Improving Site Speed – Talk About The Business Benefit

For many months now, webmasters have been responding to signals from Google that page speed would soon become a ranking factor. Very recently, Google confirmed that for a limited set of queries, Google.com (US) results are influenced by page load times (around 1% of queries). Although you shouldn’t panic about this announcement, it may be wise to be prepared for a wider implementation of the change across Google’s search results.


Photo credit: Ed Callow

Thing is, if you haven’t been optimising your site performance for improved speed until now, you may have missed a trick (or a conversion, email registration, repeat visit or some other valuable metric).
The business impact of performance changes to a site

At Velocity 09, Eric Schurman and Jake Brutlag presented “Performance Related Changes and their User Impact”. Their presentation was based on groups of site performance experiments carried out at Google and Microsoft. Microsoft introduced delays on their servers and slowed down the delivery of their search results, slowing page load or testing progressive rendering on their pages. Their objective was to collect data to help the organisations decide how important page load performance actually was to a business driven outcome.



If you don’t have time to review this (excellent) video, let me summarise a few important points:

Microsoft concluded that server side delays above around 200ms introduced strong negative impacts on the following metrics:
Repeat usage
Revenue (A revenue drop per user as much as 4%)
User engagement
Customer satisfaction
Time to first click (as much as twice as long compared to the delay introduced)

It was extremely fascinating to learn that one of Google’s methodologies to introduce a lag in their search results was as simple as setting a delay to respond to the very first server header request received from the user. Google were also able to respond with a server header response before they’d actually generated the search results themselves and test variations of ads rendered before organic results and vice versa.

Google similarly concluded that a linear relationship exists between increasing server side delays and user satisfaction metrics, abandonment and declining repeat searches.
What things can we do to improve our page load times?

If you’re lucky enough to have the development skills nessecary to implement performance changes on your site, then you might want to review Google’s recommendations for site performance tools or read this post on improving your site speed written last year on SEOgadget.
I have no direct control over my site performance

You might have no direct control over your website’s page load times. Most SEO’s don’t directly influence the site technology platform, hosting solution, content delivery network choice or proxy caching solution for their (large scale) site infrastructures, but that’s not to say they can’t offer advice, share research or evangelise in their organisation.
14% will start shopping at a different site if page loads are slow, 23% will stop shopping or even walk away from their computer.
Spread the word

Make your technology team aware of the Google announcements and draw their attention to the tools made available in the labs section of Google Webmaster tools. Fact is, most tech teams have superior site monitoring capabilities to Google’s labs based performance tools but introducing the idea that site performance can impact SEO and user engagement is a great way to capture interest. A presentation, an email or a face to face conversation can really help. In my last in-house SEO role, I would get together with the technology team to discuss SEO on a monthly basis and they were always really interested in what developments were happening in the SEO community. New issues in the SEO community, in particular, got them interested!

Distribute research available that suggests the relationship between site performance and user engagement metrics. The video embedded in this post can be found here and the presentation on Scribd, here. Data is very difficult to come by (and I suspect the outcome of this type of work will depend very much on the market, keyword, page layout / design) but experiments on page weight were being carried out a long time before Google began their performance campaign – so it must be pretty important, right? Facebook published a summary of the impact of page load on user engagement on August 28th 2009, concluding that metrics such as page views per session improved as server response time and page load times were reduced.

This excellent study on Royal Pingdom (found via SEOmoz) concluded that user expectations on site performance have changed significantly since 2006, stating:
47% expect a web page to load in two seconds or less.
40% will abandon a web page if it takes more than three seconds to load.
52% of online shoppers claim that quick page loads are important for their loyalty to a site.
14% will start shopping at a different site if page loads are slow, 23% will stop shopping or even walk away from their computer.
64% of shoppers who are dissatisfied with their site visit will go somewhere else to shop next time.

That’s amazing! Improving or resolving these factors could yield how much % increase in ROI for your business?

You could also suggest alternatives to big hosting upgrades, such as the implementation of a remote proxy cache that serves content in the geolocation you’re actually targeting.

Google’s announcement really brings a much older conversation back into play – what are the business benefits of implementing a faster, more efficient website? Suddenly a mysterious and highly difficult to predict potential ROI from 1% of US searches becomes measurable and highly actionable, with potential for a big impact on ROI.