Pingdom Home

US + international: +1-212-796-6890

SE + international: +46-21-480-0920

Business hours 3 am-11:30 am EST (Mon-Fri).

Do you know if your website is up right now? We do! LEARN MORE

Google’s clever choice of words during the Gmail incident (why 0.02% matters)

GoogleSite issues come in all shapes and forms, and no one seems immune. Google’s Gmail problems the last couple of days, where a number of Gmail users temporarily lost all of their emails, was proof that Murphy’s Law is alive and well on the Internet.

However, this post is not really about that specific incident, but rather about an interesting detail in the way Google communicated the problem.

Oops, but only 0.02% are affected

In their official account of the incident, Google was very careful to point out (twice) that only 0.02% of all Gmail users were affected. No mention of how many users that corresponded to.

[...] Imagine the sinking feeling of logging in to your Gmail account and finding it empty. That’s what happened to 0.02% of Gmail users yesterday, and we’re very sorry. [...]

In fact, during the entire conversation about the incident, Google consistently used percentages, never once mentioning in absolute numbers how many Gmail users were affected (at least we didn’t see any).

So the focus in Google’s communication was to point out that tiny little fragment of a percent. Doesn’t sound so bad, does it?

A matter of scale

To realize why this was a clever move, you first need to know how big Gmail is: Gmail supposedly has around 200 million users. It’s a huge email service.

A thin slice from a huge pie will still be significant, and even a mere 0.02% turns out to be a large chunk. In this case, it corresponds to 40,000 users.

40,000 affected users sounds a lot worse than 0.02%, doesn’t it? The latter gives a much more positive spin. “We’re having trouble, but only a tiny share of our users are affected.”

Which is of course why Google didn’t mention any absolute numbers. This was a company in damage control mode.

It’s actually a pretty smart (but not very transparent) approach for any big service, because even tiny shares of a huge user base will number in the thousands. Even “small” incidents will look bad if actual user numbers are discussed.



10 comments
Seo Montreal
Seo Montreal

In this case, we are talking about an email interruption of over 9 hours, which could be terrible for most business users (aka apps users). Imagine if you rely entirely on the cloud for all your productivity? It just shows that backup are still important to make, even if only for the sake of not hitting F5 9 hours in a row !

Tim Stiffler-Dean
Tim Stiffler-Dean

It may also be an interesting ploy to cover how many users they actually have. Some estimates for the outage on the web were much higher than 40,000, and we all know that people exaggerate during small-talk over coffee. 0.02% now could be any number between 20,000 and 200,000. And that makes a big difference if you sit and think about what 100% actually means.

Schnäppchen
Schnäppchen

40,000 affected users can be very strong in social media. you realy have to take nowadays.

Jason Crabtree
Jason Crabtree

2% is a small percentage. But 40K is a large number of users. From the corporate perspective, think percent and the problem is fairly minor. 40K users won't be too happy though.

Frank Borh
Frank Borh

C'mon! remember that this is a **free** email service. The SLAs are enforceable when you pay for a service. Any other thing is just good will from a company

Sudharsan. G
Sudharsan. G

still 40k is very less compared to 200+ million users, i think it wont affect much if Google had said the affected users in numbers.

Cam Jackson
Cam Jackson

They were also very clever to simultaneously highlight the tiny percentage, yet still apologize profusely. It makes people think "oh, this was only small thing, and they still treat it as a big deal. Gee google are such nice guys!"

Decio
Decio

I passed my eyes through some news about this ongoing problem, but it seems to me that Google has been revising their estimates of affected users downward during this time. To be fair, you would also have to take into considerations how many of those 200 million accounts are active and also the one not used for email (backups, gdrive, one time use accounts, etc). I guess that would amount to a lot less than 40k users.