Blogs

RedMonk

Skip to content

Systems Integration and the Open Source Opportunity

When I was a systems integrator, one of the moments I came to dread in the initial stages of a new project was pricing. Not our pricing – clients usually expected to pay decent sums of money for people, and the boutiques I worked at the last few years tend to be priced fairly aggressively – but rather the conversation about software cost. Customers would start hearing about twenty thousand dollars for this, another hundred thousand for this, and sticker shock was the inevitable but regrettable result. Ultimately, of course, larger customers were able to whittle down down sales people by buying toward the end of a quarter, but the conversation was never fun for either us – the systems integrator – or the customer themselves. If I never do another NPV/ROI analysis for enterprise software, it’ll be too soon.

This is why I find it somewhat perplexing that larger systems integration shops stil show very little awareness of open source products. Of course their are partnerships and resource issues that come into play, but on the other hand the economics would seem to be quite compelling. Simplified, we can look at it as an equation:

Software Cost + Hardware Cost + Services Cost = Total Project Cost

I’m not sure if services pricing has changed dramatically over the past few years, but the rule of thumb when I was in that business was that the services typically priced out at somewhere in the neighborhood of one and a half times the cost of the software involved. Assuming that or a similar figure is still true, i would seem to explain the reluctance of systems integration firms to consider open source solutions: the total size of their deals would be going down, as they’re unofficially (and occasionally officially) incented to pitch more expensive software.

But at the same time, it seems clear to me that decreasing the total cost of the software component of the Total Project Cost via lower cost open source alternatives frees up not inconsiderable sums of money. When you consider that along with the fact that many of the open source solutions are more cost competitive than feature competitive with their commercial counterparts, and it would seem that open source represents a compelling value proposition for SIs as well as customers. If you were an SI, wouldn’t you be interested in customers that had a.) more budget and b.) an increased need for customization? I know I would.

But they do not appear to be. Many of the open source vendors report difficulties in attracting sufficient attention from large systems integrators, and a couple of consultants I’ve spoken with recently have described product evaluations that didn’t even consider open source alternatives.

Until the bigger SIs figure that out however, and see the value in software like Greenplum or SugarCRM, the opportunity in front of firms like CIGNEX, Optaros and Virtuas would seem to be sizable. This is one of the areas I look forward to exploring in this afternoon’s panel.

Disclaimer: Greenplum is a RedMonk client, while CIGNEX, Optaros, SugarCRM and Virtuas are not.

Categories: Open Source.

  • http://duckdown.blogspot.com James

    Wouldn’t you acknowledge at some level that integrators simply follow what large enterprises desire? Large enterprises themselves aren’t embracing open source in the way they should…

  • http://jroller.com/page/jaimec Jaime Cardoso

    Stephen, up until last month, I was working for a reseller. while I was in the Sun area, we (they) sold everything computer related.
    First of all, the price of Software is a complicated thing. Even in Software only companies, it was pretty common to work with 80% discounts and, that, in the early negotiation stages. I don’t know if this holds for larger countries but, let’s face it, you can’t sell anything in Portugal at (say) German prices.

    As for Resellers, I came to realise that most sales people I’ve worked with have this sacred cow: “There is no tomorrow”. No sales guy I know will trade the few coins that he get’s in his sales comission selling software for the chance to get a million bucks selling services tomorrow.

    Finally, you’d be surprised (or you wouldn’t) to discover how many people still regard Opensourced and free software as low quality stuff made by kids or turned Opensource before the company EOL it.

    That was one of the fights I had for many years, this is a simplistic view on why I lost that one.

  • http://www.bluehammock.com Eric Carlson

    “I’m not sure if services pricing has changed dramatically over the past few years”

    Yeah, for operational CRM system implementations, they’ve gone down. :)

    I think James hit on it earlier than I did, but it’s not the SI that’s going to drive OS usage within the enterprise. Organizations themselves have to have the mindset and the ability to embrace the change of moving towards an OS solution.

    Typically the skill sets needed in house are not there. Organizations have lost that “development atmosphere” in recent years – more and more developers have been traded for package support type resources who can also manage outside vendor relationships (software and services). When the thought comes down to have the resources in house to be able to modify, support, and extend an OS package it usually does not sit will with today’s CIO. The mantra has been and continues to be: Outsource, lease, borrow. I think that much of the OS direction today is headed in the opposite direction of that trend.

