Skip to content

Does Jetpack slow down a WordPress website?


jetpackspeed

Update: I added more tests below, based on the comments received. This time based on shared hosting by Siteground.

Can we scare off the eight-hundred pound gorilla in the room already? Are we still buying into the myth?

Look, I get it, Jetpack has a love/hate relationship with most because of the past. It’s time to let the past go. I’m not saying you have to use Jetpack on every site to manage your updates, or to collect traffic stats, but can we drop the, “it slows your site down” rhetoric?

Load time overshadowed by function

Yes, adding Jetpack (like any other plugin of it’s feature set) will add more load time to your website, but nowhere near as cringeworthy as some people make it out to be. I’m no speed engineer, but let’s have a look at my crude test setup:

Base setup

  • WordPress 4.6.2
  • Sanse blog theme — fastest theme I’ve come across, made by Sami
  • Site is hosted on a Siteground dedicated server
  • Used GTMetrix.com website, ran all 4 tests from the same location/browser setup.

Test scenarios

  1. Baseline test, no plugins active
  2. Baseline + Jetpack active; no features enabled
  3. Baseline + Jetpack active; all features enabled
  4. Baseline + 3 Popular plugins for gallery, web stats, and share buttons

Baseline: 74.6KB .5s load time 5 requests

Speedy!

Speedy!

Here’s our baseline test results. Again, just running the theme only, no additional plugins. I set the homepage to a single page, with one line of text.

Baseline + Jetpack w/ no features on: 137KB .6s load time 7 requests

Some weight added

That escalated quickly

Okay, okay, we almost doubled the total page size by adding 2 requests. Which looks to be the stylesheet for Jetpack, and we haven’t turned on any features yet. Yikes.

Baseline + Jetpack with ALL features on: 263KB .8s load time 16 requests

test3-screen

The full Monty

I’ve never turned on ALL of the features of Jetpack — ever.

Like the Math feature? No, never. Anyway, here’s the results of turning all of the features on, just incase you ever found yourself doing so. Which boils down to thirty-three whopping features enabled!

I didn’t turn on the stupid snowflakes — sorry.

So I turned on 33 features and saw the total page size grow 3.5x from the baseline. Nearly 3x more in total requests, and over 50% increase in page load time.

Okay, Jetpack’s looking kinda bloated…but…

Imagine going out and finding 33 plugins to enable what you just did with a single instance to mange? Okay, that might be ridiculous, but what about half of that? Still too much?

Three plugins for three Jetpack features: Gallery, share buttons, and analytics

If you use Jetpack for these three core website features, here’s what you might find by searching the repo for plugins that do the same. I did what most users do, I dropped in the search term and selected from some of the most popular plugins in the repo:

Three plugins for three features – 614KB 1.3s load time 30 requests

test4-screen

We now see a dramatic difference in page speed increase, page size and total requests over the baseline test. That’s twice the loadtime of Jetpack with thirty-three features enabled!

Reminder: this is for only three features that Jetpack provides — if you wanted more, prepare for more load time.

Jetpack speed; it’s relative

Jetpack doesn’t slowdown your website. You might not like Jetpack, but it’s not a bandwidth hog.

For a plugin that does a lot, it’s pretty darned streamlined. When I added the three separate plugins, we saw a much more dramatic increase in load times over Jetpack.

Does Jetpack do everything great? No. It’s a good all-purpose plugin. Other plugins will do their specific jobs better than Jetpack, which is most likely the reason why we saw such a spike on that final test. More features, more code, more overhead (when compared to Jetpack.)

We all good now?

Download the four GTMetrix speed tests in PDF format

Updated tests

Seems like some folks wanted me to drill down a bit more on testing, specifically from a shared hosting plan. Here it is.

Video Overview

GTMetrix Reports for all tests

At the end of the day, I still don’t see Jetpack weighing down a website like some make it out to be.

Listen to this Episode

40 Comments

  1. Piet on November 10, 2016 at 8:45 am

    I really like to believe Jetpack does not slow down sites, but the testing environment you have used doesn’t seem to be the same as the average Jane or Joe would have.

    1. running on a Siteground dedicated server
    I fail to see how that is comparative for most users, especially since you seem to be focusing on bloggers by your choice of theme?

    GTmetrix has quite a few variables for setup. Without knowing which one you have used, this test is basically useless as there is a huge difference between “unthrottled” and “56K dial up” (both extremes at the far end with 3 more options in between).

