Cloud repatriation: for the 1%

Last week The Economist’s Schumpeter column asked: “Do the costs of the cloud outweigh the benefits?”, referencing, in turn, the Andreessen Horowitz (“A16Z”) post The Cost of Cloud, a Trillion Dollar Paradox.

The point of the latter is that, for SaaS companies, cloud costs can rapidly impact margin as company goes through rapid growth. A solution to that is repatriate from the cloud, hosting yourself. Schumpter’s column ignored the A16Z focus on SaaS and generalized to a more general statement about business and cloud cost. In doing so, they muddied the waters.

The TL:DR here is that for most SMEs the costs of repatriation far outweigh and benefit: The A16Z paper looked at companies like Zoom and Dropbox which, as SaaS companies, are hyper specialized to individual business problems “Video conferencing”, “File storage and collaboration” etc. Their platforms are global in scale, homogenous in architecture, features and pricing. They sell software, on a subscription, which tightly aligns their costs of delivery with revenue.

By contrast the average SME isn’t selling any software, uses numerous cloud services as part of their IT infrastructure and such costs are largely fixed from the perspective of sales.

For a SaaS company, optimization of even a small amount of the architecture (whether through software or process optimization, cloud arbitrage or repatriation) delivers big benefits due to that homogeneity. For the rest of us, our use of cloud is so hetrogenious that small changes won’t move the needle.

That’s not to say we shouldn’t optimize: one of the best things about cloud is price transparency. Tag your resources well and it’s easy to see where the dollars are going and make targetted improvements. But repatriation shouldn’t be on the menu: it’s hard enough finding engineers to work in the cloud, let along build, secure and maintain your own infrastructure. Worse, if you do manage to repatriate, you’ll loose the price transparency.

So beware the snake oil consultants selling repatriation: for the 1% it can make a lot of sense. The rest of us? It’s navel gazing nonsense.