When I posted about software testing, I never imagined that I would be so blatantly in violation of good testing principles: I added Google ads that were slightly too wide for my right sidebar, and on Internet Explorer, they made the entire sidebar shift to the bottom of the page after all the content. On Firefox, which I use, it looks fine. My only excuse is that it’s been so long since I gave up coding for pure analysis and design that I’ve become lazy about cross-platform testing, which is a pretty poor excuse.
Note to self: always test template changes on both Firefox and IE.
My formerly bandwidth-challenged friends now have wireless broadband in the house, so I’m online in London and loving it. More tomorrow after BPM 2005 day 1.
Thomas over at the Workflows – Collaboration Technology blog gave me a new title this week: BPM guru.
Although I recently changed my title to read “Strategic IT Planner/Architect” after being dissed for using the C-word, I think I like “BPM guru” better.
I’m headed for the BPMG conference in London. I’ll try to blog some daily notes about the conference; I’m staying with bandwidth-challenged friends, but there is wireless connectivity at the conference site.
At most conferences, presentation materials are handed out on CD, which is my preferred method of receiving them since my paper filing space is reserved for things that can’t be easily scanned. As an added bonus, BPMG has made most of these materials available online to attendees ahead of time: I can download and review them at my leisure (such as on the flight over), or even print them to take to the conference if I decide not to carry my laptop every day.
If you’re at the conference, look me up.
The Gartner webinar that I dropped in on yesterday had some interesting points about modelling and methodologies that started me thinking.
First, on methodologies: it’s absolutely essential to have some best practices to lend structure to your BPM project. Don’t do this alone, get the help of someone like me (okay, it doesn’t have to be me) who has actually implemented BPM projects before. Whenever you change a business process, there’s a whole lot more than just technology going on, and you don’t want to get caught in the classic IT trap of believing that the business users will be just as excited about the new technology as you are (remember, they didn’t get to play with the Java code).
There were comments in yesterday’s webinar about how the soft benefits are becoming more significant, including internal factors such as real-time business agility and a process-focussed culture. However, you can’t expect your organization to change because of the introduction of BPM technology; instead, your organization needs to make cultural changes driven by business factors and enabled by the technology.
On modelling tools, I made the statement last month that most people are using Visio to model their business processes before they are implemented in a BPM system, which is true. However, just because it’s true doesn’t mean that it’s the best way to do this. If you use a standard modelling notation such as BPMN or UML activity diagrams, you’ll do fine up to a point with Visio, but somewhere along the way to your “to be” process, you’re going to need a more serious tool for process simulation and the like. If you check out the BPtrends report on modelling tools that I reviewed last week, you’ll see a lot more tools with a lot more power than Visio for your process modelling and analysis. You’re not going to put these on everyone’s desktop, but they are needed for a few analysts who will be doing the in-depth process design.
I attended the The BPM Momentum: What’s Driving it? webinar today (not “what’s driving IT” as the host ebizQ erroneously labelled it, which has a much different meaning), featuring Jim Sinur of Gartner. Definitely worth catching the replay for Mr. Sinur’s comments on success factors for BPM projects and for his view of the market convergence.
He included Gartner’s Application Integration Hype cycle chart, showing how BPM technologies fit: apparently, BPM itself is sliding into the Trough of Disillusionment, whereas BPM Suites are still climbing towards the Peak of Inflated Expectations. (The hype cycle terminology always reminds me of the morality fable Pilgrim’s Progress with its Slough of Despond, but I digress.) He only put a 10% probability on the catastrophic scenario, which makes me feel a whole lot better.
He also had some good numbers on customers and their BPM projects:
- 85% of BPM customers are now going after their human-facing processes (presumably, the automated system-to-system ones are already in place).
- BPM projects are yielding an average IRR of 20% (although Gartner uses a more conservative figure of 15%), but larger projects can produce returns of well over 100%.
- “Soft” benefits such as competitive advantage and higher customer satisfaction are major contributors to a project’s success.
- Business and IT are becoming more aligned on BPM projects.
He also commented on the convergence that is happening in the marketplace, something that I’ve been seeing for some time as well: 130 BPM vendors all attempting to jostle their way into what I call the “BPM suite spot”:
This convergence is just a continuation of the evolution of BPM that I discussed in an earlier post, but it’s going to get a lot more painful for some of the players as they get eaten by the competition or body-checked off the playing field.
A few weeks ago, I was involved in making a proposal to a bank for a consulting gig related to business intelligence. Although we had an inside track on what they wanted, and we proposed a team that could deliver the goods, they decided to go with one of the big management consulting firms, in part because they felt that the larger firm presented a lower risk.
