Updated: July 2020
There are no affiliate links on PagePipe.
The WP Speed Guru promise: Pay $69 and learn how to make your site lightning fast. Baloney.
So we decided to learn how deep these hidden secrets are. We suckered and paid. Here’s what you get in this vacuous series of videos:
1Intro and Case Study
This is 3-minutes demonstrating WP Speed Guru’s homepage speed. That includes a screencast of GT Metrix download history. WP Speed Guru claims the speed improvements came from dropping from 60 to 15 requests. The emphasis throughout the videos is on reducing HTTP requests. This doesn’t always result in speed improvement as WP Speed Guru demonstrates later.
And WP Speed Guru considers scores important. Scores aren’t significant to us in speed testing. Only milliseconds of load time is important. We know the number of requests isn’t indicative or proportional to speed improvement. There are various reasons for this such as Time To First Byte (TTFB) and HTTPS/SSL handshaking. These are server overhead measurements. The only way to overcome a bad TTFB is to move to a different server or host. What is bad? Any TTFB over 750 milliseconds is bad. So what is a good TTFB? Anything under 300 milliseconds.
Speed results reported in these videos is using GTmetrix online testing. To get TTFB numbers in milliseconds, test with ByteCheck.com or WebPagetest.org.
2Free Testing Tools
This segment is 14 minutes long. GTmetrix online speed test show speed changes on a demonstration restaurant site. WP Speed Guru recommends a paid theme vendor. The site under test has wasteful features we’d never use on a speed site. These include a homepage slider and over-the-top animation. Other things are a YouTube video, accordion navigation, Google Maps, and a Blog section. WP Speed Guru homepage speed test is 2.6 seconds. 1.13-megabyte page weight, and 47 requests.
WP Speed Guru recommends 6 tests in a row. This is something we often recommend, too. WP Speed Guru talks about uncontrollable things like TTFB and Google Font load times. These fluctuate outside web developers or site owners’ control. That is often true. So no disagreement there. But have we gotten our $69 worth of valuable speed secrets? Not even close. This is all common knowledge shared on PagePipe and other blogs for free.
WP Speed Guru then demonstrates a speed test on Pingdom.com with the same test page. A two-year-old test is evident since Pingdom changed its interface long ago. WP Speed Guru also cherry-picks a location close to the server. This, of course, will improve the results. WP Speed Guru says grades determine if things will fall into place for speed. Everything WP Speed Guru suggests is score-based troubleshooting. WP Speed Guru disclaims scoring practices by saying, “This may not apply to your site. Some score suggestions need to be fixed and some won’t.” WP Speed Guru doesn’t tell us which suggestions matter.
WP Speed Guru then shows a Pingdom waterfall for the test page. WP Speed Guru checks for red 404 errors indicating broken links. This is so elementary and basic common sense. Broken links are bad for many reasons – speed delays being one of them. WP Speed Guru looks at requests from server or external servers in the waterfall. But again doesn’t make any suggestions on what to do about these except “Repair them.”
WP Speed Guru describes how a CDN works. A basic explanation is free on Wikipedia.
WP Speed Guru then runs WebPagetest.org. WP Speed Guru shows test server locations and browser selections. And fiddles around with selectors. After showing WebPagetest.org options, WP Speed Guru says, “Pingdom is most accurate. But the best test is checking with your browser.” Implying but not saying, “use a browser timer extension or addon.”
WP Speed Guru talks about deviation in results depending upon the test chosen. But never quantifies what to expect. At PagePipe, we use Pingdom for best-case load-time scenarios. And WebPagetest.org for worst-case scenarios. The test results in milliseconds rarely match. Pingdom isn’t a better test. Only a faster one. It’s less comprehensive.
Repeatable testing is always on WebPagetest.org. But you’re comparing relative change – and not absolute change after making improvements.
Is any of this secret? No. This is common knowledge.
3How to Set Expires Headers
The next section and 4 minutes is about How to Set Expires Headers
The goal stated is to remove the red “F” failure flag on WebPagetest.org caching and turn it green. WP Speed Guru’s method is not recommended by us. WP Speed Guru shows this method to make things seem complex or frightening. The solution is simple.
WP Speed Guru demonstrates rewriting HTaccess files via FTP or Cpanel. What? We use a simple no-coding plugin. WP Speed Guru examines server code changes and edits via Apache Cpanel. And edits the htaccess file for expires header with copy and paste. WP Speed Guru downloads to edit and then reuploads. Very clunky. But the big surprise is WP Speed Guru doesn’t tell the “magic code” that is added. This is the hardest way to fix this minor speed problem. A plugin solution is easier and requires few settings and causes no speed drag. We sense this step is hand-waving. Is it secret? No.
WP Speed Guru retests on GTmetrix. 2.4s, 1.13M, 47 requests. Scores change from red to green – but speed improves by only 200 milliseconds. This amount is not uncommon for TTFB fluctuation on a host server like SiteGround or WP Engine. This secret is insignificant. Does PagePipe add a plugin anyway? Yes. But only as a matter of principle. It doesn’t improve UX or SEO in the slightest. It’s a vanity speed metric.
