What about BizTalk Indeed?

Oliver Rist wrote a column in his blog titled "What about BizTalk?" Historically I haven't responded to these types of articles, but that's what blogging is supposed to be for - expressing your opinions and this seemed like a good article to respond to so here we are. Starting at the beginning - I never really like the Michael Vaughan character on Alias. He kind of reminded of a secret agent for a certain European country and a Greenpeace incident in NZ a while back so I was happy to see him being written out of the script (assuming he doesn't reincarnate).

Oliver asserts that for the vast majority of its implementations BizTalk will require plenty of time and money. When I started in this business 5 years ago integration products were sold by sales folks who owned boats, fancy cars and the type of houses with those cute Victorian layouts and views of large lakes. Why? The software was super duper expensive. Microsoft has been singularly responsible for reducing the price-point of integration software by over 100 times. Indeed we are still the price-leader in the space and in several deals that come to mind our competitors have dropped their catalog prices by 40% just to be within the realm of our pricing when they hear that the competitor is BizTalk Server.  On the other hand, Oliver does have a point hooking up existing systems across multiple platforms is inherently a tricky proposition and for that reason BizTalk developers get paid more than Web developers and the projects tend to be more strategic to the very existence of enterprises.

Oliver talks about BizTalk Server being revved back - I'm not sure why.  BizTalk Server is not being revved back in the slightest way - that won't make sense as the business is doing very well.  Indeed we have shipped more innovation more frequently than any other server product from Microsoft in the last 5 years, with 4 new versions, and that will continue.  We now have 5000 customers making us the #1 in our space in terms of customers and we are growing strongly as a business due to the value that our customers find in our product.  We are currently planning BizTalk Server vNext which will include Windows Communication Foundation, Windows Workflow Foundation and lots of specific BizTalk innovations.

According to Oliver BizTalk is a proprietary platform which costs money. I guess I could argue this both ways. One way would be to define the term proprietary. If he means that it isn't open-source and it is a for cost product then that's absolutely true and is equally applicable to IBM Web Sphere, SAP, Oracle and all the players in the integration and business process space. Another might be that it is somehow closed in the sense that it only supports "Microsoft stuff". BizTalk Server provides the largest level of interop of any Microsoft product to numerous platforms and recently we acquired 9 adapters to help our customers broaden that through connectivity to Oracle, Siebel, TIBCO and so on..

So what about BizTalk and Windows Workflow Foundation? Did you bet on the wrong horse by choosing BizTalk Server? No you didn't. BizTalk Server is, was, and will be the right architecture for business process and integration on Windows. To properly serve the needs of business process and integration you need a "man in the middle" server that adapts, routes, tracks, manages, scales and has all the appropriate BPM capabilities such as BAM. Windows Workflow Foundation gives you none of that. On the other hand there are many scenarios that Windows Workflow Foundation is useful outside of business process and integration such include page-flow in ASP.NET web applications, workflow within line of business applications that ISVs build, workflow within Office. By targeting ISVs primarily the Windows Workflow Foundation technology can be embedded within these applications.

Put simply:

A. Workflow within applications = Windows Workflow Foundation

B. Workflow across applications = BizTalk Server

I love working with the press, but on one or two occasions some "selective editing" has provided some out of context quotes. Let me reverse the situation and end on my favorite part of Oliver's article "In fact, the synergy of [Windows Workflow Foundation] and BizTalk will probably make you look like a visionary". You bet we are working hard to make that 100% a reality.

Comments (28)

  1. Jeff Lynch says:

    Thanks for speaking out about this article. It’s this kind of "quick & dirty" article, written by someone that’s never even used the product, that really annoys the &^%#$ out of me. I wonder how they would like it if someone reviewed their magazine without having ever read it!


  2. Jeff Mayeur says:

    I have to admit my experience with BizTalk is limited and, at best only mildly disconcerting. While arguments about cost and difficulty as fault points are not necessarily strong, when it comes to criticizing BizTalk. I find it fails on 3 major points.

    1) It over estimates the ability of the average business analyst, Logic by it’s nature is bad, and just because you can wysiwyg an orchestration doesn’t mean you should.

    2) It under estimates the benefits of a roll your own, in terms of critically analying underlying processes and infrastructure.

