Debunking myths about consultants who debunk myths

It's important to understand the advantages and drawbacks of using consulting firms to advise and assist your business.

First, the drawbacks:  Some would say that consulting companies tend to run up their hours, and milk their customers for every last penny - that they cloak their deliverables in buzz-speak and gobbledygook, and live in terror of their clients discovering that they're not really that smart.

As for advantages, well, some others might say that consultants can be of some help, sometime.

<Thud.>

Not a very fair or rigorous analysis, was it?  You can imagine how I felt after reading Eric Kimberling's "Debunking myths about open source ERP software."

It was especially disappointing because I've read a lot of Eric's posts, essays, etc., over the years, and find him to be one of a handful of high-visbility ERP consultants who emphasize process over product, measurable return on investment, and the importance of following well-planned projects through to fruition.

But he's off the mark on this one.  Following is my debunking of his debunking.  (Does that make me a rebunker?)

Myth #1: "Users of open source ERP often need to modify the software in order to gain some of the functionality already in place in traditional ERP packages."  Overly broad, to say the least.  As anyone who's ever tried to put together a meaningful feature-comparison grid can attest, this is the most mobile of moving targets.  Some open source ERP packages have more functionality than others, just as some proprietary packages have more than others.  And - wait for it - some open source packages have more functionality than some proprietary packages.  Have we learned anything yet?

Myth #2:  "Open source often requires more robust programming and software development skills ... total costs may actually be higher over time because of the specialized skills required of IT support staff."  This might be true of some open source solutions - some are more "toolkits" designed for internal IT programmers than end-users - but it's certainly not the case for all of them.  Most notably, our own xTuple ERP solution is frequently lauded for making the complex simple, and bringing a real productized discipline of usability to open source.  Once again, Eric is painting with a very broad brush here.

Myth #3:  "It can be quite costly to continuously update and improve the functionality of [open source ERP] software. Traditional ERP vendors use the relative scale and size of their customer bases to constantly improve software and offer new functionality."  Now, this one's just flat-out wrong.  If anything, the knock on proprietary vendors - particularly older packages that have been acquired by rollup shops like Infor - is that they don't invest much of anything in R&D.  This is a drum that people like Ray Wang have been beating for years.

Open source offers a direct antidote to this condition, and the math is very simple.  More users = more eyeballs on the code = more hands on the oars.  Bugs get fixed faster, and with properly managed community development, new features get implemented in ways that are more applicable to more users. 

Myth #4:  "Myths 1, 2 and 3 can actually escalate the overall cost of open source relative to traditional ERP."  Boy, this conclusion is right out of the Steve Ballmer mid-decade scare book.  Even Microsoft has moved past this talking point.

In all seriousness, Eric's piece restates and compounds a lot of misconceptions about open source generally, and open source ERP in particular.  It reads very much like defensive vendor boilerplate from ten years ago - perhaps not uncommon for firms that are tied to a particular vendor or point of view, but unusual for a consultancy that makes a distinguishing point of its independence.

My own point of view is that xTuple ERP is the most feature-rich, powerful open source ERP available.  The commercial Editions for Manufacturing and Distribution add even more functionality, and since they are the same code base licensed under a community-source license, users of those products can participate in most of the same open source benefits as users of the free product.  We are told almost daily that our product strikes the best balance of functionality and simplicity among all the open source ERPs - and that even the free PostBooks Edition of xTuple ERP is vastly superior to most of the commercial Tier 3 products out there.  We are the only solution I know that runs natively - on both client and server - on any combination of Windows, Linux, other Unix, and Mac.

But does that mean xTuple ERP is for everyone?  No.  If you're a company with a large IT shop that wants to program custom applications in Java, for example, you'd be better off with Compiere.  If you want all of your users to work in a web browser, then you'll likely prefer OpenBravo.  And if you're married to Microsoft .Net technologies, then you're probably stuck with MS themselves, or one of the surviving midmarket acquisition shops like Epicor.

In conclusion, I guess with any ERP selection and implementation, as the kids say, YMMV (your mileage may vary.)  I don't believe Eric's sweeping generalizations about open source ERP were well-advised or supported by any actual data; moreover, like all generalizations, they were wrong :)

Let me close with a peace offering to Eric and the many other consultants who make a living evaluating and implementing ERP systems.  Open source ERP - most definitely including xTuple - can be a fantastic tool to generate new business, and to help bring the world-changing benefits of ERP to companies who might otherwise have not been able to ever get there.  And unlike with proprietary vendors, you don't have to take our word for it - download the software and see for yourself.  Speak with actual users in every kind of industry on the forums.  Watch our public progress fixing bugs and adding new features.

And then look down the road, think about the accelerating pace of open source adoption, and ask yourself, honestly - isn't this a better way to develop and maintain software?