IG v.s. FB: A simple A/B testing (paired T-test)

Since I want to post my blog article in Traditional Chinese, I wonder whether Instagram or Facebook is better to post this kind of article. Therefore, I use the connection between IG and FB to post the same article simultaneously.

The assumption or the setting of posts are:

The posts are all in traditional Chinese.
The number of words in each post is below 350.
It must posts follow by a image.
The number of friends: FB - 981, IG - 180

FB:

IG:

Because I post the same articles in IG and FB at the same time, I think I can do the simple A/B testing. The method I use here is paired T-test. I just use the Paired T-test online calculator. The data I collected is the number of like in each post. The latest post is at 5 days ago.

The data is here:

Post	Facebook	Instagram
Infinitesimal (book review)	7	3
Blue Giant (movie review)	8	5
code interpreter introduction	21	7
code interpreter application	13	3
Flash (movie review)	1	2
Mission Impossible 7 (movie review)	6	3
Barbie (movie review)	8	7
Across the spiderverse (movie review without spoiler)	9	4
Oppenheimer (movie review without spoiler)	21	9
Across the spiderverse (movie review with spoiler)	5	8
Oppenheimer (movie review with spoiler)	6	2
Asking a queation between God and AI	8	6
The best way to get the answer from web is to post a wrong answer	2	7

Using online paired T-test calculator, and my hypothesis $H_0: FB = IG$. The result is following:

Parameter	Value
P-value	0.03079
t	-2.4464
Sample size (n)	13
Average of differences ($\hat{x_d}$)	-3.7692
SD of differences ($S_d$)	5.5551

Since p-value is 0.03 < 0.05, we reject H0 at $\alpha=0.05$. The result is FB is not the same as IG

However, I cannot know which is better from this result, so I change my hypothesis $H_0: FB <= IG$. The result is following:

Parameter	Value
P-value	0.0154
t	-2.4464
Sample size (n)	13
Average of differences ($\hat{x_d}$)	-3.7692
SD of differences ($S_d$)	5.5551

Since p-value is 0.03 < 0.05, we reject H0 at $\alpha=0.05$. The result is FB is not the same or smaller than IG

Therefore, The conclusion is: The article with a image is better to post in FB than IG

By the way, I also have some findings:

If sharing a post than writing a post in FB, the number of like drops dramatically
If using a series of infographs in IG, the number of like is the same as the most like normal post. However, it still cannot show the difference between normal post and inforgraphs. (By the way, my next side project is automatically creating infographs for IG)
If you want to collect most of like in IG, the best way is using story, not post. Maybe that is why TikTok is popular nowadays...