Is a billion dollars worth of server lying on the ground?

Years earlier, old brick & mortar company brought in tech consultants to rapidly pivot to e-commerce, who then farmed out most of the work to InfoSys.

My team's own spend was ridiculous. Our hottest data set, which could easily fit in RAM, was on DynamoDB. So much "event sourcing" and CloudWatch. Because you needed all those logs for troubleshooting such a brittle system, right?

And since our core function was the recommendation engine, of course we hoarded and munged data like gnomes, with almost no value add, result negative ROI. (+70% of our "lift" was from the customer's own "recently viewed" list. The highest cost recommendations accounted for less than 3% of "lift".)

There was some director level push to migrate from AWS to Google. Which meant first Kubernetes and then Google. Such fun.

There was next to zero consideration of questioning assumptions. Like unwinding the unholy tar pit of mainframe era batch processing mentality fork lifted onto cloud era web services hotness.

And I don't think they could question their assumptions. The skills and experience of the traditional brick & mortar types simply couldn't.

This "consultants ruin everything" story has played out across all industries.

It was far better to invest a little up front, and maintain at $600 my operations than the same for $96K a month, that's why.

I never "lifted and shifted", I built and deployed, with physical servers, a 3-way duplicated environment that flew like a hot rod. At a fraction of cloud's expense.

1) It isn't just that AWS (and others) provide loads of services, it's that I can run everything I need in the same data centre. SQL Azure might be a good pricing model for your system but if it makes sense for everything else to be in AWS (features, cost whatever) then I am more likely to spend more on AWS RDS or whatever. Might cost £1000s more but I cannot simply split my load across two providers, that is two points of failure.

2) The level of support you get from OVH is nowhere near as good as AWS, which is to be expected when comparing a "budget" provider with the "gold standard". OVH can rent stuff to you cheaper because their support team is presumably much smaller.

3) Even the UI for OVH gives me the heebie jeebies. I use them for personal stuff because they are cheap but the UI is far too slow to use for fast work. I see notifications that I have previously cleared and I get random emails occasionally about things I don't quite understand. I can live with that for cheaper personal sites but if I was using that every day, I would quit!

Of course, for many people, there are issues with marketing (I haven't of some of the other providers) and the basic, and probably fair, assumption that with AWS you know the sort of quality you expect - a bit like buying Toyota. The time and risk involved in evaluating other providers is not worth it in many cases, especially when you only learn 6 months later that OVH don't have a configurable load-balancer or they don't have proper zone redundancy or you cannot upgrade certain hardware without literally copying your files into storage and trying to copy them back to a larger machine.

From a marketer's perspective, (as long as the site doesn't go down), this is actually facebook doing their job.

The scale can happen once you've validated your startup, and when that happens it's a lot easier to just turn on 30 more EC2 instances than to get on the phone with Dell.