    How about doing the same test with Jetpack turned on (by default it activates a whole bunch of painfully stupid modules) on a Siteground hosted server, let’s take the GrowBig package and for GTmetrix use the DSL connection for Chrome via Dallas.

    Then it will be interesting to know the results and I bet that those will be far less in favour of Jetpack.

    • Matt on November 10, 2016 at 12:26 pm

      Piet, thanks for stopping by. Pretty sure I prefaced the test as “crude.” So now that that’s out of the way…the hosting environment doesn’t impact the total load size and number of requests — that will stay consistent on a cheapo-plan all the way to an Enterprise Pagely hosting account.

      Sure GTMetrix has it’s flaws, but again, crude test. I feel like this test does do the average Joe or Jane justice because the variance isn’t going to be enormous when they bring it to their own hosted sites. I’m also not testing hosting accounts here, what you’re eluding to is a cheaper plan will load slower, which might be the truth, but not to blame on Jetpack (or any other plugin.)

      • Piet on November 14, 2016 at 5:19 pm

        I must have overlooked the word “crude” that’s placed somewhere in the text.

        Pretty sure that “prefaced” usually means something different than adding a single 5 letter word to the last sentence of the 3rd paragraph though.

        With an excuse like that I cannot help but think that the title is mere clickbait, which is strange, because of all people you certainly don’t need that!

        • Matt on November 17, 2016 at 5:28 pm

          new results posted above 🙂

    • Neil Murray on November 14, 2016 at 1:21 am

      I think this would be a much more realistic test IMHO which I would be really interested to see the results from.

      • Matt on November 17, 2016 at 5:28 pm

        new results above, Neil.

  2. Nate on November 10, 2016 at 10:34 am

    An interesting follow-up would be to get some actual website setups using or not using Jetpack and then swap out their plugins for Jetpack features and vice-versa. Then perform tests in both states.

    That would give people a real tangible look at the cost of running poorly optimized plugins.

    An interesting project (maybe for PluginTut) to do would be performance comparisons between similar plugins. Like for contact form plugins, you could set up a site with a basic form and a complex form using 3 or 4 different plugins, and analyse each plugin’s impact on site performance in relation to their feature-set trade-off. Same with gallery plugins, event plugins, etc.

    • Matt on November 10, 2016 at 12:27 pm

      Hey Nate! Grand idea, mate.

  3. Sam Hotchkiss on November 10, 2016 at 12:07 pm

    Hey Matt!

    Thanks for taking a fair look at Jetpack and performance, it’s something we’re working very hard on improving. We’ve made huge strides in the last year and have some exciting further improvements in the pipeline for 2017, so I hope this becomes an annual post!

    Best,
    Sam Hotchkiss
    Jetpack Team Lead

    • Matt on November 10, 2016 at 12:27 pm

      Thanks, Sam, good work 🙂

  4. John Keegan on November 10, 2016 at 11:00 pm

    Ok but you are not measuring its impact on the sever itself. On the CPU resources it consumes or the increase in size of the php processes or the increase in MySQL queries.

    Sure impact on the load time of a single web page is marginal. But really that’s not at all the issue.

    If the use of a plugin such as Jetpack increases the resource consumption of a base WordPress install by 2x (which is probably a conservative estimate) then your particular install (shared hosting account, server, or whatever you host at or on) will be able to scale half as much as it otherwise would have.

    So to put it another way, you hit the ceiling of your server setup at only half the traffic level you otherwise would have.

    Might not be an issue for you if you can increase your capacity easily or cheaply. But if increasing your capacity is not a simple matter or is not something you can afford, then disabling plugins such as Jetpack can give you additional headroom to handle a traffic increase or to handle more concurrent sessions.

    Remember this is “all else being equal.” I.e. You may have other resource intensive plugins installed which make use of (or removal of) Jetpack less impactful.

  5. Elaine Mingus (RadicalChristianWoman.com) on November 11, 2016 at 12:44 pm

    Matt,

    Have you done the test the other way around?

    I took my heavily loaded (but optimized site) to GTMetrix at a recent blog conference. It scored at F.

    As soon as I installed ONLY Jetpack, my score went to a C.

    Just wondering, because I was 100% sure as a coder/developer for over 12 year I could do better at optimizing than Jetpack (considering that it has other plugins that I don’t use), but this test proved me I was wrong.

