Every once in a while I seem to get caught in a discussion with a customer / potential about what solution would be good for him. And sometimes we have different opinions.
Take this one, for example.
The customer wants a DMS platform, we analyzed and proposed EMC Documentum. The first business application he wants to deploy on this is an approval process for some forms. I can already hear everybody screaming: xCP!
Staring to read the RFP, from 30.000 feet it seems it’s a match for xCP, indeed. Digging into details… not anymore. You find features which are not possible in Forms Builder. You find features which are not possible in Process Engine. And you find that the approval process configuration is a matrix which in no way can be emulated or implemented reasonably enough with ootb stuff. One after another, this leads up to a significant custom development task. Development which sometimes will break the “rules of engagements” about TaskSpace (“you thou shall not code inside the UI, or else the wrath of the new version will come down to you in 6 to 12 months!”).
Add the fact that this will be used by top management and middle management only. And it’s business critical flow.
Well, my money is on custom development. Build a specific frontend and use the platform underneath. Forget about TaskSpace.
What? And then where is xCP useful? How about the promised enabler for “business agility”? It’s still there, it’s still ok. Just that it’s not the good answer to this business requirement. It simply isn’t. And as long the business requirements don’t change to take into account the technology.. there is not much we can do.
Of course it can be implemented in TaskSpace. But it will result in an ugly solution. Both to maintain and to use. It will not be something I’ll be proud of. Will I take the customer money if he orders me to do it in xCP? Yes, and I’ll do my best to give him the great solution he wishes. We will both learn something in the process, and I will even be happy if I learn I was wrong. But until now I was quite right on these things… as many seniors I tend to have a nose for them, not bragging.
Then why not Webtop, why go for the challenging custom development on UI?
Because doing the UI developments on top of Webtop will sooner or later require effort to port them to whatever new version of WDK or – more likely – new client technology EMC pushes forward. Imagine yourself changing the customizations you did on Webtop to CenterStage. A custom client will be more resilient on future changes. Build it on top of DFS and you have a really long term facing solution in front of you.
Now, let’s convince the customer. After he have seen a beautiful presentation of xCP and heard the market talk… how could you? “Trust me, I’ll be implementing this for you… I want the perfect thing for you… we’re in this together…”? Helps, but does not make it.
Show TaskSpace shortcomings at a technical level, related to the specific RFP details? Aren’t you shooting yourself (the vendor) in the foot? Well, with all the respect to the product, this needs to be done. But is it enough?
Incidentally enough, just before this customer I was in another one where I advocated the usage of TaskSpace. Why? Because is really good at what it does. But it does not do everything. And comparing with older BPM tools it still has some mountains to climb.
Custom developed UI is the answer sometimes (I reckon around 10% of ECM business applications). Because sometimes default clients simply can’t get it up.