What I Learned: Open Source and PLM - Compatible?

Hi JP, In the PLM implementation at my last company we worked
pretty much as you described. We developed our goals up front with
executive sponsorship, had a cross-functional team develop a spec,
an RFP was put out, then vendors responded. We made a decision,
implmented our spec (and worked very hard to keep scope creep
out!), and put it into production. It went into production 5 years
ago, and is fully deployed and well utilized today. Some updates
and improvements were made but, it is basically running as
originally designed. The point of that story, is it supports your
points, that PLM can implemented without the expensive zigzagging,
and everything developed in the implmentation could be ported to a
different application. It's much easier to implement something from
a reference. I think your strategy for piloting on Open Source is a
good one. It should help reduce the upfront costs, and improve the
success rate if you do end up needing to spend capital for a
proprietary solution. Best Regards Tom Gill www.aras.com

I have been involved in a number of PLM implementations with
different solutions and I would agree that the ratio of total cost
to software cost is closer to 10:1. However, I think a large
portion of the total cost can be attributed to research and
education on PLM in general, and is not specific to any one vendor.
So I'll pose this question to the experts: how much money would a
company really be wasting if they started with an open source
solution and found out later that it didn't fit their needs? I have
personally seen millions of dollars wasted on
evaluations/benchmarks of competing solutions based on vague
notions of what PLM is and how it can be implemented for a specific
company. A vendor choice is then made, licenses are paid for, and a
bunch of consultants show up to help guide the company through the
implementation. Almost inevitably, the company ends up with a
fraction of what they asked for to begin with and a bill higher
than they planned for. Why? Because they're figuring out what they
really want after they've locked into a vendor and while they're
paying half a dozen consultants to zig and zag their way to a
workable solution. If I were in the market for a PLM solution
today, I would start with an open source solution. If it works, I'm
a hero for saving the company money. If it doesn't, then I've
learned a whole lot about what I really want without paying anyone
a dime. In the long run I have no idea if the open source business
model will work in PLM, but right now if someone is willing to let
me evaluate a full-fledged PLM indefinitely and for free you'd
better believe I'll take them up on it. What do I have to lose?

Tom, thanks for your post. Most companies don't track the TCO of
their software. That is true in all of enterprise software (from my
experience at least) and not just PLM. I have seen some studies
that quantity it well, but I don't have anything at hand to point
to. Some rules of them are 4:1 ratio, but that typically only
covers implementation. I think your experience of adding an extra
"0" to the end, or a 10:1 ratio of total cost to software cost, is
probably closer to the truth. The biggest thing missed is the
allocation of peoples' time, particularly those outside of IT,
which includes users' time (not productive time, but "extra" time
spent on implementation, training, support, etc.) I am following
@aras_corp, I am @jim_techclarity

Randall & Jim, Thanks for the encouraging words on Open Source.
I contributed to the discussion on www.virtualdutchman.com. It was
great to get the visibility for Open Source PLM software. I would
like to expand a bit further on cost measurement. Have you seen any
end user companies publish TCO studies for their PLM Software
implementations? The only ones I have seen have been vendor
sponsored. A few years back, I developed a PLM deployment proposal,
and in a draft cost justification I included estimated internal
costs of the deployment. This consisted primarily of employee time.
The 7 digit software and services number drove an 8 digit internal
cost number! The internal cost number was not included in the final
presentation. Internal people [primarily end user] costs were not
tracked, so, they really didn't matter, or so we rationalized. Best
Regards Tom Gill www.aras.com PS We're on twitter too @aras_corp

Thanks for commenting. I agree that there is a certain amount of
FUD (fear, uncertainty, and doubt) coming from people that have a
vested interest in the current licensing models. Often times, I
find that many of those vendors are also misinformed about open
source. Their only view is what they are used to, and open source
seems scary. What you and I both know is that open source is just
one more business model that a software company can adopt. Should a
company jump into open source without understanding what they
will/won't get from the supporting vendor (or vendors in some
cases)? Of course not. They need to do their due diligence - in the
same way that they should be doing due diligence on a software
supplier with a traditional model. Your fellow Twitterer,
@jim_techclarity

As I Twittered to you (@randallnewton), I would be harsher on the
critics of Open Source than you are. They often rely on FUD
messaging to disparage Open Source for business. Not just PLM, but
in general.

The author of the blog replied to my comment on his blog, and we
are not as far apart in thinking (nor do we entirely agree). Where
we agree is that open source PLM doesn't solve the COST problem of
PLM. Please read his comment, and take a look at the rest of his
blog as well, it is called the Virtual Dutchman. The link to his
post can be found here:
www.virtualdutchman.com/2009/02/15/free-plm-software-does-not-help-companies/#comment-144