Mr. Programmable Web: Salesforce.com
Morten E. Wulff Mar 22nd
Salesforce.com, the leading on-demand CRM solution, is taking the idea of the programmable web (e.g. Mash-Ups via APIs) for on-demand business applications to the next level by introducing a development platform so extensive and so motivating leaving me quite stunned. They call it Appexchange. I see this bring so many added values to all the clients of Salesforce.com and I see so much potential for companies like TraceWorks.com.
Just how big is this? Simon Fell of Salesforce.com says “Sforce [now called Appexchangee] is one of the world’s most widely used enterprise Web services, accounting for over 40% of the total transactions on salesforce.com” - and with $310 in revenue we don’t need help from Einstein to figure out if this is a success or not …
Open Outcry to all (well, most) TraceWorkers
All TraceWorkers directly or indirectly associated with product development should register for an AppExchange Seminar. All TraceDevelopers should visit http://www.salesforce.com/developer/ and gain an insight on to what extend developing a mash-up would be possible. You might also want to add the company’s developer blog to your RSS reader.
Need a better understanding on what the programmable web, web2.0 (ask Pelle), Mash-Ups etc. is all about? Take a look at this website: http://www.programmableweb.com. (Perspective: today TraceWorks.com is only working with two of the APIs listed here, Yahoo and Google).
I can’t wait to begin developing a Headlight/Salesforce Mash-Up solution and I can’t wait to launch our planned enterprise web service for Headlight.
Entries (RSS)
What you call the programmable web is what developers know as service orientation. It has existed for some years now, but really took speed after the release of the .NET Framework in 2002. It is cross-platform and is an XML standard called SOAP.
The http://www.programmableweb.com website is nothing more than a marketing stunt for non-developers that isn’t aware of the good old service orientation or old wine in new bottles if you will. They are selling the service orientation principles all over again, and I’m sorry, I fail to se how this is relevant to Traceworkers. We already do service orientation. Google, Yahoo, AdServing etc. Our own API will be written as an SOAP service.
The AppExchange sounds pretty cool, but it has some apparent drawbacks:
It’s not open (or defacto) standard:
This means that the application you develop upon the AppExchange API can only be used there. You cannot use it anywhere else. No reuse there.
Limited user base:
You not even have to be a SalesForce user to benefit from this, but you also have to know about it and register to use AppExchange anyway. You also have to prefer the AppExchange over our own web solution.
External API:
We don’t control the API, which mean that when SalesForce make changes, we have to be fast to adapt. It’s important that we trust them, as we would if Microsoft or IBM made a public API. We had our share of problems with the Google API…
Is it that good?
If the AppExchange is such a good idea, why isn’t somebody else also doing it, e.g. Microsoft, Google, and Apple?
IMO, the HTML based application within the enterprise is dying. The way to go is smart clients; thin clients that utilize the power of the PC instead of the website approach which is much, much slower. Thin enterprise applications like this one is very easy maintained and offer so much more than the equivalent web interface.
I do not agree that AppExchange is unconditionally beneficial to Traceworks and I do not buy the mumbo-jumbo of the programmableweb.com.
Is all the software developers in the world not good enough?
The AppExchange is an opportunity for other developers to develop software/ addon’s for the SalesForce CRM system.
So why is Apple and Microsoft not doing this… It’s all they do mate.
They would be nothing if companies didn’t make software for their systems. As well as companies like eg. Adobe, would be nothing if they had no system to make software for.
It is exactly that beneficial symbiosis that is starting to popup in web based applications. Which is a good thing.
Never mind if the technology has existed for years or what people call it. It’s what you do with it and why, that matters.
Just because the wheel was invented ages ago, doesn’t mean we can’t continue developing new or better vehicles.
But I do agree with you that, not using a de facto standard isn’t the best of ideas.
Microsoft, Google and Apple do not provide a plug ‘n play web service last time i checked. This is not the same as the widgets/gadgets of live.com and google.com/ig which is not applications but little info boxes.
But your right, it isn’t about when things we invented but how they are used. As long as we don’t fall victim to the buzz word hype without doing a proper cost/benefit analysis. Just because something sounds cool or looks great does not mean it IS cool. http://askalexia.com is a good example of doing something that sounds cool (Web 2.0) but it brings nothing new. Where is the cost/benefit analysis there? I would say, that it was never made.
Microsoft and Apple make operating systems. Adobe etc. make software for these operating systems. Neither would exist without the other.
