Why Does Google Kill Products and Services?

The recent announcement of the demise of Google Domains and sale to Squarespace brought out the usual reaction whenever Google kills a product — along the lines of, “How dare Google discontinue a product I have come to love?” Dead Google products and services are even memorialized on a website.

And that is usually followed by promises never to trust Google again, and (among software developers) threats to not build on Google Cloud Platform. Honestly, I doubt “Google might one day kill a service we use” figures too highly in most businesses’ cloud decision-making processes. But the emotion is genuine.

(In the interest of disclosure, I am a Google Domains customer, and I know a number of the original engineers who worked on it back from my Google Kirkland/Seattle days. It’s a great product, well-designed and well-engineered, and I will miss it. But I’ll survive, as will everybody else.)

So why does Google keep doing this to itself?

For the same reasons Microsoft doesn’t operate a standalone contact manager or continue to sell Money or Encarta (remember those?):

Companies need to focus on their core strategic lines of business, particularly in a down cycle.

Anybody who was at Google in the ‘10’s has Larry Page’s “more wood behind fewer arrows” mantra seared into their memory. That’s exactly what’s still going on here today a decade-plus later.

Presumably Google Domains is a solid business and probably does not cost a whole lot to operate at this stage of the product lifecycle. Why shut it down?

A small-but-solid business like Google Domains would be the envy of hundreds of loss-making pandemic-era startups. To a company that operates Google Search and AdWords and (yes) Google Cloud, it’s a rounding error.

I have no idea what kind of cash flow Google Domains generates, but let’s just pretend it does $10M a year in revenue. That’s not bad, right? But what if Google can take the engineering and infrastructure resources dedicated to Domains and instead allocate them to AdWords and increase AdWords revenue by $20M? That seems like a no-brainer even to a non-MBA like me.

You might ask why Google doesn’t just hire more engineers and expand data centers if necessary to capture ALL of that revenue. First of all, anyone who has done a lot of work in either tech hiring and firing or data center procurement will tell you it’s not quite that simple (saving some of the specific reasons for a later post).

But beyond that, when it’s clear that a project is no longer strategic given the current operating conditions, it shouldn’t be considered unusual for a tightly-run company to shut it down, sell it off, or spin it out.

Maintaining focus — not just of a single engineering team but also the supporting staff and executives — is basic business management.

Finally, if you are a Google Domains user — or user of any domain services provider, for that matter — rewind back to when you first signed up and ask yourself what other services you might have used. Was there one that would have looked more likely to still exist n years down the road? Are there any now that look more likely to exist n years in the future? I would guess probably not.

Nothing in business, engineering, or product management lasts forever. That’s why if you’re going to work in tech it pays to be agile.