James Taylor's Decision Management

Don't forget to use decision management to improve enterprise applications

Thu, 03 Jul 2008 08:43:11 -0800

This report from Aberdeen is available on ebizQ this week - Aligning IT to Business Processes: How BPM is Complementing ERP - and it covers how IT and business groups can use Business Process Management to complement ERP and other Enterprise Applications. The report discusses the widespread concern in companies over the lack of ROI in their ERP projects and has a number of interesting statistics. For instance, only 15% say their ERP systems offer the flexibility they need while over half say they must perform manual steps to get things done. The report goes on to describe how companies are using business process management (BPM), SOA and web services to fix some of these problems. In particular it discusses how companies feel BPM gives them more agility, improves speed of implementation and gives business more control.
While I don't disagree with any of these, I do think that decision management is also important when thinking about how to maximize the value of existing enterprise applications. I have blogged before on why agility needs process and decision agility and of the role of decision management in composite applications and what Forrester calls Dynamic Business Applications. Building decision services and making these part of the new composite applications and agile business processes is important too. Interestingly more than half the responders to the report said said that "key application parameters and rules are locked in program source code, making changes difficult" - a clear cut case for decision management if ever I heard one.
Check out my guest post over on Vinnie's blog too on Raising your enterprise application IQ

Decision management, now more than ever.

Wed, 02 Jul 2008 18:10:42 -0800

Dave Kelly wrote a nice little article on BI as a Boon to Business: Now More Than Ever. In it he makes some great points, particularly about the value of understanding what is profitable and what is not as well as the value of being better able to understand, and therefore, target your customers.

to make it worth it, though, it must affect your results - you must drive what you learn into your operations and your operational systems. Decision management, the automation and improvement of critical decisions in your operational systems so that they are smarter and better informed, is the best way to get a competitive advantage from your data. So do as Dave says, but don't stop with BI - go all the way to decision management.

Here's another "If only "

Wed, 25 Jun 2008 18:12:23 -0800

Carey Schwaber from Forrester published a nice little post this week - The "If Onlies" Of App Dev. The list included a couple that I really like:

If only requirements were complete
If only software were easier to evolve

Me I think these are related - after all, if the software was easier to evolve it would not matter so much if the requirements were not stable - we could always change the software to match the changing requirements.
If only developers did not think everything had to be coded, perhaps we would not have these problems. If we recognized that requirements are not the same as business decisions and that decision making logic - business rules - should be created, managed and evolved differently from code we would not get into such a mess. Perhaps we could get business users to evolve their own applications as their needs change. If only.....

Are humans faster than computers?

Thu, 19 Jun 2008 16:05:22 -0800

A post titled "Humans are Faster than Computers" over on the Return Customer blog made me think. I completely see his point - far too much automation is dumb and irritating and simply gets in the way. But let's think about his scenario.

Firstly, the person he was asking knew the book he wanted. Had they not, increasingly likely with the ongoing explosion in customer choice, they might have had to go look it up on a computer anyway. Even so, perhaps the employees knowledge of the store layout would probably have given them an edge (though a well designed application could have mitigated this with directions, maps, perspectives etc). But his core point, that personal interactions have real value, is a true one and one that can be supported by computerization.

Decision management is key in both cases. Automating the decisions involved intelligently and supporting the decision making of staff who are interacting with customers really matters to customer service. Customers who self serve want to get stuff done - to take action - and thinking about the decisions involved makes this smoother. Understanding the decisions a staff member makes - offering a deal, upselling, cross-selling when something is out of stock - and supporting those intelligently can make every staff member seem experienced and knowledgeable.

Here's how to use decision management to empower mobile devices

Tue, 17 Jun 2008 15:19:58 -0800

I was asked a question about analytics on mobile devices today and it occurs to me that I don't see weary road warriors wanting to do "traditional" BI on a smartphone or other mobile devices. Some new decision management functions are needed to empower a smart phone platform.

One of the the differences between BI and decision management is the focus on taking action using insight gained from data rather than showing someone the data and helping them gain some insight. This means applications that use the information the phone has (position), insight from the data the company has (fraud likelihood, wait times) to take an action (dispatch the person with the phone to a particular place, tell them to do or not do something). I don't see traditional BI vendors having much to offer here - the whole reporting/OLAP infrastructure they have developed is predicated on knowledge workers doing analysis. If you want to take advantage of mobile devices you need to think about automating decisions for the person holding the device. For instance:

Use mobile phones held by maintenance engineers to track their location and then use analytics to predict which pieces of equipment are most likely to fail soon and rules to assign the nearest, qualified engineer before sending the directions on where to go to the engineers phone.
Don't show them reliability graphs or travel times, tell them where to go to make best use of their time
Use the mobile phone of a real estate appraiser to find out which risk zones a property is in and what the predicted difference is between a house inside and outside that risk zone
Don't show them a picture of the risk zones
Use a doctor's mobile phone to route them to the most useful hospital during an emergency based on predictions of patient load, the hospitals they know and their specialties
Don't show them graphs of wait times and pie charts of specialties needed
Use a customer's mobile phone to make them an offer at a store that is nearby having predicted that they are likely to buy it, checked that is in stock there and estimated that they are more likely to respond in person than to an email promotion to the website

And so on. Automate decisions and use mobile devices to provide context for those decisions and to deliver decisions to people out and about. Don't send them reports. Please.

