A curious reader asks…

One of my favorite things to do on my blog is respond to questions from readers. To make it easy to find these going forward I have added a new category – Reader Questions. Email me (james at smartenoughsystems.com) if you have a question you’d like me to answer on the blog. Anyway, this week I got an email from Richard Ordowich with some interesting thoughts and questions and, with his permission, I am going to answer them here for everyone to share. His comments are shown as blockquotes, everything else is me.

James, I have read about ½ of your book and I would agree with most of the concepts you put forth offer some benefit. As I’m sure you’re aware, implementing these concepts in legacy environments is the greatest challenge. Business policies, processes and rules are ambiguous, conflicting, inconsistent and subject to change without notice.

First of all Richard gets 100 points for buying and reading the book! Any of you who have not yet done so can take advantage of a discount on amazon.com here or buy it direct from Prentice Hall here (also at a discount). I’ll pause for a moment while you all do that.

Getting from a messy current state to a future state where decisions are managed and improved systematically is, indeed, the hardest part of adopting EDM. I would argue it is the hardest part of adopting any new approach or technology. EDM offers real hope to organizations burdened by legacy systems precisely because it does not require them to be replaced wholesale. By identifying decisions, externalizing them as decision services (wiki) and managing the rules within these services using a business rules management system (wiki), an organization can take control of its policies and procedures. Not only are the externalized decisions and rules easier to manage, the rest of the legacy application requires less maintenance as many if not most of the constant changes needed by applications are around decisions and decision making. The three problems Richard identifies also map to the benefits we identify for EDM – precision (making more accurate decisions) to replace ambiguity and conflict, consistency (the same decision in multiple systems and processes) to replace inconsistency, and agility (ease of identifying the correct change to make and then of making it) to handle constant change.

Establishing the discipline to manage the business following a rigorous methodology when the organization has been operating based on mythology is difficult. To me it’s analogous to getting organizations to adopt a software development methodology. The results have not been exactly exemplary. Projects continue to be late, over budget and don’t meet the businesses needs. So why would the business adopt EDM?

Richard is clearly a hardened realist. Adopting EDM does mean taking a more rigorous and less mythological approach and so has some of the challenges he identifies. Personally I think that the persistent confusion of rules and requirements (wiki) is a major cause of project failure and EDM would, at least, address that. In all seriousness EDM will take a while to adopt and should be done in stage, that’s why there is a whole chapter on this in the book (wiki).

Also, I think your ROI model is inverted. Business folks are most concerned about revenue. I suggest the following represents the reality of business:

If EDM can be shown to increase revenue, then you have a better chance of getting senior management buy in. Cost Reductions and Strategic Control are “nice to have”.

Now Richard is quite correct here. For those of you have not read the book (and if you haven’t ordered it yet, please do so. Someone you know must need it for Christmas), this pyramid is presented the other way up. This is one of those times where I have to say that we were trying to show a sort of “Maslow’s Hierarchy” with cost reduction being the most basic, revenue growth being next and strategic control or executive empowerment being top. Richard’s point that revenue growth is top of mind is quite correct, however. One of the reasons for the three core benefits of EDM being precision, consistency and agility while the two minor ones are speed and cost is precisely to emphasize this. Precision, consistency and agility of decision making can grow revenue and so should be the primary focus. Someone who only thinks about speed and cost is going to do workflow and other efficiency-oriented things, not EDM.

I think the undisciplined way organizations implement business rules makes it overly onerous to consider having a repository of business rules. Some like credit scores and risk analysis for fraud can be “centralized” but I suggest that once you attempt to unravel the convoluted business rules embedded in systems and manual processes you end up chasing your tail. I think this is one of the main reasons organizations buy ERP packages. All the rules are predefined for standard business processes that do no offer competitive advantage (sales, manufacturing, HR, finance etc.). With SAP and Oracle offering ERP as a SaaS, medium and small organizations will begin to use these solutions as well. Then you don’t have to worry about rules. Anyway, when has an organization succeeded because it had a better HR system than it’s competitors?

This is one of those “not all rules are rules” kind of moments. It is true that trying to externalize all the rules in a system or even in an organization is probably futile. Where you have a decision you want and need to automate, however, it is often extremely practical and rarely impossible to find all the rules that are relevant. Using the ability of a BRMS to allow rules to be based on templates (so they can be rapidly and accurately created), partially exposed to business users (so they can make the most common modifications directly) and organized into layers that complement or override as necessary and you can do this. The rules for a standard business process or in a standard package are often not decision-centric but process or flow or data-centric. In most standard packages and processes the decision making is left to people. EDM is about taking these dumb applications and mindless processes and making them “smart enough” by hooking them up to appropriately automated decisions.

Business intelligence and analytics is another topic in your book. Once again an important discipline for an organization to adopt. But mostly it appears that the intelligence is lacking between the chair and the keyboard. Data warehouses are built and then people try to figure out what to do with them, generating interesting analysis. In my opinion most of the intelligence derived from data warehouses is really artificial to begin with. It’s most a figment of people’s imagination and helps perpetuate the business’s mythology!

Quite agree. Just like it makes no sense to take all the rules you can find and then see what you can do with them, it makes no sense to store all the data you have and then see what you can do. Much more effective to find the decisions about which you care and then try and see what data you have (or can buy from someone else) that might make that decision more effective. Could your data help you put a customer into a meaningful segment? Could it help you predict retention risk for a customer? Can you estimate the likely future value of a customer? These kinds of questions can be answered by analytics but notice that they typically relate both to a single customer or transaction and to the future. This is the essence of the kind of analytics EDM needs – focused on helping with a specific decision for a specific customer or transaction and focused on the future not the past.

So, Richard, how did I do?