Google+ Case Study: Why did my lunch go viral?

A photo of my lunch went viral on Google+ over the weekend with 28,616 views, 312 +1’s and 54 comments and I had no idea why. The level of engagement on the post goes well beyond the norm of somebody with only 16,000 followers.

Even though none of the main metrics stood out as possible catalysts for the success of the post I still gave each one an arbitrary contribution score based on evidence available to me:

1. Hashtag [10%]
2. Ripples / Re-shares [5%]
3. Comments [0%]
4. Plus Ones [35%]

Hashtag

Hashtag investigation showed the following image views on ten random, variably aged posts in the same stream:

1. 113
2. 902
3. 27906
4. 2214
5. 350
6. 1027
7. 1212
8. 605
9. 387
10. 538

The leading image with 27,906 views belongs to an influencer with 451,466 followers, so it wouldn’t be reasonable to attribute the success of his post to the hashtag alone. The average view count on the rest of the images is 816. If hashtag alone contributed to more than 10% of the post’s popularity it would be a likely case of low ‘competition’ in the hashtag section at that particular time.

Note: Plus ones on each post measured in proportion with image views.

Contribution Score: 10%

Ripples / Re-Shares

Ripples showed a rather low level of public sharing activity. Unless an influencer shared the post in a selective mode, I don’t see how re-shares would have made a significant contribution to this post’s success.

Contribution Score: 5%

Comments

I am yet to see any evidence of comment-based post discovery. Unlike +1’s user’s comment activity doesn’t show on their friends’ social stream, unless it’s a YouTube video.

Contribution Score: 0%

Plus Ones

Getting a full list of +1’s on popular Google+ posts is tricky, as the post itself shows only a sample of 50. Where there’s a will, there’s a way though. After a bit of digging I found a reasonable method of retrieving the +1 data associated with a post.

[Skip the technical part]

Firstly you’ll need to know what the post ID is and the easiest way to do that is to check its ripples:
https://plus.google.com/ripples/details?activityid=z13hclv5nq3nuvx33235xd3h2xqrdtjfz

If this doesn’t work you can use this tool and retrieve the post ID from the list of your recent posts.

Once you have the ID navigate over to the Google+ API explorer and enter the ID and set the collection type:

Note that the collection field has two modes, plusoners and resharers. Set maxResults field to 100.

Output shows user ID, full name, profile URL and image:

To get the next 100 results, copy the “nextPageToken” value, paste it in the and re-submit the query.

The Results

I had a little script made to streamline the whole process and allow me to quickly export and sort the data.

To be honest I was quite surprised to learn that there weren’t any 100K+ follower users involved. The top four users, which all have more than 10,000 followers, have a total of 86,073 followers. This leaves us with nearly as many in the long tail (79,993). This hints at a possibility that the post may have truly progressed organically.

Contribution Score: 35%

Full Data Table:

Other Factors

There are other likely factors which have contributed to high visibility of my weekend lunch:

1. What’s hot? [35%]
   What’s hot algorithm is generally triggered by interest, velocity and popularity. It would be very easy to attribute the success of the post to this mechanism entirely, but I simply don’t have any evidence of this taking place.
   
   
2. Content [10%]
   The type of content in this case played a significant role in the spread mechanism of the post. Technical articles and posts hardly break a 100+ barrier in my case.
3. Timing [5%]
   I call it timing as ‘luck’ doesn’t sound like a legitimate metric. I believe if I posted this photo on a week day (as I have in the past) it wouldn’t have had the same effect.