SaaS vendors, continue to debate which flavor of SaaS will ultimately win out. Vendors fall largely into one of two camps – the purest such as WorkDay, Taleo or SuccessFactors who host their customers out of data centers running on the same code base. The other camp is occupied by those that host separate instances of their product for customers to not only consume, but also to control. Is the small savings worth the lower cost?
Generally speaking, the term ‘Cloud Computing’ refers to applications that are hosted with a more ‘purest’ approach. To understand this a bit better, let’s walk through an example of how it works. At Bob’s Business Systems’ data center, they host clients of their core product. They have each client’s data stored in a large shared database and then have a series of application and web servers all running the same version of the same software for their customers. This is great for Bob because he’s able maintain only one line of code and one server configuration for all his clients. This make upgrades easy because everyone is upgraded at once. All this translates to lower cost for Bob and savings for his customers.
There are a few issues with this model though such as security, quality and control. For now, let’s talk about the quality side of the house and those cost-effective upgrades.
Bob’s company is a young company that started in his Mom’s basement but quickly grew in once he received that first round of VC funding. Sure, he’s now a top player with $150 million in revenue and has seen double-digit growth for the last few years. Reality though is that it is still a young company that has spent its entire existence focusing on getting product to market quickly and adding as many features that will attract new customers so that they can continue to see 30% – 40% growth.
Let me ask you though, have you ever noticed that when you ask your IT team to make a significant enhancement to your core system as quickly as they can that you seem to have more bumps and bugs than when you don’t rush it? Or maybe you’ve noticed that things are as smooth when you are working with that new developer. The truth is that when you are rushing to get product to market, you don’t always take the time to test and rarely have the hundreds of hours been spent developing your processes to manage release, changes, testing, etc. The result? Young teams can run the risk of developing products that aren’t clearly thought out, well designed or thoroughly tested.
In a recent article syndicated on BNET, SilkRoad co-founder Brian Platz said one of the reasons his purest vision of SaaS will win out is because if the vendor sells software that can run on internal computers, in other words something I can bring in-house, is it will be more expensive because ‘extensive testing may be needed in advance of new releases.’
Hold the phone their Brian, so what you are saying is that you can get by with minimal testing of products since you are hosting them? My guess would be you feel comfortable with that because you feel you have control of your environment. Glad you aren’t losing sleep at night, but I’m as nervous as a long tail cat in a room full of rocking chairs. Personally, I like the comfort regression testing gives me.
The other thing the purest don’t like to talk about is those bugs that they create (every vendor has them so I’m not suggesting others don’t). When they give it to one client, they give it to every client. Everyone is on the bleeding edge together.
At the end of the day, the risk is that cloud computing will continue to leave customers without a firm footing.