I fought against this prejudice for years when I ran a small (40-person) systems integration company, and were shut out of some opportunities because of our size. Customers who did hire us realized that small is good: we tended to attract a better quality of team member, keep them well-trained, and motivate them appropriately through the right combination of profit-sharing and foozball tables. From the customer standpoint, they held a great deal more leverage over us than they would over a larger SI because they were a larger percentage of our business, and we were very motivated to make them happy, repeat customers who could be used as references. Our reputation was our main marketing asset, and the only way to make that work was to out-perform the big guys.
These days, given the abysmal track record of a couple of the big firms, I find it hard to believe that the myth of “big company = low risk” is still alive and well. The last time that I was involved in a project with one of the big SIs, it was more than a year late (at last check) and I’m assuming that it’s also way over budget. Also consider some of the recent layoffs announced at big companies, including a whopping 13,000 at IBM. Do you think that they considered the impact on your project before they laid off a staff member who works with you, or supports the project in some way?
Risky business, hiring these big companies.
OASIS announced yesterday the creation of a technical committee to develop a core reference model for Service Oriented Architecture (SOA). To quote their goal:
The SOA reference model will offer an understanding of the core elements within a service oriented environment and the associations and relationships among those elements. The reference model itself will not be directly tied to any standards, technologies or other concrete implementation details. Rather, it will be an abstract, designed to be used as a tool by software and enterprise architects developing specific SOAs.
The end-user organizations seem to be strongly behind this, given the participation of Boeing, General Motors, Lockheed Martin, Mitre, VISA, USA’s Department of Homeland Security, Japan’s Electronic Commerce Promotion Council, and Canada’s Public Works and Government Services. However, The Register reports today that some of the biggest names in enterprise computing have not yet hopped on board: Microsoft and Oracle, to name two. Shame, shame; you guys aren’t planning to go the proprietary route, are you?
I’ve been putting together some ideas for a presentation at FileNet‘s annual user conference, UserNet, on how to make the link between business process models (such as those that form part of an enterprise architecture set, modelled with a notation such as BPMN) and the implementation of those models in a specific product (in this case, FileNet BPM). I’m interested in exploring the BPMN/BPEL link between modelling and implementation using various products, since it represents the transformation from Zachman’s Row 2 to Row 3; however, for UserNet, where BPEL is not yet in play, I’d choose to focus on the conceptual (business) view of how this works specifically with FileNet BPM.
I’ve spoken at UserNet many times, first as the chief architect of a FileNet partner, then later as FileNet’s Director of eBusiness Evangelism, and I always enjoy it because of the focus on user/business issues instead of just technology, and the interaction with business users of BPM products. But today, as I was looking at the conference dates to make sure that my calendar is clear, I had a flashback to my earlier post on vendors charging exorbitant rates to their “partners” for training. Sure enough, according to the conference presentation guidelines:
One customer presenter per presentation will receive a free registration but they must be the primary speaker. Partner presenters will register for the regular conference fee.
Excuse me, I thought that I just read that FileNet wants me to spend my time creating an educational presentation that will be of benefit to their user community, but then wants me to pay to deliver it.
Just like with the training issue, this cuts out a lot of the highly-skilled smaller partners, because the conference fee on top of a week of lost revenue, travel expenses and the time invested in creating the presentation just isn’t worth it. Instead, count on seeing presentations submitted from partners based on their marketing budgets rather than their abilities.
The attendees should be hoping that the conference selection committee uses a more appropriate assessment tool than a partner’s amount of loose change.
Almost missed the date: the BPMN 1.0 specification is one year old today. Check out the list of 23 product vendors that already support it, plus a couple of user organization testimonials.
Business Integration Journal (formerly EAI Journal) now makes their magazine available via free email subscription, for those of us who are not in the U.S. and therefore not eligible for a free paper subscription. A lot of the content is “advertorial” written by vendors, but there are a few gems in there, such as this month’s article on process-centric business intelligence by Keith Gile at Forrester Research, wherein he tackles the problem of out-of-context BI data by looking at ways for BI platforms to associate data with processes in order to deliver decision-making capability to the operational level.
Given BIJ’s policy of not publishing PDF copies of the current edition’s articles on their website until the next month’s edition is available, this e-subscription is the only way to get the content in electronic format at the time of publication, and I actually prefer an electronic copy of these “read and toss” magazines anyway. I think that I was sent an invitation for the subscription, but have no idea why; poke around on their site and you’ll probably find something. They also have issues back to January 2000 online, some of which are really a blast from the past.
My only complaint is that the e-subscription issue is hosted on Olive Software’s “ActiveMagazine”, which is really not a nice way to read online. It also doesn’t produce very readable copies if I print to PDF, so if I want to save a copy of an article or send it to a client, I have to either put up with the poor quality or wait until next month for it to come available on the BIJ website.