What is our primary use case?
BPM is very horizontal solution, so it can be used across any kind of industry. Financial services is the largest one - banking and insurance would be the biggest sector for BPM. But really, it's anything where there's a situation where a request comes in the door and has to be serviced, or at least that use case fits most of the time. But my biggest project at the moment is telecom infrastructure for mobile phone tower installation of antennas.
In terms of how our clients use it for workflow, first of all they're just going to implement what the process is. It's very similar to any kind of BPR methodology, mapping out the process - the "as-is" process - and then using that process to identify what the "to-be" process should be, how it should be different. Sometimes, it's even just implementing things using automation with their "as-is," having automation so that they can actually analyze how it performs, and then using that to make decisions on how it should be improved. Then, in an iterative fashion, going ahead and making those improvements.
For our clients, I'd say the more common case of using it in conjunction with other IBM automation products would be the ECM, the content management, because a lot of business processes will be delayed with documents. So sometimes they will use the one that comes with BPM itself, but for customers that really want to focus on the documents, they will have IBM FileNet, the document management system for that, and that's usually fairly tightly compatible with BPM.
What is most valuable?
The most valuable thing is actually the Analytics, which is actually a little bit not where it should be.
What needs improvement?
I'm hearing things might be improving, to really deliver on BPM as opposed to simply workflow. That really should be emphasized a lot more than it has been, because a lot of customers will simply implement the process and leave it there, because the product maybe doesn't emphasize BPM as much as it should, as much as maybe they talk about it in the sales process.
The whole idea of BPM, is to iteratively improve the process, and in order to do that you have to have the Analytics tool with it. A lot of times that doesn't go as far as it should simply because there's a lot more work to be done for that to happen, and just some sort of technical limitations that don't make that as easy as it should be.
I think they are working on it. I'm looking forward to seeing how much of an improvement it will be. They did announce that they have some stuff on the way for this, so I'm hoping it will be good. I think that at the moment, BPM as an industry in general, and IBM in particular, is not differentiating itself enough. I think the hype for it is down a bit, and I think that is a key way that they can reinvigorate the whole thing, by going back to emphasizing the entire cycle and fitting in with business improvement, as opposed to just being a technology with a layer type of thing.
So, that is, in a way, the most important aspect of it, for it to be BPM, as opposed to simply another way to implement a software solution that could be the same as anything else.
What do I think about the scalability of the solution?
I would say it's very good but it's not perfect. It is much more scalable than it has been in the past but there are some things, it does require some work to keep it stable. So that is an area that should be improved, actually.
The engine itself tends to accumulate a lot of data that needs to be cleaned up, and that's the kind of thing that keeps it from, in some scenarios, scaling as much as it needs to. And then, when you're building solutions, if you're not careful to keep the screens from being associated with too much data, if you're going to just do things the way that a lot of people would just assume that they can do, without having experience of having made those mistakes before, it will accumulate a lot of data, and that will cause it to perform very badly.
It would be great if you didn't have to worry about that, but the reality is, at the moment, that you have to pay attention to that. If you do, then you can do just fine. But if you pay attention only to the business requirements, and just throw in everything that the business imagines that it wants see, you can easily build something that is way too heavy on the front end. But if you handle those things, if you tune the underlying platform, it can achieve the target, sometimes with more hardware than you want.
In terms of ROI and scaling use of the product, I don't have a client where I can say that we specifically measured that as much as you would say we should, but my feeling is yes, they do see return on investment.
How is customer service and technical support?
Usually it's the client is using their account. The initial technical support is fine. When there are escalations, and their Tiger Team comes in, they are excellent.
Which solutions did we use previously?
My clients that go with IBM usually do so because they have a sense of, if they have problems they have an organization like IBM to complain to.
If we try to use open-source, that's usually a problem if it's a relatively big company, something like banking or insurance, they're not going to want that. But the manufacturing companies, they tend not to care as much, and as long as it works they're fine. For banking, they want something that looks a bit like the big-time.
How was the initial setup?
It has improved a lot. They built some tooling for the BPM platform in particular, in the releases that started about three years ago; they streamlined that quite a bit, it is quite a bit better.
It is still a massive installer compared to some of the heritage software that it was built from. It's quite large and it can take quite a bit of time to install. So that definitely could be better, but it is what it is, it's an IBM product.
Regarding working directly with IBM in the setup, if it's not an IBM project, usually I don't, because I'd have to pay. If it's actually an IBM customer then I would. And I do, sometimes, informally talk to some of the people that I know about what is the current latest and greatest way to handle certain things, but not so much in formal channels if it's not an IBM-direct engagement.
Which other solutions did I evaluate?
In the BPM space, the undisputed leader is a company called Pega. In terms of the market share, they win a lot more deals, but they have a very different approach, and they have different types of customers.
In the region where I work, Oracle is doing pretty well, although their software is not as good by any stretch. Their services organization is a bit more respected for some reason in the region. So they have won some things, although they struggle to actually deliver.
Globally, I think the other one that I find interesting is called Appian, and Appian does quite well. I think that they have more of a lightweight cloud solution that's a bit easier to go to. They have a strange sort of closed mindset, you can't read their documentation unless you have an account with them, this kind of thing.
And then there are the open-source vendors, which used to not really be considered much, but they actually have started to do very well, for certain types of scenarios. Among them are Camunda and Activiti, and another one called Bonita, although I'm not really sure - its heritage is definitely open-source. But all of those open-source solutions have an enterprise version. Their approach to marketing is open-source, but they'll still sell you an enterprise license.
What other advice do I have?
I rate IBM BPM a seven out of 10. It's good but it needs some help.
Disclosure: My company has a business relationship with this vendor other than being a customer: Partner.
Mar 28 2018