Monday saw Fastly inadvertently wipe out much of the Internet. It’s not just Fastly: all of the SaaS providers have their off days. For instance, earlier this year we saw Azure AD Outages which were sure to cause a pucker if you use AAD as your Single Sign On solution.
Cue the grey beards saying it would never happen on premise. So, should we ditch the SaaS and revert back to the ‘good old days’ of rolling our own? Of course not! Here’s a couple reasons why not:
Strength in numbers
Number #1 reason is that if you’re using AWS/Azure/Google you’re in the same boat as much of the rest of the Internet. You suffer an outage, yes, but so do your peers and competitors. Part of the trick of being a CIO or CTO is backing the right horse. Right now, that horse is SaaS and any savvy CTO will point at the global impact to explain to the C-Suite: “it’s not just us”.
Nimbleness, Flexibility & Agility
That’s a very defensive posture. SaaS software offers inordinate benefits over on premise: more configurable, rapid iteration and feature development, a sales & pricing model that more closely aligns the priorities of supplier and customer, better cost transparency. If you can’t articulate those to non technical types you might be in the wrong role.
World Class Ops & Security
There is massive incentive for SaaS providers to avoid outages. Each one shakes customer confidence and will cause some churn. So the bigger providers invest in talent and facilities to keep their systems up and secure. Can you?
No, unless you have exotic uptime requirements (and associated budgets) and organizational size to retain talent, rolling your own really doesn’t work at this point. For a small or medium size business do you really believe you can do better? Embrace the occasional chaos and ensure your eggs aren’t all in one provider’s basket.