Demand Capacity Real World Example on Windows Azure
I recently had a fascinating conversation with CoreConnex, a company that has a Financial Analysis dashboard called Corelytics. Corelytics is an Azure based SaaS product with about 1,000 customers supported in 11 countries. Last week Corelytics was announced as the Intuit Application Showcase Grand Prize winner. Intuit has a huge customer base for their QuickBooks accounting system and this award creates an interesting dilemma for Corelytics: like, how do you support a flood of new customers that you were not expecting?
Corelytics uses Azure and they were able to connect to Intuit QuickBooks through the Azure App Fabric. This allows a tight, fast and highly secure integration between the accounting system and the dashboard. But the platform decision wasn’t expecting such a big spike in new customers.
But as we all know, whenever you introduce any new product, you have to be ready to service the need – especially if it happens to grow explosively. There’s always a curve of introduction, ramp-up, maturity and so on. Sometimes that curve is incredibly quick, and other times it has a different shape – and that’s the difficulty. You can’t always predict how many people will use your product and when. Not only that, you’re not certain when that need will change, or even at what frequency.
As we discussed sales and adoption cycles, they pointed out that one of the reasons they chose to use a platform like Azure was specifically for the elasticity to accommodate an unknown growth rate. With stateless code, the system can handle massive changes in workload automatically – something that is far more difficult with an on-premise or even an Infrastructure as a Service system where you have to build, configure and patch operating systems.
There’s a subtle but important point here - the point is you need to be prepared for all possibilities, great and small, before you decide where it needs to live. This is another real-world example of that.