Important Note: The original post remains below unedited — but Belghast has run the same experiment (with a much larger userbase, likely to be more representative of Twitter at large) and found engagement levels dropped when using the below method.

As noted down in the comments, he found that while what I’ve posted looks great to users — users with a third party app will not have the image pulled through. So they end up with a plain text tweet with zero image of any kind.

So! While it might be worth some experimentation for a few days to see how this works for you, in general? Might just be best off trusting it to WordPress and Jetpack attaching an image for you after all, rather than relying on a meta data pull.

The Jetpack plugin — core to your site if using — has a number of helpful Social Media connection options. And as it turns out, perhaps one that is less than helpful. The general idea of these connectors being that when you publish a post, you write a small written blurb to go along with a link to your post. Then that, along with your featured (or first used) image will be published to any or all of Twitter, Tumblr, Facebook or LinkedIn.

The Jetpack plugin — core to your site if using — has a number of helpful Social Media connection options. And as it turns out, perhaps one that is less than helpful. The general idea of these connectors being that when you publish a post, you write a small written blurb to go along with a link to your post. Then that, along with your featured (or first used) image will be published to any or all of Twitter, Tumblr, Facebook or LinkedIn.

While I’ve not tried LinkedIn — this isn’t the sort of blog where I’d like to share content there — for the most part this auto publish functionality is excellent.

But steer clear of the Twitter one! Don’t use it if you can help it!1

Why? What’s Wrong with the Publish to Twitter Connector?

The tweet on the left was posted manually. The one on the right came from the Jetpack plugin.

The difference at first glance is subtle but it can cost you in post clickthroughs.

I’m not talking about any difference in tag or text content of the Tweet’s either. Although those are important too and Syp has some great general tips for game blogging which can help here.

The problem is how difficult the Jetpack plugin makes it for readers to actually get to your post via Twitter.

The highlights show clickable area that will take someone to your post.

With Jetpack powered tweets, I found a high degree of ‘Media engagements’ — or clicks on the images. But Jetpack is uploading these images as attachments, so clicks on the images will not bring anyone through to your post.

If you publish a tweet manually however — and don’t upload an image attachment — Twitter will convert your link to what you see on the left. The text-only link disappears and your featured image will be pulled and used automatically as part of the clickable area that will take people to your post.

The difference now is anyone clicking the image, or any of the text below it, will be brought through to your post. Much easier for them.

In my limited experimentation so far I’ve seen very positive results by manually tweeting over using Jetpack for Twitter. To the order of 250-400% increase in Link clicks over what was typical previously.

My sample size in both tweets and followers is small though — so I’d be curious to find whether ‘bigger bloggers’ can replicate these results.


  1. It might be difficult not to if you regularly schedule posts for times you’re not up and about, or able to access Twitter.


Gamer, reader, writer, husband and father of two boys. Former WoW and Gaming blogger, making a return to the fold to share my love of all things looty.


Chestnut · August 4, 2019 at 2:06 pm

Interesting. I don’t use Jetpack to auto post, but I often use an image when I manually tweet. I wonder if Twitter has tools to monitor interaction. I’d be interested in doing a comparison.

    Naithin · August 4, 2019 at 2:12 pm

    If you click through to the ‘Details view’ of a tweet, one of the buttons you’ll see as owner of the Tweet looks like a set of little bar graphs.

    Clicking on that you can see the impressions and engagements, and if you drop down details from there you can see a breakdown of the types of engagements too.

    If you have a love of data, you can also dig into — just of course replace out my username with yours — and in the top right of that screen you can ‘Export Data’ and open it into Excel or similar for a pretty full on breakdown of your tweets over the selected period.

    The actual metrics shown on the analysis page itself are interesting, but possibly not the most useful without an export for this sort of comparison.

Shadowz · August 4, 2019 at 2:13 pm

I use it to Auto post, but I don’t put a huge excerpt in the field of text. It has the hashtags, then link, and I’ve never had issues with clicks via Twitter.

    Naithin · August 4, 2019 at 2:17 pm

    I still had clicks coming through with Jetpack posting, but they’ve gone up *markedly* with manual Tweets. I did a quick check on yours and even with the shorter excerpts your images are still separate from the link through to the content.

