A lot of attention is being given to BI or Business Intelligence these days. There are new BI products, BI conferences like the recent Dynamics Communities event in Dallas, and lots of people talking about BI. But when you ask someone what it is, you get a variety of answers that generally revolve around some form of reporting and dash-boarding. Business Intelligence needs to be more.
For decades, the computer industry has touted the “paperless society”. But as we stored more and more information in the tables of applications, we became better and better at producing reports. Why with all of that data stored, it was easy to produce a list of sales by customer, identifying the best customers, and then sales by items, identifying the best items, and so on, and on, and on…..consuming more paper with each variation.
But, to focus this conversation on sales for a few minutes, sales managers like to know who is buying and what is selling and which rep is selling those products. And, those reports show trends. If a top ten customer this month becomes a bottom ten client next month, something happened! A sales manager wants to know why. Who stole the client? Why did they stop purchasing? So each week, the sales manager gets a new “Top Ten” report and is expected to review it and check out the issues and problems found.
The problem with this scheme is the volume of reports provided. Yes, they all contain valuable information but someone needs to review them. Every day or every week, a new batch of reports can be printed, distributed, and reviewed. Until data numbing sets in. Week after week of reviewing the same data and something more important happens one day and a day is skipped. Nothing exploded so the next time 2 or 3 days of reports get skipped, then a week, or two, until something happens and everyone starts back-peddling.
Using the correct tools, a BI team can replace the daily/weekly reports with an analysis tool that sits quietly watching the data and only presents the user with something to review when a problem occurs. Let’s take AR as an example. Most firms have an established historical set of numbers indicating what percentage of receivables should be current, 30-60 days late, 60-90 days late, and later. Certainly the over 90 day accounts need to be addressed and decisions made on their credit levels. This, however, can be a month end exercise. What needs to be watched is the trending of account ageing. When too many accounts move from current to 30-60, someone may want action to be taken. Does the AR clerk need to pull the ageing report every day and examine the percentages?
No, a business alert can be built that looks at the open amounts aged and, when the percentages shift, send an alert, email, text message, or some notification to the correct team members. THEN, the ageing report gets pulled and an analysis made. It may not be the customer’s fault. It may be that a rash of bad material went out that the firm is correcting. It may be that the industry is tanking and overall adjustments on the credit policies need to be reviewed. But the point is, an automated system can watch the indicators, freeing the user for the daily doldrums of reading a repetitive report and providing timely alerts. And only when actions are needed are the key management people involved.
So how does a firm implement this BI monitoring system? Look first at the entire process of monitoring, analyzing, and responding to events. Then, and only then, build the appropriate tools to enable the process. The Top Gun School (the actual school, not the movie) teaches a 4 step process: Occurrence, Orientation, Decision, Action. Occurrence is an event that triggers a need for action. Orientation involves pulling the needed data then and only then that allows the team to make the needed Decisions. With the information in hand and decisions made, Actions can be taken.
I have talked above about the AR monitoring process. This same process applies to many different facets of a business and depends on the type of business. Sales managers will have their own Key Performance Indicators. Production Teams will have their own KPIs. Purchasing even needs to know when orders are running late. The difference in this approach is, rather than drowning managers in reports to read or gauges to watch, identify the limits, the triggers, the events that indicate an issue and start notifying those manager when an event occurs. Follow this with the availability of reports or queries that can provide information needed for the decision making process. Identify in advance who will be making the decisions and, generally, that same person will drive the actions.
Which tools are best for a BI implementation depend on the needs of the firm, the data points that need to be watched, the skills of the firm’s IT team, and how key people need to be notified. In many cases, the tools exist with in the current system. MS Dynamics GP, for example contains a Business Alerts system. Add to this some custom queries or views and the alerts and notifications can be made. Add to this a layer of detail reporting and a good BI system exists. In some cases, special BI software may be required. No one tool can serve all requirements or needs, it is up to the organization to define their needs and the IT team to determine how to answer those needs and what tools to use.
No matter what tools are selected, the BI implentation process is the same. And it is a constant and on-going process.
First, meet with key people that have a need to monitor events and take corrective actions. Identify what those events are and what the limits are. Then monitoring systems can be designed.
Second, identify the information needed to resolve the issues. With this data, the appropriate reports can be identified and deployed.
The last two parts of the action/reaction plan are up to the key people themselves. They must react to the alerts, review the data, make the decissions on the appropriate responses, and take or schedule the appropriate actions.
With an approach like this, the data numbing that occurs from daily and weekly reports that never get read is easily replaced by a system that says “Wake UP! WE have a problem” and starts the analysis process. Business managers can now focus on the important jobs of managing the business and leave the reading of the data numbing reports to the system!