FluxHQ - Rambings of Beings from a Parallel Universe

The last few days I’ve been evaluating various Open Source/Free CRM and ERP solutions for SME’s(Small & Medium Enterprises) and let me tell you, I was in for quite a ride. Before I go into further detail, these are the products I evaluated:

• ADempiere
• CentraView
• Centric CRM
• Compiere
• Cream
• ERP5
• Daffodil CRM
• Hipergate
• OpenCRX
• OpenBravo
• OpenMFG
• Opentaps/OfBiz
• SourceTap CRM
• SugarCRM
• TinyERP
• vTiger + CK-ERP
• XRMS + CK-ERP

Some general comments before my analysis of the individual packages. I’ve found that these offerings can be broadly classified into two, Java based applications and LAMP-stack based applications (ERP5 being the only exception). MySQL is pretty much the common denominator amongst all of them, barring OpenBravo and Daffodil. As for other database support, Oracle seems to be a pretty strong player in this market, especially with their new 10g offerings in Linux and the free XE alternative. PostgresSQL gets a lot of love too as does MS SQL Server. Another classification that can be made is desktop application and web-hosted application, the latter is clearly the preferred option by many, as I have discovered. Some also offer it as a SaaS (Software as a Service) product, which has a lot of cost benefits over hosting it on your own. Add to it the fact that you can have the vendor customize the app to meet your exact requirements, makes SaaS an enticing play.

My initial objective was to find a system that was free, easy to use and met the requirements of my fictional company ACME Beer Distribution. I was shooting for lowest cost of entry and potentially the highest possible ROI. The company I made up, ACME Beer Distribution, was a small beer distribution company who needs an ERP solution to manage their procurement, sales, customers, shipping, digital assets and accounting. Some of the other features I was looking at were ease of administration, usability and functionality match. Both self-hosting and SaaS option were possibilities, but cost was the driving factor. My findings were very interesting and so I decided to grade them based on my criterion, so here they are.

Grade F: Cream, Daffodil CRM

• This is a CRM app focused on media organizations. It’s got a poor UI, is written in Java and supports Linux, FreeBSD and Windows. OS X is interestingly not one of the supported platforms, considering the focus, I think is not a very redeeming fact. Other than that, the development of this product looks to be at a standstill, so not much hope here, at least you don’t have to pay for it.
• Here’s another CRM product trying to emulate a mini ERP, add to it that development of this product has seemingly come to an end and the fact it uses a non-standard database, you get a recipe for a product to avoid. Daffodil has a non-working demo on it’s homepage and all signs point to a dead project. It’s mentioned here for posterity.

Grade D-: OpenCRX, SourceTap

• OpenCRX would not have been given such a grade had I been evaluating CRM systems only. It is quite the full fledged CRM application with a fairly decent albeit confusing UI. Support and internationalization is very good as is the multitude of databases it supports. If it evolves to anything more than a CRM system, it could very well be worth another look.
• Here is another candidate for the UI reclamation project. Since the world apparently needed one more CRM application, SourceTap stepped up and filled that need. It has a fairly comprehensive feature set for sales force automation, things like E-mail integration and mobile device support are missing. Their support system and documentation could use a little help.

Grade D: CentraView, Hipergate, vTiger+CK-ERP*, XRMS+CK-ERP
This grade category is very indicative of the state of open source CRM & ERP implementations today, mediocre.

• CRM with a little bit of this and little bit of that, that seems to be CentraView’s motto. It’s a J2EE application and is truly open source. CentraView offers consulting and support services for a fee. It’s internationalization support is poor and it’s self-touted sales force automation is woefully lacking any mobile device support. It’ll be interesting to see if it’s developers can get it right in subsequent releases. One of the nice things about it is that it’s got a fairly usable UI.
• Hipergate is another truly open source application. It is web and java based, so speed is not its forte. It’s feature set is a bit skinny and it’s implementations haven’t really been spot on. It’s documentation, support and user base also leave a lot to be desired.
• Middleware. How cliche. Well the next two are good examples of it. vTiger is a open source CRM application based on the LAMP stack, much like SugarCRM, and actually almost emulates SugarCRM to a T. It does have some compatibility issues and the developers are pretty slow to integrate new changes. The user base for it, though, is fairly large and hence it has a very usable support and feedback system. The UI is pretty nice, but can get confusing at times.XRMS is a barebones CRM system with a limited UI and confusing setup. It’s only graded in this category because of CK-ERP, which when integrated properly makes this a semi-usable product.As for their ERP capabilities, they are able to extend themselves into that arena by being the middleware for another open source project, CK-ERP. The feature set is very comprehensive, but the documentation and the support group are not too compelling. Neither is the implementation. It makes no attempt to integrate into the middleware’s UI and workflow, but instead just sits on top and becomes an application the middleware links to, because of this there are some application redundancies and no inter-application communication. However, it is a rapidly deploy-able solution for small organizations with limited ERP needs.