    We’ve had multiple discussions, and even some small package selections that have included OS solutions – the problem always comes down to the long term survivability of an open source package, the percieved supportability of that package, or more commonly the supplied feature set.

    I believe the feature sets of the currently mentioned OS platforms are pretty basic. You can make a small argument on the functionality of, say, SugarCRM vs. a hosted solution like salesforce.com, but these applications really don’t have even a small percentage of functionality of the larger CRM vendors. They have a ways to go to really be considered ready for the larger enterprise.

    For small-mid market companies, I think they’re great. We use Sugar internally and it does pretty much everything we need.

    Problem is there is very little revenue in selling to an organization that would consider it. Good for 1-2 person shops looking to build a client base, but you’re not going to see any $100k SugarCRM implementations happening….at least not yet.

    Hope you’re well.
    - E

  • http://www.redmonk.com/sogrady stephen ogrady

    James: certainly i think that’s a big part of it. but i can also say that having done a fair number of product evaluations and bakeoffs on behalf of clients, we were responsible for getting products on their radar because they lacked expertise.

    a couple of the examples i referred to as well involved products not very well known – niche type offerings – that were most certainly not implemented at the request of clients. they’d never heard of them.

    Jaime: interesting examples, and fully agree that the actual cost of software is a very complicated equation indeed – something my equation glosses over. as for perceptions, i think that’s true but changing. i can’t speak to the Portugese market as you can, of course, but i can say that within the US perceptions of open source as implicitly lower quality have abated significantly. which is good news.

    Eric: great to hear from you, sir. any number of interesting points in there, but seeing as i tackled the SI as driver point before, let’s look at a couple:

    * Typically the skill sets needed in house are not there. Organizations have lost that “development atmosphere” in recent years – more and more developers have been traded for package support type resources who can also manage outside vendor relationships (software and services).

    i don’t agree with the first part, necessarily, around skills but i do tend to agree with the latter. when i look around at sources like Craigslist, i find skills in various open source technologies – say, Linux, Apache, MySQL and PHP – to be relatively plentiful. and given the volume of business folks like MySQL and Zend are doing within enterprises i’d say the skills are definitely in the enterprise.

    but on the outsource v build front, you have a point. i think it depends in part on precisely what kind of development we’re talking about, but objection sustained.

    * We’ve had multiple discussions, and even some small package selections that have included OS solutions – the problem always comes down to the long term survivability of an open source package, the percieved supportability of that package, or more commonly the supplied feature set.

    this is very interesting, but i think we’d have to parse on a case by case basis. i doubt, for example, that many enterprises would question the survivability or supportability of packages such as Apache or Linux. but might they be more skeptical of SugarCRM? certainly possible.

    i think it will become more obvious over the next year or two however that several of the application packages are here to stay, and it’ll be interesting to see how that news is received.

    * I believe the feature sets of the currently mentioned OS platforms are pretty basic. You can make a small argument on the functionality of, say, SugarCRM vs. a hosted solution like salesforce.com, but these applications really don’t have even a small percentage of functionality of the larger CRM vendors. They have a ways to go to really be considered ready for the larger enterprise.

    true, i’d say. true for how long is a good question – as some of the open source packages are evolving at phenomenal rates – but there’s little question that something like SAP, say, is overwhelmingly more functional than something like, say, Compiere.

    the question of course comes back to the 80/20 point. are all of those features really “necessary?” like our brains, do any of use any more than 10% of Word?

    now that question won’t matter for many enterprises as they’ll simply take the safe, conservative route and overspend on functionality they don’t require, but in the meantime other businesses are looking at the number of failed implementations and thinking twice.

    * Problem is there is very little revenue in selling to an organization that would consider it. Good for 1-2 person shops looking to build a client base, but you’re not going to see any $100k SugarCRM implementations happening….at least not yet.

    well, there have been some fairly sizable implementations of products such as Zimbra, but again point taken.

    i think SaaS will play a role here, and i also think it behooves would-be vendors to either a.) tackle this approach technically by building value add stacks on top of the open source that are light in terms of their services requirements or b.) collect proof points that these packages are good and viable at levels far about the couple of person shop.

    thanks for stopping by – let’s chat soon.