That is what web software developers ,like us, is starting to learn. Because it is demanded by our customers. They expect browser based software to work together, just as software you have to install does, and they very well should.
Companies like Adobe and Macromedia (now also Adobe) have had solutions similar to AppExchange for years.
So you can hide behind technical details and buzz words all you want, the answer to your - “If the AppExchange is such a good idea, why isn’t somebody else also doing it” - will remain - “but, everybody is doing it”.
Ok, then give a link to a web site that you can write your own plug ‘n play applications for. And not the little widgets/gadgets sites like live.com. A site similar to AppExchange.
You seem to have missed my point.
Who cares if it is done exactly like AppExchange. The point is that other developers/companies have the possibility to develop plugins/addons to our software, and we to theirs.
Most CMS applications have had this possibility for a while including http://wordpress.com/ and if you want something more commercial try http://www.sitecore.com/ - which BTW integrates with StatLynx.
It’s the idea that is good. Not necessarily the exact same solution/method as AppExchange.
I agree 100% with that. That’s why we make an API. My original point (scroll up, way up), was that I’m not sure if AppExchange is such a good idea, because… bla bla bla, read the commment up there.
And just to make it absolutely clear. Nobody is doing what AppExchange is doing, which is a website framework you can write your own plug ‘n play applications for.
If you say “a website framework you can write your own plug‘n play applications for” you are probably right. I’m not sure I even know what that is exactly, and I doubt that this was what Mr.Wulff meant was the good idea (as he probably don’t know what it is either).
But hey! our customers probably don’t know either, and why should they. All they should care about is if they can do what they want, when they want.
Which means; you can use your excellent skills to build it anyway you want to, as long as it works. Of course you can make something better than AppExchange, I was never in doubt.
But that does not make the concept of an “AppExchange” a bad idea.
I agree, it is still a great idea, but it might gain more ground if you could reuse your apps in other app exchange systems.
As I read the original post, Wulff believes that AppExchange has much potential for companies like TraceWorks.com. I disagree in the fact that AppExchange is beneficial for Traceworks. This isn’t the same as saying I don’t believe in using API’s (the programmable web).
(I wonder why Wulff hasn’t contributed with some comments in this discussion?)
“Marketing stunt”? Come on! Give up this paranoid conspiracy plot. A marketing stunt by whom? The well organized homogenous group of non-developers? And to achieve what (besides trying to explain what this technology is all about and what you can use it for)?
The-Programmable-Web / The-Web-As-A-Platform / Web2.0. These are all what I would call “concepts” or “frameworks” of understanding and use … These are examples of simple evolution. It’s a first step in order to advance a technology (e.g. “service orientation”) by most often merging it with other technologies / theories / disciplines (e.g. sociology, anthropology, mercantilism) and problem areas. The outcome then sometimes becomes relevant use cases adding actual value and/or ease actual pain to certain scenarios. So basically what concepts do is that they elevate technology from being just technology to be about people and real-world problems.
Sure concepts such as The-Programmable-Web & Web2.0 tend to become Buzz. I wonder why? Maybe it’s just a typical characteristic of being relevant and of people actually understanding what the hell you’re talking about?
But sure there is a lot of BUZZ without any real relevance. But the-more-mature-than-just-technology-concepts like Web2.0 and The-Programmable-Web are, in my opinion, pretty solid - also for enterprise software companies like TraceWorks.com.
From a business perspective what makes Appexchange extremely interesting is of course that it makes it possible for small companies like TraceWorks.com to hook-up with Salesforce.com and profile our unique software in front of 25,000 SMBs. If say 2.5% actually choose to manage their marketing activities using an AF/HL mash-up priced at e.g. EUR500 per month then it would mean EUR 3.75M in annual revenues.
But Mads, you’re on to something. Our approach to software development is all about simplicity, quality, and beauty. Will the cost of building a mash-up WITHIN the UI of Salesforce.com just be too great? Will we remain to be unique? Can we accept this compromise? Should we instead insist that people should use two separate solutions: The CRM Solution and The Marketing Solution - integrated with each other via APIs of course (in the same way as with TraceWorks versus WA Solutions)?
BTW I don’t really see any real evidence that browser based applications are dying … And regarding the smart client technology; at this point I don’t see any real world impact. I simply don’t consider this really smart until it becomes part of a pragmatic concept generating some solid BUZZ …