Grade C: ADempiere, Centric CRM, Opentaps, SugarCRM

• Now, normally I dislike projects that have arisen due to a flame-out between developers, but I make an exception here since it is a noteworthy product as well as a distinguished product when compared to it’s parent project. ADempiere is largely based on Compiere and aims to be the truly Open Source and community focused application out there. But in doing so it has forgotten that it’s companies who are the end-users of this product and not developers, at least not what developers think companies need as an ERP solution. So the feature set is a tad obsolete and as usual the UI is nothing to write home about.
• Centric CRM is another java based offering out there, and just like SugarCRM, it does not belong in this discussion, but nonetheless it’s features and implementation, showcase what happens when open source is done right. The install is a bit tricky but once you get past that, it’s smooth sailing. It utilizes Tomcat to serve up it’s web based Java-CRM-ness, but it’s well implemented and seems to have quite an active community behind it.
• Our next contender suffers from the same UI disease that almost everyone else has. It utilizes an Apache project OfBiz as the code-base for it’s offering. That being the case it tries to be a jack of all things and succeeds in being a master of none. It’s invoicing, quoting and some of it’s CRM modules leave a lot to be desired. To be fair though, development is still underway, and OfBiz seems to be a good base to build upon. Hopefully in the upcoming releases they can fix the UI and some of the shortcomings. Opentaps has the potential to be one of best Open Source ERP solutions out there, it’s just not there yet.
• The most impressive of this group is SugarCRM and though it technically does not have many ERP features, I was compelled to put it in here. From install to UI and customization, this takes the cake. But all is not great in Sugar land, the free version comes with limited functionality and the wiki and docs are very disorganized. Along with it’s Asterisk integration, this makes for a fine web based CRM system.

Grade B: ERP5, OpenMFG, OpenBravo, TinyERP
These packages had quite useful features but yet were lacking on some fronts.

• ERP5 has an impressive feature set and it’s UI is one of the nicer ones in this comparison, had a lot going for it, until I had to install Zope. I’m not going into the shortcomings of Zope in this post, but it’s definitely not what I would choose to run my “enterprise ready” app. The other factor against it was that it decidedly has a European focus, not unlike TinyERP and OpenBravo.
• OpenMFG has some serious user interface engineering problems and is not for everyone, as it’s focused on manufacturing. It’s also not a web based tool, it requires a client for the customer’s platform of choice. The only reason it gets a B is it’s redeeming feature set.
• TinyERP is a desktop app, but has connectors for web apps like Joomla. Pretty decent feature set, but not being web based and it’s European flavor has it at a B.
• OpenBravo has most of the features but like most of the others lacks a nice UI. Like it’s compadres in this grade category, it’s main customer base is European, Spanish in this case.

Grade B+: Compiere

• Well, for one it has the most unique name of them all. It also has the most complete feature set of the group. It runs well on different versions of Oracle, and though there is no official documentation about it, it apparently runs on other relational databases. It’s a Java based app, a knock in my book, but the bigger fault is it’s UI. Very clunky and has a 90’s computer feel to it ala CDE. Access is through a client agnostic Java webstart client or a HTML client.

*I wasn't able to install vTiger on a LAMP stack with PHP 5.2, but on 5.1 and below it seems to be fine. This was a big factor since the upgrade to 5.2 basically broke the application.

Well basically there is no solution that perfectly fits my mould, which was disappointing. A few came close and none of them I could install on a LAMP stack, most of them are Java based, which is cool and all, but somehow I have a Java stigma that I cannot get over, and it’s got ‘administration nightmare’ in it. SaaS is my only viable option, the cost of entry is not prohibitive and it’s someone else’s responsibility to maintain the app. The net net is that ACME Beer will have to shell out some money for a SaaS ERP solution to perfectly match their needs, and Compiere is a front runner for that cash.

