Wednesday, September 6, 2017

The power of a photo in a tweet

Last week I did a little experiment on Twitter. No, I did not have IRB approval.

I wanted to see what the impact of attaching a photo to a tweet would be.

On August 30 at 10 a.m., I tweeted a link to a blog post I had just written about two new types of water—one that supposedly has “activated stabilized oxygen” in it and another that is “living crystal” water. Both are touted as having health benefits.

Here are the first tweet and its statistics.

Click on photo to enlarge it.
Of the 1299 impressions or unique Twitter accounts that could have been reached at 10 AM the time the tweet was posted, 11 (0.8%) people engaged with the tweet.

Three hours later I tweeted about the blog post again using similar wording but this time adding this photograph from the blog post.

Here are the tweet and its statistics. 


As you can see, the number of impressions was almost double that of the first tweet, but the percentage of total engagements, 106/2122 or 5.0%, was 5 times higher.

About three hours later, I tweeted it with the photo again and at 4.8% the third tweet’s engagement numbers were fairly similar to the second one.

Even after the third of three tweets, all of which potentially could have been seen by most of my active followers on that day, the engagements held steady.

A limitation of the study was that the times of the three tweets were not the same. It is possible that the audience significantly differed which may have confounded the results. The post has had 976 page views to date.

However, I concluded that adding a photograph probably resulted in more people engaging with the tweet. It certainly did not hurt.

The floor is now open for comments or questions.

11 comments:

Old FoolRN said...

You proved that a picture is worth a thousand words!

William Reichert said...

What is "twitter"?

Skeptical Scalpel said...

Old, thanks for the comment. I'm pretty sure I didn't prove anything here.

William, google it.:-)

Anonymous said...

You should have controlled the timing. Professional marketers even use schedulers to post on social networks at the times when the content is most likely to be read.

Old FoolRN said...

I should have said you proved a picture probably results in more people engaging. Now all you have to do is wait patiently for that phone call from Sweeden.

Skeptical Scalpel said...

Yes, the time of the tweets is important. All three of the ones in the study were during midday which is when my Twitter feed is busiest.

I hope the Nobel Prize committee can get my number. It's unlisted.

frankbill said...

Any tweets about DHMC shooting? More information here http://www.vnews.com/DHMC-Evacuated-After-Reports-of-Active-Shooter-12449909 And here http://www.wmur.com/article/man-accused-of-killing-mother-at-dartmouth-hitchcock-medical-center-to-face-judge/12234208

Skeptical Scalpel said...

Yes, I tweeted about it twice yesterday but it got very little attention. I guess a shooting at a hospital is so routine now that no one cares.

frankbill said...

One of the questions asked was how was the shooter able to bring a gun into to hospital? Since I have been to DHMC many times I can say they have no way know what you bring in the hospital with you. Maybe this will change but with as many people that go to DHMC it will be a big task to check every one.

As of yet the reason hasn't been given as why this son shot his mother.

Skeptical Scalpel said...

I agree no one knows who is carrying a gun into a hospital. Setting up metal detectors at every hospital entrance (and some hospitals have several entrances) is not practical.

The motive for the shooting is still unknown. Mercy killing?

frankbill said...

From this report don't think it was a Mercy Killing. The affidavit said that Pamela had suffered an aneurism and had been in the hospital’s Intensive Care Unit. She was scheduled for discharge on Friday. http://www.unionleader.com/article/20170913/NEWS03/170919649

Would agree metal detectors would not be practical.

Post a Comment