a bargaining process. The buyer first makes an offer and the seller makes a counteroffer. By making an offer, the buyer, in fact, produces useful data. When the seller makes the counteroffer, we have the more useful data. Maybe at that point the buyer will make another offer or maybe there will be an impasse. Simply put, there is proactive data in this interaction. The data is extremely useful for both parties because now they both have a better understanding of what the market price of the house is. If the buyer or the seller were to conduct the negotiation process with more buyers or sellers, the data collected will be even more accurate. That data represents what the market conditions are at the time the parties are faced with the complex decision of buying and selling. Before this process, both the buyer and the seller have opinions about the price of the house based on the historical data and that certainly is a key component of the equation. By going through the bargaining process, the buyer and the seller collect proactive data and incorporate it in the decision-making process.
When we are faced with a complex decision, we must go through the process of collecting proactive data. We do this by asking a question of the marketplace and recording the answer. Then we modify our model parameters which represent our current situation and goals. For the home buying process, that model may be how much we can afford, how long we want to keep the house, what neighborhood we want to live, etc. For a pricing decision, the model may represent cost and revenue figures. We price our products or services based on historical data that we collect about the market. Since all the competitors do that as well, we need to do something better to maintain or gain competitive advantage. Going back to the original question: How do we obtain the complex data that we do not have initially when we model our decision?
Figure 1: Complex Analytics Data Flow
In building the business model, we make assumptions. The data for which we make assumptions is the data we need to collect proactively. We load the data into our data warehouse so that we can query the new data along with the old data. From our data warehouse, we extract the data that goes in the business model. We continue the loop until we reach an optimum solution. In doing that we need solid methodologies, some new tools and all of our basic tools.
The competitors in the business world have implemented all existing data warehousing technologies today. The answer we’re all looking for especially nowadays when the economic waters are rough is how to stay in business and grow our existing business. And that answer lies in the concept of complex analytics.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access