Instagram: What Happens When Instagram Goes Down

25,554 Followers

I tried to post a picture. Failed to upload. I tried again. Failed to upload. Tried again. Finally uploaded!

Checked it 20 minutes later: 19 likes. Checked twitter to see if there were any #InstagramDown posts and:

In honor of the July 3rd Instagram outage, let’s take a second to talk about how you can find out if Instagram is down, if you should still post if you can, and what ‘being down’ actually means.

1. How can I identify when Instagram is down for a lot of people and not just me?

Sometimes you have a bad internet connection, and sometimes it’s really just not you. But how do you tell the difference?

My first step is usually to go to Twitter to see if people are complaining about it. So many people do this that it’s almost a joke at this point.

There are also several accounts on Twitter that cover this topic:

My second step is to go to Down Detector which is a website that allows people to report website errors. There are always some errors being reported, but if there are thousands of errors being reported suddenly then it’s probably a bigger problem.

2. Okay, Instagram is down, how does this affect posting/engagement levels?

Let me tell you, an outage does not increase engagement

The screen shot on the left was posted (unknowingly) during the outage. It was also my first sign that something was wrong. For two reasons:

1. Typically the first ten minutes of a post have the most rapid engagement, for my account that means around 10 likes/minute.
   By the time this screenshot was taken I would have expected closer to 150-200 likes, not 19
2. The first ten minutes also usually has the highest ratio of impressions to likes, then over time the engagement rate usually closes out around 8% (impressions are the 1014 and 196 numbers in the screenshots)
   Having 1,014 impressions and 19 likes is a 1.8% engagement rate, which is highly abnormal behavior (see in the right screenshot that initial engagement is 37%, this is more typical).

As soon as I realized there was an outage, I archived my post and decided to repost the next day when people would be more likely to actually see the content.

24 hours later the post is at 2.5k likes, an engagement of 7.6% and has been performing well. I doubt that it would be doing as well if I hadn’t decided to archive it.

3. Great, don’t post during outages. Got it. But why do outages happen in the first place?

Unfortunately (and fortunately) our virtual worlds rely on technology in the real world. This means that the things that lets you post photos, comment, upload stories, and keep yourself distracted at work all rely on real-world-infrastructure operating correctly.

Here’s a Kindergarten-level explanation on the tech behind these websites

All applications (web, mobile, desktop, whatever) are created by developers (read: humans) who code them to do what they’re supposed to do. Think of it like writing an essay for machines to read.

These applications are then put on servers (read: really big computers located around the world) and are made accessible to everyone through a variety of technologies

Why the technology matters

There are a couple overarching reasons why a website that’s active might be having issues:

1. A server crashed (or became inaccessible)
   Servers are just big computers. Much like people, they can get overwhelmed if too many people are trying to access them at the same time. Unfortunately, this isn’t always something that companies can prevent (especially if they don’t own all of the technology that they use).
   From our perspective it’s similar to dialing a phone number that doesn’t exist, you likely won’t get a response and you won’t know why
2. Someone made a bad update
   Developers are people too, and sometimes we make mistakes. Someone might have deleted a line in the code they shouldn’t have, or forgotten to run a test to make sure everything works.
   It takes time to fix these kind of problems. You have to first find them, then figure out what the solution is, then actually fix it, and then share that fix with the rest of the world.

Why is this worse for big companies?

The bigger you are, the more that people are paying attention to you and the longer it takes to fix things.

Scenario 1: My website can be down for a day and no one in the universe will care about it but me (I know this because it has been down for a day and no one complained).

If Instagram is down for 30 minutes they’ll have 10,000 tweets and damage control to do. People are asking them what happened. people are angry, people are foretelling the company’s utter demise in the next 6 months.

Scenario 2: If my website is down, I can fix it in my own sweet time (because of Scenario 1), and the fix is usually as simple as re-uploading a file and I’m good to go.

One person, one file, one environment, no problem.

If a website like Instagram or Facebook goes down they have an issue because they have massive amounts of code that is touched by dozens of people (I assume) every day that’s going through multiple environments (read: it has to be loaded on several computers) to eventually get to the state that you see it in.

The level of effort there goes from finding a needle in a pincushion to trying to find a needle in a haystack while people are yelling at you about how silly it is that you can’t just find it already. I am not envious of those teams.

TL;DR

First of all, if you think that there’s an Instagram/Facebook/Social Media outage, there are a few places you can check to verify that it’s not just your internet being awful. If you’re someone who cares about engagement rates, you should probably not post during this time if you actually want people to see your content. Finally, try to give tech companies a little grace during outages. The companies are made out of people who are trying to do their jobs and make mistakes, they aren’t trying to ruin your lives.

This is the best technical explanation I’ve seen so far for the July 3rd outage.