    I continued to turn off and on plugins to see if it affect my GTMetrix score, but non affected my score nearly as much as Jetpack did (positively).

    My guess is that it probably depends on the site. My site is very image heavy.

    • Matt on November 18, 2016 at 9:04 am

      That’s great to hear! Probably because Photon helped improve the load time of your website.

    • Luke Cavanagh on December 9, 2016 at 2:20 pm

      Score or grade is not as important as actual site load time. WebPagetest is normally a better option for site load time testing.

  6. Acil Kredi on November 14, 2016 at 7:52 am

    Jet pack is not good to use if hosting has low CPU and REM limit!

    • Matt on November 18, 2016 at 9:04 am

      Hi Acil, I updated the post with testing on a shared hosting plan from Siteground.

  7. Antony Agnel on November 22, 2016 at 5:54 am

    To be honest, the features of Jetpack easily outweigh the so called ‘Jetpack slows down your WordPress site’ myth. Jetpack has got a whole lot of features to be integrated into a single plugin. I consider Jetpack as a boon because it saves me the job of hunting for & installing multiple plugins for various purposes such as Stats, Social Sharing, Image CDN, Bruteforce-protect feature, Image Widget, Social Widgets, Uptime Monitoring, Custom CSS, etc.!

  8. James McInally on November 27, 2016 at 4:40 am

    Hi Matt, as a relatively non-technical person who simply wanted to find out if Jetpack increased or decreased a WordPress website, this article and your research is is a boon. I’ve tested Jetpack on a coupke of our website which as pretty slow (on on the same hosting company and same hosting package) and both slowed down a little. However, rankings went up in Google, which I believe is due to trust factors in Google by using the plugin as well as managing comments etc.

    Ultimately, this is what I (and most website owners wish to happen).

    I haven’t tested this empirically but something I will do when I have time, unless someone else has?

    Best
    James

    • Matt on November 27, 2016 at 9:57 am

      Thanks for commenting, James. Glad things are working out for you using the Jetpack plugin. I’d imagine you saw ranking increases because of the pagespeed timings improving via photon, and some of the search optimization features like Sitemaps and Enhanced Distribution…or it was all coincidental 🙂

      Either way, great to hear!

  9. Jigar Shah on December 16, 2016 at 4:49 am

    First of all, I would like to thank you for the showing extreme Jetpack Performance Review and it looks perfect to understand the Jetpack strength. Really brainstorming research but I also agree with Antony about Jetpack. It’s really cool option and we can save lots of time by using this.

  10. Mich Chavez on December 19, 2016 at 1:43 am

    It’s great that you have shared your test Matt and I agree that Jetpack is not the main reason why a website slows down, yes it can be consider since plug-ins eat-up loading time but overall it’s not the main reason why it’s slows down. It’s great that you have share this one Matt, it can also be the basis of continuous improvement .

  11. Tamhas on January 7, 2017 at 8:33 am

    I really liked jetpack features, but recently when really focusing on speed, I find that turning off jetpack reduced load time of my pages from about 1.5sec down to about 0.5sec.
    My setup is Vultr $5 VPS running Ubuntu Nnginx with CGI fastcache plus free Cloudflare.

  12. Wordpress customisation on April 29, 2017 at 3:02 pm

    Probably because Photon helped improve the load time of your website…That’s great to post!

  13. kv famana on July 15, 2017 at 4:25 am

    Hi,
    After activating Jetpack, I am getting ‘404 Not found’ error and it adds ~3 sec to the total time.
    http://mysite.com/wp-content/plugins/jetpack/css/%22https:/fonts.googleapis.com/css?family=Open+Sans%22

    How can I fix this?

  14. Gino Côté on July 20, 2017 at 12:14 am

    I do not need a speed test, just normal browsing to show without a doubt that Jetpack slow down my website.
    A test with a WP plugin who show plugins speed show that Jetpack slow my websites by 30%

  15. David Ranck on July 26, 2017 at 10:08 pm

    Hmm. Old post but I found it because Jetpack great increased page load times for me. I’m using the Avada theme With Jetpack enabled the hero image took about 1 full second to load and then the menu would pop in about a half second later. Very noticeable delay. Deactivating Jetpack let the pages load in less than .5 seconds. I didn’t use Y-Slow to try to figure out the root cause. Anyway, I’m not using it for now.

  16. Bryan Veloso on August 22, 2017 at 10:57 am

    Ok, I’m deactivating Jetpack completely. Haha!