    So I’d be super curious if you ran an experiment whether you saw similar results as I did, or not. :D

      Shadowz · August 4, 2019 at 9:14 pm

      I’m not worried about all that though. Plus when I set it to a schedule. I can’t be bothered to go in and manually tweet it out. I get more clicks ironically from WordPress reader, and search engines that my social media, even though my Facebook page for the blog gets a few there too. Search engines seems to be my biggest stat with 120

        Naithin · August 4, 2019 at 11:19 pm

        Yep, all that is true. Ditto for me to much of it actually, but Twitter as a referrer has shot up significantly since making this change. That’s all I’m saying.

        If you’re happy with how things are — or, as you say, rely on the scheduler to do the posting for you, then all good!

        But if you were so inclined to improve the click through from Twitter, this is a way to do it. :)

Neri · August 4, 2019 at 4:09 pm

Yes, so much this! I used the auto publisher the first few times and yeah, not even including stats, they just look straight up ugly 😂

I’ll probably still use it occasionally while I’m away, but I generally have those tweets go out during off peak time and then do a manual one much later on. At least they look different enough I guess!

Excellent advice as always 😊

    Naithin · August 4, 2019 at 6:03 pm

    Yeah — the auto posting will still be mandatory for scheduled post announcements and that sort of thing. But where there’s a choice in the matter I don’t think I’ll be using them any more!

Tea · August 4, 2019 at 4:24 pm

Looks like I’ll turn that feature off for future posts! Thanks for pointing the differences out.

    Naithin · August 4, 2019 at 6:05 pm

    No problem! The auto-publishing of tweets is still going to be useful for scheduled posts where you can’t be around, but otherwise it seems a much better idea to post to Twitter yourself.

    The Facebook one seems absolutely fine though, and for Tumblr while I actually prefer the ‘Share as a Link’ appearance better, it isn’t toooo bad if you let the autopublisher in Jetpack do it.

Roger Edwards · August 4, 2019 at 9:08 pm

I recently started adding pictures to my post promo Tweets. Especially for movie reviews. It kind of gives potential readers some sort of context, especially if they’re unfamiliar with the film in question (which can often be the case).

I stopped using auto-generated Tweets a long time ago as they were devoid of any character. A good pithy Tweet can do wonders for “clicks”, just as a good headline can.

    Naithin · August 4, 2019 at 11:16 pm

    Pictures are definitely key attention grabbers — at least for WordPress sites, when a post has a ‘featured image’ (or even just an image full stop) this will come through in the post meta data and Twitter will pick it up.

    The key difference between inserting it to the tweet as an attachment and this though, is that this way the image click goes through to the article rather than just the image.

    But with SquareSpace posts, I have no idea whether would still be true or not!

Rakuno · August 5, 2019 at 1:29 am

I have always posted the links to my blog posts manually to Twitter however I’ve never seem it do the preview thing with the link. I wonder if it has to do something with TweetDeck, as I prefer to use it over regular twitter?

    Naithin · August 5, 2019 at 3:09 am

    By Belghast’s next comment — that may well be the case, that (some?) third party clients are not generating the preview, and only have a text link either way.

    More info in reply to Bel below. :)

Belghast · August 5, 2019 at 2:05 am

So on a counter point to this. I’ve done some testing and while this makes the posts look really good on… for those of us who do whatever we can to avoid using the official twitter posts get left in the dark. This morning I sent out AggroChat manually and it just looks like a block of text and a link in TweetDeck. However it does look cool on So you will likely get better engagement among the official client users but worse for the folks who are using a third party mobile client or tweetdeck.

Like for me personally if I see a link without an image I am apt to completely miss it scrolling through in my timeline. But if there is an image I am way more likely to click on it.

    Naithin · August 5, 2019 at 3:18 am

    Thanks for highlighting this and the quick chat. :)

    This would suggest that results may vary from person to person depending on the makeup of their followers, potentially. Results for me have been pretty categoric that posting manually has had a significant uptick in link clicks (essentially a 1:1 replacement of previous ‘Media engagement’ clicks).

    This suggests to me that most of the base is using either official Twitter or a third party app that does roll out image in the meta data ala Discord and

    I can’t find any *recent* statistics yet to back that up, but only had a cursory look so far. From 2012 numbers it was just under 30% using a third party app. Since then it could’ve gone either way. I would’ve thought more to be pushed off the official Twitter solutions due to the *ahem* ‘improvements’ made over time, but just as equally shutting down some of the more useful Twitter APIs last year might’ve brought some back again.

    I guess what this boils down to is: YMMV! For me, categorical success. But now that I know this issue exists… I wonder if there might be a better / balanced solution to it.

Comments are closed.