Some of my other observations:

Almost all the applications have a pretty significant learning curve. Something, I think, that needs to be addressed. If you believe all those Microsoft adverts on the telly, you might just think that Microsoft, through acquisitions, has solved that mystery. It does highlight the need for easy user adoption though, which is key in SME’s. They have a hard enough time trying to be in business, the last thing they need is training sessions for a system that is supposed to make them improve business.

Most of these entities who create the kinds of software listed above seem to be a bit confused about CRM and ERP. ERP stands for Enterprise Resource Planning while CRM is Customer Relationship Management. ERP is about managing just about everything a company has and optimizing it to further improve their business prospects, earnings and to facilitate cost savings throughout the company, at least in theory. Whereas CRM is about customers and how businesses want to treat customer relationships as assets, which they hope can be leveraged to improve sales. In order to do so you have a CRM application which helps facilitate this process through improved communication, better and more targeted marketing initiatives etc. To better understand this difference, I highly recommend reading this article.

Share This

My heart goes out to all those affected by the shootings at VaTech. I don’t know any of you, but this has touched me nonetheless. Be in peace.

photo courtesy of flickr user Claudia1967

Share This

This is a follow up to my post about my experiences with Asterisk, the other day.

So there we were armed with a freshly pressed TrixBox 2.0 CD and a carafe full of coffee, the latter just in case. After reading through docs on the TrixBox website we were skeptical at best of their promises of having everything installed and configured within an hour. Just to prove to ourselves and to justify the hype, we decided to time ourselves. The CD went in the tray and we booted the server up, a few keystrokes and about 30 minutes later I was configuring FreePBX with extensions and whatnot. About 20 minutes later, we had setup 5 extensions, with their requisite Linksys IP phones, configured our IVR (Interactive Voice Response) menu complete with custom recordings, added & configured all the incoming and outgoing routes and barely made it through one cup of coffee. We then got around to testing the lines and having people call us to test out our various configs. 25 minutes later we were out the door, our job there being done.

Yup, the hype is REAL! and it’s well justified to boot. We now have a fully functioning PBX complete with IVR, DISA, Callback, Voicemail etc. configured to give us the lowest possible cost for routing a call. Incredible, that it took us just over an hour to get all this accomplished. To the Trixbox folks, thank you. You have made it infinitely easier to install, configure and use a wonderful product along with a whole bunch of other useful related products.

Share This

So Apple has decided to delay the launch of Leopard until October and here I was all ready for a June Mac purchase. I guess I will save a few more pennies and wait until then. This is going to affect their second quarter mac sales figures, but I think it’s a wise choice especially since it was done so that the iPhone launch would not be delayed. The iPhone is much more anticipated and hyped, and considering most people who were waiting for a new OSX would have a reaction similar to mine anyway. An iPhone delay could mean a lot of other things,  including a big dip in the stock price, though in after hours trading AAPL has taken quite a dive since the Leopard delay announcement.

In other developments, I tried out VMWare for the first time the other day. Installed the latest version of Ubuntu and did a little server virtualization. I haven’t actually had time to measure performance but I’ll take wildbill’s & greenfly’s word for it, they use it very successfully with HP Blades at their work. So far I’m very impressed with the ease of use, install and setup. I need to investigate this more thoroughly and maybe have it talk to a few SANs and have it actually do some work, then I can measure performance and potential cost savings.

The other thing I’ve been tinkering with apart from my regular workload is Asterisk, no thanks to Samir, my long postponed interest in asterisk has been rekindled. So last weekend we got to work installing and configuring the sucker, but woe is us and we did not compile asterisk with the required zaptel modules(we didn’t know that then) and so we were stuck for a several hours trying to figure out why it wouldn’t work, all the time reading everything that was out there on asterisk. But clearly fatigue had gotten hold of us and we missed the part where it mentioned asterisk and zaptel together, we had them working fine individually though… Ah well, lesson learned. We also, in our quest for asterisk knowledge, came across Trixbox(formerly called Asterisk@Home), and we are going to give it a whirl tomorrow night, while the rest of our peers drink beer and party. Needless to say I will keep y’all posted. The other cool resource I must point to you is Asterisk Tutorials, they have great little screen-casts about what asterisk can do for you and how to setup the various features. Let me tell you, I’m very very impressed by what this piece of software can do, I’ve seen several thousand dollars being spent on proprietary hardware/software that do 50% of what this does. Anyway, a full description of our Trixbox saga coming up soon.

Share This