  17. jehzlau on September 29, 2017 at 8:43 am

    On my blog, my GTmetrix score is 100. After activating Jetpack, it went down to 97. Not really a big issue, because the actual load time using Chrome or Safari seems to be the same. The only difference is the score in GTMetrix. And great thing about Jetpack is I can use the WordPress app to update my blog. That’s a great + despite of the millisecond load time difference.

    PS: Currently Jetpack is disabled in my blog, because Im a sucker for 100. Haha!

  18. Bill on October 6, 2017 at 7:56 pm

    Jetpack was slowing down my woocommerce sites. Took 3 sec to load. After deactivation they all work like a charm < 1 sec. Guess it happened due to Publicize

  19. Craig Constable on October 19, 2017 at 6:45 pm

    I have a WordPress website with WooCommerce. I wasn’t overly impressed with the speed of the site or the WordPress admin interface but it was ok.

    I then installed ‘Woo Floating Cart Lite’ which allows you to update the cart using ajax calls and this is where I noticed the speed was horrible. It would take anywhere from 4-13 seconds to update.

    I installed ‘Query Monitor’ which was showing around 160-170 queries and 69MB memory usage.

    I then deactivated ‘JetPack’ and the queries went down to 60-70 and memory down to 53MB.
    The WordPress admin interface was super snappy and fast and the update cart via ajax was now 0.6-1.2 seconds.

    The funny thing is I had hardly any features enabled in JetPack, it was mainly the statistics and security features that were being used.

    Huge improvement from stopping one plugin. I read a lot about JetPack being well built and perhaps it is fine if you have a great hosting environment. But the improvement was huge in my case.

  20. Tai Chinh on November 2, 2017 at 8:39 am

    I am a speed fan. I tried 1 time and don’t use any more as it made my website slower

  21. David Sim on November 27, 2017 at 11:58 am

    I’m another WooCommerce / WP / Jetpack user – I was only using the security and website stats features.

    Load times went from 3 – 5 seconds per page to <1 second when I disabled Jetpack. I still have Wordfence loaded.

    It could be an interaction between Jetpack and the plugins I have to support WooCommerce (I have several) but it seems to be making calls or conflicting on every page load.

    The MySQL database on my shared hosting is relatively slow, so it may be the hugely increased number of queries that Craig identified in an earlier comment. Either way, it took me a while to find the problem – it's staying off!

  22. Richard on December 14, 2017 at 11:05 pm

    It sounds awful. It’s supposed to improve performance but does not.

    Seems like false advertising on the page.

    More complexity and it doesn’t deliver. Why even touch it?

    No tanx.

    But thanks for everyone for contributing here. Very helpful!

  23. Rustyn on January 1, 2018 at 12:22 pm

    I have to disagree. We launched our new site last night and within 30 minutes the resources eaten up by Jetpack killed the site. Once we deactivated it, it’s run like a charm ever since. We used the same testing site as you, and markedly different results. I really only used it for stats and autoposting features, But it is terribly bloated and a true resource hog.

  24. Deepak Gupta on January 22, 2018 at 1:31 pm

    Thanks for sharing this theory. Though, I still see that Jetpack hurts the website speed intensively!

  25. Joe on February 8, 2018 at 9:02 am

    I wish I could agree with your post, I’ve been using jetpack for years, and a recent update made my site return a 500 internal server error

    I had to deactivate all plugins (moving them out of the folder)

    Put them back in the folder, and activated them 1 by 1.

    The culprit was Jetpack.

    Then, with Jetpack off I reran my speedtests and my page speed increased by 3-4 seconds on some of the larger pages.

    This is a shame as I love the Publicize and Gallary options on jetpack.

    If Yahoo Pipes was still around I would get back to using that for my social sharing, but I haven’t found anything as good 🙁

  26. Deepak Gupta on July 18, 2018 at 7:32 am

    Even I agree with you. I’m using the JetPack on my website, but I’ve removed the JetPack CSS and one more file using the function.php file.

    Now it works good. Just using traffic status module!

  27. TelFiRE on December 2, 2018 at 3:44 pm

    JetPack literally makes pages take an extra 5-10 seconds. How can you say that doesn’t weigh it down? If other plugins made it take that long, they were bloated too. No page should ever take longer than 2 seconds to load, and with JetPack no page will ever load in less than 5 seconds.

    • Matt on December 20, 2018 at 10:23 am

      This post is 2 years old.

Leave a Comment