4Remove Query Strings
This section is 2-minutes long. It demonstrates installing a single discrete plugin with no settings. The plugin is “Remove Query Strings From Static Resources.”
The test results are the same. GTmetrix retests: 2.4s, 1.13M, 47
It improves the score but not the speed. We repeat: score improvements are meaningless.
Are this free plugin and others (that do the same thing) a hidden secret? No.
5Free CDN with Cloudflare
This section is 21 minutes and is about how to set up Cloudflare CDN. Boo! We never use this third-party service. Ever. If it’s on a site, we remove it first. It’s a band-aid to mask real speed problems. WP Speed Guru even demonstrates later how CDN is unnecessary if you use caching plugins. Cloudflare is a complicated setup for a free account. This section is full of BS about the theoretical benefits of CDNs.
This section is a waste of 21 minutes on which you’ll never get a return.
You can duplicate minification and concatenation on Cloudflare with free no-setting plugins. WP Speed Guru tells us CDN can break pagebuilder plugins. And needs special settings for selective deactivation of minification.
If you want Cloudflare CDN, all this information is available on their website. It’s not secret.
But PagePipe doesn’t recommend free CDNs. They produce 503 server-not-found errors. A broken page or site results. Not worth it.
6Great Free Caching Plugins
This section is 10 minutes of detrimental garbage.
He installs Wordfence Security since it has caching. This slow multi-function plugin causes delays of 250 milliseconds globally to pages. Slower speed overhead. It can cause unnecessary server resource overages. A bad plugin choice.
With caching, the initial page load remains unchanged. But repeat visits with primed cache yields homepage results of 0.9s, 1.12M, 47 requests. What percentage of your traffic are first-time visitors? We bet it’s the majority. They don’t benefit during visits when the first impression is most critical.
Our note: Minification often breaks sites. So don’t use it unless needed – and usually, it’s not needed.
WP Speed Guru then installs WP Fastest Cache.
1.4s, 1.18M. 39 requests. Even though this is slower. The author thinks it’s a speed win by reducing the number of requests by 8. Ridiculous conclusion.
7How To Optimize Images
A 6-minutes tutorial of how to optimize images manually in Photoshop. WP Speed Guru recommends not using extra plugins for image optimization. While we do hand optimization, it’s bad advice to not use image compression plugins.
Is any of this knowledge secret? No. Free tutorials are all over the web.
How to do image optimization is available elsewhere on the Internet. WP Speed Guru shows how compression didn’t hurt the image quality. This is common knowledge. Not secret.
For 8 minutes, WP Speed Guru demos an “About” page with an embedded YouTube video.
Before test on GTmetrix: 2.9s, 945k, 31 requests
WP Speed Guru recommends using a lightbox plugin. Video then opens in the lightbox and plays. WP Speed Guru used an unidentified paid X-theme plugin.
New results: 1.3s, 507k, 24 requests
Achieve these same results with a free lazy load for video plugin. Secret? Nope.
4-minute explanation of a map on a contact page.
This shows how a developer used a still image of a static map. The still-image link opens a new window with the Google Maps page. Again, this clever offloading trick isn’t a secret.
2-minute gloss over. WP Speed Guru recommends offloading social media to a gateway page. 10 requests are from 2 buttons. But how to do this isn’t explained.
WP Speed Guru examines requests added by WooCommerce – even without features. This is global loading or site drag. WP Speed Guru uses FTP access to edit the child theme’s functions.php file. WP Speed Guru copies and pastes some code into the file. But doesn’t give us the code. Crazy. Get our free PDF instead to tune up WooCommerce.
WP Speed Guru saves 7 requests but the WooCommerce page speed is slower. Lame demonstration.
Make Fewer HTTP Requests
ON: 2.4s, 1.17M, 33 requests
OFF: 4.0s, 1.18M, 41 requests
Big deal. WP Speed Guru showed caching reduces requests. But still doesn’t get the page under the 2-second performance goal.
These load-time failures are never addressed.
Was this a secret? Not ever.
WP Speed Guru then installs instead Autoptimize plugin and
glosses over settings.
Results are then: 2.9s, 1.04M, 23 requests. WP Speed Guru claims speed victory because the requests are fewer. But speed still isn’t under 2 seconds. Absurdity.
The Final Speed Test
A final 3-minutes demo of Pingdom using different server locations.
1s, 1.32M, 29 requests with unprimed cache.
809m, 1.2M, 29 requests after caching.
This demonstrates that speed is non-geographic influenced. It also shows how Pingdom always gives best-case results for cherry-picking speed. Really? Now he uses Pingdom to measure?
Inconclusive. These recommendations aren’t secret – all are available for free.
WP Speed Guru subtracts your $69 tutorial fee if you buy tuneup services. Services for the confused.
Useful Performance Plugins – links given. They are only a few and of minor consequence.
Knowledge Base – a link to WP Speed Guru’s blog.
WP Speed Guru evaluates by the number of requests. This is wrong. It’s load time in milliseconds that matters. Save your $69.
Original article © January 2020
If MailChimp messes up delivering your free report, email us and we’ll kick their monkey butt.