    3) Too vulnerable to the Hit by a Bus Factor. to implement a best of breed BizTalk solution, you really really have to be a domain expert, that creates bottleneck, and "information silo" issue, that just don’t exists will a home grown solution to integration.

    I know this is a faulted comparison, but BizTalk smells all to much like the VB6 of software, it certail works, and can be tuned to work fairly well, but it doesn’t set the best expectations, or create the idealized growth environment.

  3. James Hart says:

    Cheers for the spoiler warning – those of us in the UK are still months away from getting episode 1 of the next season of Alias.

    But thanks for the clarification of WWF vs BizTalk 🙂

  4. Lance Hunt says:

    I disagree with Jeff’s comments about the lack of value of BizTalk vs "home brew" integration.

    First, toss out BizTalk, and lets talk about Integration Frameworks in general. A good integration frameowrk provides a standardized set of tools for connecting disparate systems together in a managemeable way, with guaranteed delivery, security, flexibility and a strong set of management tools to enable quick diagnosis and resolution of problems.

    Building a custom integration framework is nice because you have 100% flexibility in the implementation, but you spend 90% of your time building & maintaining a framework rather than focusing on the business needs the framework was supposed to deliver.

    There is where an out-of-box solution is helpful. If you don’t have unlimited development time and resources, you need to spend them wisely. By spending some money early on a pre-built Integration Framework like BizTalk, Rosetta, or other tools, you focus the majority of the time on the real business solution of connecting systems together. Also, with such a product you get plenty of admin tool support, which is typically the first thing to get cut when time and resources are adjusted during development of a custom framework.

    Add to this the movement towards SOA, I think that a pre-built solution has great ROI and helps your organization start thinking like a collection of services.

    Now, back to BizTalk – the big question is whether or not the overhead of setup, customization, and maintenance of this particular tool is worth the investment. Here is where I basically just say "It depends on your needs". Like any tool, you have to identify your integration needs and choose the right product to fit. If you are like us, and have a system that handles nearly one million transaction per day, then you definitely can see a quick ROI on BizTalk 2004 (and newer).

    My 10cents…


  5. Chris S says:

    Scott great reply, clearly Mrs. Rist has little or no valuable experience in the business integration space. I can’t help feel that this article was written by an individual who loves Open Source, has little understanding of Open Source, but yet embraces it with all his heart because he feels that it is at odds with the #1 software company in the world. Oliver clearly shows his ignorance on the subject when he said, "WWF will most likely be a free or very low-cost upgrade to BizTalk." Does he know what a framework or an API are? Does he even know what a software application is? Even with a small amount of education one should be able to read this article and realize that it does not make a strong case of any kind nor does it make any sense.

  6. Deepak says:

    I had to agree with Jeff Lynch on this. Somebody who does not have a clue about Biztalk has a review written poorly!!!!

  7. Mick Badran says:

    "All press is good press" – right?

    Scott – you can take away from Oliver’s (limited knowledge) article.

    He’s talking about Microsoft, he’s talking about BizTalk and he’s talking about WF – nothing else. (where’s the competition to compare/contrast in this space)

    I think a key take away from this article is – CONGRATS to you guys (BTS team!) for providing this pseudo standard product that addresses the integration space!!!



  8. Joku says:

    Good clarification but what did the Alias spoiler had to do with it? I vision a world where people get US TV shows the same moment through IPTV as they air in US. So does MS probably, but since it’s not here yet…

  9. Ted says:

    I wouldn’t agree with the "meat" of the article but I will definately agree with the title. While @ PDC and sitting in on a presentation by Don Smith and Beat Schwegler(among others) they showed some pretty cool tools(I think the 3 letter acronym was GTA) that are great for web service and SO development. After the presentation I posed the "What about Biztalk" question to Don and he did get that glazed over look. I am not quoting but I think the comment went something like "the biztalk guys are hard to get with and get these issues ironed out". I know the adapters for WCF will be there but if there are wizards for WCF to help in developing good service based systems, will Biztalk have the same support?

  10. Vineeth Raja says:

    Great post. Thanks for clarifying the role of WWF and where it fits in.

  11. Sandeep says:

    With so many products / technologies being created by Microsoft that appear to rely on similar ‘Visual Programming’ concepts (DTS, BizTalk and now WWF), would it not be a good idea to create a common base framework for Visual Programming? Call it Windows Visual Programming Foundation, maybe?