Don't forget operational decisions for analytics

Fri, 13 Jun 2008 14:54:45 -0800

Kyle McNamara responded to a post of mine about analytics and how confusing it is as a word with one of this own What is Analytics?. Kyle had a nice definition that focused on competitive advantage and predicting outcomes and used a graphic from Competing on Analytics which I too find useful. The most useful thing was last, a graphic with an illustration of the use of analytics across the value chain.
Looking at this, however, it seems to me that few of the examples are the kind of operational decisions on which I focus. Here, then, are some additions by area of value chain:

R&D
This is one without much in the way of operational decision making
Production
Optimizing supply chain decisions is one but others like production scheduling and process control often require operational decision automation and analytics.
Marketing and Sales
One of the classic areas for operational decision making, optimal pricing and targeted/personalized offers are both operational decisions that use analytics. More generally a Best Next Offer or Next Best Action approach to sales/marketing/support requires strong analytics in operational decision making
Distribution
Route optimization is one if it is done dynamically. As RFID becomes more widespread, real-time decision making using the RFID information will become more important too as I noted here.
Customer Service
Kyle gets started with personalization but retention offers to specific customers as well as determining if a particular claim (warranty or otherwise) is valid require other complex operational decisions.
Finance
Not so many here but fraud detection as well as compliance with things like anti-money laundering decisions might go here.
HR
Very large organizations might automate decisions like scheduling

There are lots of ways to use analytics, don't forget the operational decisions when you think about them. Strategic/knowledge worker decisions can be easier to think of but they may not be the best or only ones for which analytics can help.

What RFID really needs to mean - action!

Thu, 12 Jun 2008 15:36:07 -0800

Michael Dortch had an interesting post this week What RFID Means, REALLY – Real-Time, Fully Integrated Data! in which he argues that the coming wave of RFID data will deliver:

real-time, fully integrated data. The more you can know about what's going on at the edge of your network the closer to when it's actually going on, the more opportunities you have to inform, refine and optimize business processes

Now this is all true - RFID information from the edge of your organization will create an opportunity to change your business processes at the operational interface - down at the sharp end. The volume of data, and the speed at which it arrives, are critical however. Streaming RFID information onto a dashboard or reporting on it is simply not going to get it done. So I start to disagree with Michael where he talks about the need to

minimize "time to information"

This sounds a lot like "Those who know first, win" and I think this is misguided. What good does knowing something before your competitors do if you don't act on? What if you know first but take so much longer to decide what to do than your competitors that they respond before you do? If you cannot reduce the time from when you could know what to do to the point where you actually do something then you are not going to win. I wrote about this once before in the context of Richard Hackathorn's work on latency. So I would say

Those who act first, win (provided the action is smart enough)

If you want to take advantage of RFID then build systems that will take decisions based on that RFID data and ACT!

Decision management and event based marketing

Wed, 04 Jun 2008 14:12:18 -0800

Jim Berkowitz had a post on event-based marketing that led me to this piece by BJ Morgan from Unica. The posts talk about the need to monitor individual customer behavior, the ability to explicitly take time into account, openness, ease of use, and speed. The value of decision management in delivering event-based marketing is pretty compelling:

You must make a decision that is focused on a single customer and do so often. This means lots of micro-decisions and so implies both automation and ongoing management of those decisions.
Data mining and predictive analytic techniques are ideal for turning time-based data into usable information - they turn uncertainty about the future into probabilities allowing you to act despite the uncertainty.
Deployment of decisions as Decision Services helps ensure that multiple channels and systems can deliver the same decision, improving openness
The use of business rules to manage decision logic makes it possible for business users to manage the details of the marketing decision, crucial to ease of use
Automation of decisions means they can be made in a very short time, crucial in any event-based approach.

Indeed event-based marketing is one of the major drivers behind companies adopting decision management.

Decision management makes BI come to (SOA) life

Tue, 03 Jun 2008 09:03:15 -0800

Mike Kavitz had an interesting post on Integrating BI and SOA that made me think about BI in SOA. Personally I have never much liked discussing BI in SOA as BI's focus on reporting and presentation to people seemed a more fit for SOA's focus on interconnected services with well-defined interfaces. Mike's post was nicely done, though, and made me see his point of view clearly. His platform nicely externalizes and manages the business rules as well as the data and business intelligence services. If, instead of a business rules layer, he had a decision management layer I think he would be in even better shape. Instead of separating out decision making and saying that business rules OR business intelligence can be used, a decision management layer would support the use of data, information and rules in whatever combination made sense. Thus some decisions would be completely automated using rules and executable predictive analytic models, some would be automated using rules derived from data mining, others would be supported using reports and rules in combination to make suggestions while others would be left entirely to people with BI being used to support them. Lots of flexibility, a nice focus on decisions and even more value from integrating BI and SOA.

