Google Tools Not Insights Optimized

Willie Louthen-Brown

unread,

Jul 8, 2014, 1:48:13 PM7/8/14

to pagespeed-ins...@googlegroups.com

To whom it may concern -

Over the past several months, we had spent significant time optimizing our sites, and we were able to achieve perfect scores (100 Mobile Speed / 100 Mobile User Experience / 100 Suggestions Summary) for many of our sites.

As we worked on optimization, our team has been frustrated by the lack of speed optimization within Google's products. We have noticed that when we use the tools provided by Google to enhance our sites appearance and functionality - we are penalized within the Google Page Speed Insights test. Some of these penalties are severe - causing immediate failing scores.

I have created several simplified test cases below to illustrate these penalties.

http://acornserverone.com/speed-examples/perfect-test.htm (Test Case #1)

This page has most of the basic features of a "real" Web site, although they are simplified. The site design is also fully responsive as recommended by Google. There is CSS, jQuery, and text, but this page has no additional features. The Insights scores for this site are perfect 100/100/100.

http://acornserverone.com/speed-examples/google-font-test.htm (Test Case #2)

This page is an exact copy of the original page - with the addition of a request for the Roboto font family from http://fonts.googleapis.com. This request was made immediately after the opening <head> statement, following the installation instructions within Google Fonts. With this single request - the Insights scores for this site drop to 79/100/93 - giving this page a failing score for Mobile Speed.

http://acornserverone.com/speed-examples/google-map-test.htm (Test Case #3)

This page is also an exact copy of the original page, with the addition of an embedded Google map. This was done using the code provided by Google maps with no modifications. With this addition - the Insights scores for this site drop to 83/84/88 - giving this page failing scores for both Mobile Speed and Mobile User Experience.

http://acornserverone.com/speed-examples/google-translate-test.htm (Test Case #4)

This page is also an exact copy of the original page, with the addition of the Google Website Translator. This was applied to the site following the installation instructions provided within the Google Website Translator. With this addition - the Insights scores for this site drop to 61/100/84 - giving this page failing scores for both Mobile Speed and Desktop Suggestions Summary.

https://www.google.com (Test Case #5)

This is the front page of Google. The Insights scores for this page are 91/100/97.

As we work to enhance the user experience for all of our sites, we would like to use Google tools and features. The functionality of these tools is important for our clients, and the visitors to our Web sites. However, the test cases above show that we cannot use these tools, and continue to receive good scores within Google Page Speed Insights. We have continued to implement and test these features for several months, hoping that these issues would be resolved over time, but we have seen no progress in this area.

Are you working with the teams that develop these products with a goal to optimize these features for Insights? Is there a timeline available, or a development blog where we could track the progress of page speed optimization within these features?

Our team is happy to collaborate with your developers to help balance the need for fast Web pages with real world design and development needs. If there is anything we can do to help you improve the functionality of the Insights test, please let me know.

Thank You,

Willie Louthen Brown

Senior Developer

Acorn Internet Services, Inc.

Miguel García Sánchez - Colomer

unread,

Jul 8, 2014, 2:43:46 PM7/8/14

to pagespeed-ins...@googlegroups.com

Hi,
I understand you, but pagespeed it's a "suggestion" tool to optimize the web. It's not mandatory and i think that not all the code need to be 100/100, it's not a ranking.

With this said, when i found these issues i try to make "workarounds":

-You can "inline" the font css. It's not a problem in the font, it's better to try to inline and minify little css.
- With Google maps, try to apply an async function but i think that would be better a image (with the size fixed) and a href to the Google map location.
- I wouldn't use Google translate directly. It's not a perfect translation and i think that would be better to use the Google Translate API directly or make redirections with https://support.google.com/webmasters/answer/189077

Best regards,

Miguel.

Dave Mankoff

unread,

Jul 8, 2014, 2:54:16 PM7/8/14

to pagespeed-ins...@googlegroups.com

Howdy Willie.

Thanks for your feedback. Let me answer a few of your closing questions first:

> Are you working with the teams that develop these products with a goal to optimize these features for Insights?

Yes. We regularly communicate with teams within Google about improving the performance of their offerings. Unfortunately, they are limited by the constraints of web-technology as well as their own internal resource availability. Improving speed sometimes comes with costs that they can't afford in the short term. We're still working on it ;)

> Is there a timeline available, or a development blog where we could track the progress of page speed optimization within these features?

I don't know that there is any dev-blog dedicated specifically to this feature set, but there are many resources that Google makes available. I would start here: https://developers.google.com/web/fundamentals/. We also regularly attend conferences and do developer outreach. All teams within Google are constantly soliciting feedback about what developers want.

If you want a faster embedded maps, for instance, I would reach out to the maps team directly. They maintain their own developers blog - http://googlegeodevelopers.blogspot.com/ - have presence on G+ - https://plus.google.com/+GoogleMapsAPI/ - host live office hours - https://developers.google.com/live/maps/ - and more - https://developers.google.com/maps/.

Returning to your original issue, (and speaking specifically of the speed testing). The goal of using PageSpeed Insights (PSI) is not to get 100/100. I can't stress this enough. PSI is a tool intended to help website maintainers understand what may be hurting their page performance. When a point is taken off your PageSpeed score, it is an indication that "something" can be done to make the site faster.

That does not mean that _every_ "something" needs to be fixed. There may be small issues, such as a cache-header on third party content being too short; or a dynamic image that could have an extra 500 bytes saved if optimized. Small issues like those do hurt performance, but not so much that they need to cause frustration. Ultimately, PSI flags them because they _are_ issues, if only nominally. PSI doesn't give free passes just because it would be convenient to do so.

Their may also be larger issues - the three examples you provide highlight this well. The reality is that using any of those three features _does_ slow down a page, Google provided or not. We establish the PSI testing mechanisms independent of other Google services. Then for example, when we put a page with web-fonts into PSI, the scores you provide are what came out. This is because web-fonts really are bad for speed.

We wish this weren't the case. However, as you suggest, every development project has to "balance the need for fast Web pages with real world design and development needs". We couldn't agree more. PageSpeed is a tool that exists to highlight that which slows down a page. If the features of your page are in conflict with speed, it is up to you and your development team to weigh the costs and benefits of the features vs speed. Often times there exist work-arounds (usually loading content asynchronously).

Ultimately, your customers decide what is important. We know from our research that page visitors value speed. If you find that they value a particular feature more(an embedded map, a nicer font, a translation), then do implement that feature, regardless of what PSI is telling you.

I hope this clarifies things a bit. If you have more questions or feedback, don't hesitate to write back.

-dave mankoff

man...@google.com

Thomas Hey'l

unread,

Jul 8, 2014, 3:14:10 PM7/8/14

to pagespeed-ins...@googlegroups.com

Hi Dave!

Now your's is a really, really great answer! +100 for it.

Cheers, Thomas

Willie Louthen-Brown

unread,

Jul 18, 2014, 12:18:07 PM7/18/14

to pagespeed-ins...@googlegroups.com

Dave -

Thank you so much for your reply. This does help me to understand the goals of the PSI team - and also how to better plan our site development for the best possible user experience. We will try to reach out to the individual teams (Web fonts, Maps, Translation, etc.) to communicate our needs. We will continue to use the PSI tool to evaluate the speed of our sites - and continue our work to make the Web work best for everyone.

Thanks

Reply all

Reply to author

Forward