    WWF does appear to have a lot in common with BizTalk, and according to the the WWF website, it is going to be the core technology underlying BizTalk in its next generation. Sort of similar to what WSS is to SPS. I would agree with Oliver’s observation in that regard.

  12. Great post Scott! I have been using BizTalk ever since Dan Rogers first introduced it at TechEd99 in Dallas. I thought the product (ok jumpstart toolkit 🙂 was so cool and filled a busines need that Barry Varga and I built an SI company based solely on BizTalk! Call us crazy.

    Since then I have been involved in the design and delivery of over 25 BizTalk projects and can tell whomever will listen that is an excellent integration product indeed! So much so, we built a product on top of BTS called BRIDGEWERX that resembles Eddie Churchill’s cool BTS Solution Designer. If interested, you can read about it here:


    Thanks to you Scott and your team!


  13. Mark Sargent says:

    I do find it a little difficult to reconcile the statement "Workflow across applications = BizTalk Server" with the existance of the WebServiceResponse & WebServiceReceive activities in WWF.

  14. Hans Valcke says:

    I can’t help myself in responding to Oliver’s thoughts on the Biztalk product.

    I live in the real world here and we use Biztalk extensively for EAI and B2B scenario’s.

    I must say that the product meets and even exeeded our expectations of an integration tool.

    You can deliver integration solution fast, programmers can use the programming environement they are used to. You can easily write your own custom pipelines and adapters, you have the rule engine, have BAM and HAT. And on top of that you can buy lot’s of adapters and accellerators.

    Our company has never been able to ajust to new business and integration needs as fast as we can do now.

    Also my CONGRATS to the BTS team. Keep up the good work. Afterall, someone has to show the road ahaed.

  15. Don Smith says:

    In response to Ted’s comment above, I’m sure I didn’t say the BTS team was hard to work with. In my experience, they’re not hard to work with at all. Now while PAG isn’t fully engaged with them right now, I certainly suspect we will be in the not so distant future.

    The only "issue to be ironed out" might be the messaging around when to use ASMX/WSE/WCF versus BTS for exposing Web services. This should actually be pretty clear for most people, but when I (as a Microsoft employee) am asked, I obviously have to be sure to give the agreed-upon-by-everyone answer. Now that the WCF and BTS teams are in the same org, I suspect this will be communicated very soon (if it hasn’t been already).

    Sorry for the confusion.

  16. Scott, I agree that BizTalk has been a great force for good in the Integration middleware market, and is currently hard to beat on value, functionality and interoperability for enterprise integration solutions.

    However, I am a bit uneasy about the rather simplistic view of the future roadmap which seems to be the official Microsoft line at the moment. You are of course absolutely right that WF is an API framework whilst BizTalk is a full-featured server platform, and those that suggest WF alone might supplant BizTalk are clearly wrong. However, it seems to me that the meatier comparison is between building integration solutions on the BizTalk Server platform on the one hand, and the upcoming prospect of being able to build them by wiring together service endpoints built with WCF (with hosting and activation courtesy of IIS/Windows Activation Services), process orchestration built with WF, and SQL Service Broker somewhere in the mix. Such a comparison is especially pertinent if BizTalk vNext is not due until 2008, since this would mean there would be a period when the choice would be between BizTalk 2006 (with its "private" orchestration technology, already slated to be supplanted in vNext by WF) and WCF/WAS/SSB with WF orchestration.

    I’m not saying that BizTalk would necessarily do badly in such a comparison, but I think it is much less clear cut than the simple WF v BizTalk argument you (and other MS commentators) have addressed to date. To inform this comparison I think there needs to be some clear picture painted reasonably soon about the likely shape of BizTalk vNext and the migration path from BTS 2006 (especially orchestrations->WF).

  17. Joe says:

    Scott, why don’t you talk about that marvellous piece of … software called HWS that ships with todays BizTalk?

    I sincerely hope WWF is better than this but I still don’t see any "humanity" in the Workflow space in your plans, apart from using Office 12 (why draws a lot of dependencies in the enterprise space)

  18. Last week an ISV approached me after I showed Workflow Foundation and to parphrase, said "Given Workflow…

  19. Workflowwithinapplications=WindowsWorkflowFoundation


Skip to main content