Decision management does not make programmers obsolete

Mon, 02 Jun 2008 13:44:51 -0800

Eric Roch had a great post last week on SOA, Reuse and Programmer Obsolesce. Like Eric I believe that dynamic, agile applications will require more than just SOA but will use technologies like rich interface tools, Business Process Management, Decision Management (rules and analytics) and event correlation technology. The standardized, componentized approach represented by SOA is crucial as it allows components built with different technologies to interact easily and be composited into new applications. Similarly I agree that this will not mean the obsolescence of programmers in general. It might make it hard for "old school, give me the requirements and go away until I have finished coding" types to get work but it won't eliminate the need for programmers.
Even decision management technologies, which are primarily designed to manage business logic, do not make programmers obsolete. Not only is some business logic very complex (meaning that a programmer is likely still to "own" the definition even if a declarative approach and a business rules management system are used), but the object model and other "bones" of an application must be hidden if business users are to successfully manage decision logic. The purpose of decision management technology is to allow business users to manage decisions, rather than programmers, but this is mostly about improving collaboration and integration not replacing one with the other.

Come to the Intalio conference and get the book free!

Thu, 29 May 2008 15:25:34 -0800

Neil and I are going to be speaking (on Applying Decision Management to Make Processes Smarter, Simpler and More Agile) at the Intalio User Conference on June 17th and the nice folks at Intalio have decided that everyone who attends deserves a copy of the book. While I am sure that all the attendees are smart enough to already have a copy, this will ensure that they have a second copy to lend to people, which is terrific.

You can find out more about the Intalio user conference or register. If you did not already know, Intalio offers open source BPM and SOA software.

Using decision management for self-service and good service

Fri, 23 May 2008 09:41:34 -0800

Ginger Conlon wrote an interesting piece this week - The Real Reason We Love Self-Service in which she wondered if the real reason people like self-service options is that it avoid bad service from people. I have to say I think she has a point but I also think people prefer the control of a self-service environment. Unlike Ginger, I would always prefer a kiosk that got it done than a personal interaction just because I control the environment and don't have to use up interaction energy. But what does this have to do with decision management?
One of the ways companies can make self-service work better is by implementing decision management technology to ensure that customers can get decisions made without referral to a person - check out this post, for instance. But decision management can also be used to empower and support customer service reps - they don't have to bounce you to someone else or refer you to their manager or make decisions about how valuable a customer you are. A decision management system can do it for them.
Will this eliminate bad attitudes, long shifts or other elements of bad service? No, but it will help some.

TDWI Executive Summit

Wed, 21 May 2008 08:00:52 -0800

The TDWI Executive Summit takes place this August in San Diego. It is titled "Be a Hero: Unleash the Value of BI". My business partner and co-author Neil Raden is one of the keynotes - Reinventing BI with Real-Time Analytics: Embedding Analytics into Business Processes - along with Howard Dresner (ex Gartner and Hyperion) and Wayne Eckerson (of TDWI) among others. It looks like it would be a really interesting session for anyone looking to transition from BI-as-reporting to using data to improve decision making at every level. You can register here.

Best Buy and "micro" marketing

Tue, 20 May 2008 07:50:39 -0800

I have blogged about Best Buy and it's program for customer-centricity at various times and I was reminded of the program by an article in Business Week this week - "At Best Buy, Marketing Goes Micro". This particular article was about how Best Buy is trying to get more input from it's local stores so that they can customize their behavior not based solely on central analysis but on local expertise. While at first sight this may look like a case of manual decision making taking over from automated, the truth is more complex.

The use of analytics in retail is growing fast and Best Buy is one of the primary proponents of a more analytic approach to their stores as well as their marketing. This has been very successful and it is the existence of a data-driven, explicit decision making process that makes it possible for a big retailer to bring in local know-how. Without the existing analytics and automation, no what-if would be possible and the corporation would not be able to assess the impact of these local initiatives nor compare them with any kind of real benchmark. The local staff are providing their expertise and overriding standard options today but it seems likely that in the future the decisions will be automated in a way that brings that expertise in more directly. Only once that is the process can you really balance the local experience with the multi-channel reality of a modern retailer. Customer segmentation, predictive analytics about customer buying patterns, analytically-driven personas can all help make a better store and a much better online/direct mail environment. Bringing local experts into the picture can only help.

Some FAQ on decision management

Mon, 12 May 2008 16:42:44 -0800

After running an initial article on decision management - Business Decision Management - Part 1 - the BPM Institute published an FAQ by me and my business partner (Neil Raden) as Business Decision Management - Part 2. Together these are a pretty good overview of decision management. Enjoy.