How To Reduce Your Website Load Time To Less Than 2 Seconds

Google is continuously evolving the 200+ signals they use to rank web pages. It is becoming more apparent with every update that your website needs to be on top of its game from a technical point of view. At The Whole Caboodle we acknowledge this fact and have decided to share our knowledge to show the service we can offer to you. (smile face)

We decided to take a look at website load times and how this may help your sites rankings and performance. Below we will look at

• How we improved one of our clients websites load time
• What effect did it have on their website's rankings and performance?

Why Should You Improve Your Website Load Time?

The main reason why you should improve the load time of your website is to provide a good user experience. We’ve all been there; when you click on the website you want to view but instead you’re presented with a white screen, which seems to stay there for what feels like an eternity. Even though only 10 seconds have gone by, this is far too long to wait for a webpage to load. Internet users of 2015 want instant access to information they are searching for, and it's only going to get worse as we go into 2016.

So what is an acceptable webpage load time? Have a guess and see what you think! Would you say it's between; 2-3 seconds, 5-7 seconds, 8-10 seconds, 13-15 seconds or 20-30 seconds? (Great if we could add a data capture here of guesses or add HotJar poll) Studies have shown that users are not prepared to wait longer than 2 seconds for a webpage to load which doesn’t seem long but these days, users are very impatient.

Here are a couple of facts for you;

• A 1 second delay in a webpage load time can result in a 7% reduction in conversions.
• If your e-commerce site earns £100,000 per month, a 1 second delay for a user could potentially lose you £16,000 in sales every year - every second counts.

View this 'Every Second Counts' Infographic for more information on webpage delay time.

With that potential loss in mind let's look at what we have done for one of our client’s websites load time.

Our client, www.cartyres.com is a new player in the tyre industry. We thought we would benchmark cartyres.com against some of its major competitors to see where they sit with load times against the main companies providing tyres online.

This is what we found:

Car Tyres Desktop Site Speed and PageSpeed Insights Score;

PageSpeed Insights Score is very low even though the Pingdom tool gave the site a load time of 1.40 seconds.

ASDA Tyres Desktop Site Speed and PageSpeed Insights Score;

PageSpeed Insights Score is at 76. This result is ok but could do with being improved. Pingdom tool gave the site a load time of 2.19 seconds. Remember the above comment about a 1 second wait and how it can affect your conversions.

Blackcircles Desktop Site Speed and PageSpeed Insights Score;

PageSpeed Insights Score is at 53 which is very low. The Pingdom tool gave the site a load time of 2.79 seconds. Remember every 1 second delay counts.

We benchmarked our client’s site against their competitor's using video imagery which WebPageTest.org provided.  Take a look at the results here:

What Did We Do?

Google are very helpful and provide you with all the data you need to improve your webpage load time using Google's Page Speed Insight tool. We plugged our clients URL into Google's tool and within 3 seconds we had our answers. Google lay out exactly what needs to be done to improve your load time and display their PageSpeed Insights score.

We needed to optimise the below points, and we have also included what needed to be done;

• Enable Compression – You should be looking at your file size that needs to be transferred when loading a webpage. If you haven’t already enabled gzip, you should as this can reduce the size of a file transferred by up to 90%!
• Image Optimisation – Images can have quite large file sizes, especially if you are working with high resolution images for your website as many do these days. You should look at reducing the size of the images by either putting your images through an image compression tool before adding them to your website or installing a plugin if you use WordPress.
• Eliminate Render Blocking – This is one of the main issues Google flags, the blocking of JavaScript and/or CSS files. To solve this problem you should look at in-lining your CSS or JavaScript code directly into the HTML document where possible.
• Minify JavaScript – JavaScript is widely used across the internet. We advise that you look at reducing the use of unnecessary JavaScript code such as shorter variables, function names, formatting, etc.
• Minify HTML – This is the same as above. You need to streamline your code where possible, remove bloated code, use shorter variable and function names etc.
• Minify CSS – This is the same process as the above just in your CSS file.
• Leverage Browser Caching – To load a website it can take many round trips from user to server to load the webpage. If your website isn't changing every day you should look at adding a caching policy of around one week. The result is once the site is rendered on a user's browser its cached ready for their next visit and it won't need to take so many trips to load the page.
• Prioritise Visible Content – This refers to the above the fold content. Limit the size of the data needed to load the above the fold content.
• Reduce Server Response Time – The recommended server response time is 200ms. We would suggest talking to your hosting provider to look at this issue.

What happened after we had optimised the above was quite amazing, the results are below:

Car Tyres Desktop Site Speed and PageSpeed Insights Score;

PageSpeed Insights Score is now looking positive at 87/100. This has improved the load time by 0.19 seconds, decreasing the overall time to an impressive 1.21 seconds. This means that our client's website is well under the recommended 2 second load time. You can even see the reduction in the Page size which has reduced by 0.3mb.

Not only has the load time and page score improved we have seen an improvement in Google Analytics. Comparing analytic results from 30^th September – 7^th October vs 28^th October – 4^th November.

Average time spent on site has had a slight boost from the website load time improvement. The biggest improvement in analytics is the Bounce Rate. It has reduced by 18.99%, just from improving webpage load times.

When comparing the time periods in Google analytics Site Speed overview, we saw a much greater improvement.

• Average Page Load Time - Improved by 18.42%
• Average Domain Lookup Time - Improved by 94%
• Average Server Connection Time - Improved by 84%
• Average Server Response Time - Improved by 12%

After double checking the competitors' websites, they still have the same page scores and load times as before. They obviously have not read this article.

As you can see from the video comparing the three websites loaded side by side, our client's page renders 1 second faster than Blackcircles and a whole 7.3 seconds before ASDA Tyres. You may see the mobile version of the CarTyres page loading instead of the desktop verision; this is a visual issue with the tool as it is actually rendering the desktop URL NOT the mobile URL.

You can see from the data above, that by just implementing these changes to your website, it will dramatically improve your load time. This is especially important if you want to stand apart from your competitors. But what does it all mean? Yes you are satisfying the user’s intent with a faster load time but are there any other benefits?

We mentioned earlier that for every second a user has to wait for a page to load you lose 7% conversion on your site, which is a large chunk over the year. We also wanted to see if improving load time would affect our competitor's rankings.

The below results are very interesting:

As you can see from the image above there has been a huge improvement in keyword rankings across the tracked keywords. We would like to tribute all of this to the improvement of the load time and Google page score. But Google was kind enough to possibly mess with our experiment slightly by releasing an algorithm update. We feel that this update may have changed the rankings slightly but not as much as the results above.  

Conclusion

With this case study we have shown how it's possible to get ahead of your competitors in SERPs and develop onsite conversions by simply improving your website's Google PageScore and Load Time. We feel the time spent on this website improvement is well worth the benefits. Therefore we carried out this process across all of our clients and would recommend the improvements to all our potential clients. We hope you have taken some interesting points away from this article and will look at improving your own website's page speed.  If you would like any help, feel free to get in touch with us.

This helpful Load Time How To was written by Craig