A Compelling Case for For Speed

Posted on 11. Jan, 2012 by in Thoughts

Like most of my colleagues in the technical SEO team, one of the most common problems I face is making a compelling business case for fixing the technical difficulties I find on clients’ sites. The positive relationship between, for example, high quality content and a strong social buzz is easy to see and easy to grasp; but the relationship between, say, combining multiple JavaScript files into one and a positive effect on the bottom line can be hard to establish and even harder to communicate. But as the speed issue gains momentum, we need to find some way to sugar the pill.

Making page load speeds sexy

Page load speed is one such unsexy SEO issue which is starting to gain wider attention. Only last week on the bigmouth blog, Blair Walker wrote about the importance of optimising page load speeds and looked at the great Site Speed tools available in Google Analytics, while back in November 2011 my tech team colleague Gerry wrote on the same subject, partly in response to Alex Graves’ post Google Page Speed Optimisation is a Waste of Time.

Lipstick on a pig

While it’s great to see this issue getting the attention it deserves, Alex Graves’ point of view (the essence which is that “dedicating huge amounts of time to battle against the statistics […] is time that could be better spent elsewhere”) is probably common to a lot of website owners. So just how do you make a convincing case for a potentially expensive list of seemingly pedantic tweaks?

I’m alright, Jack

In my opinion, the first hurdle is to dispense with the “it looks all right to me” fallacy. It’s true that for most, if not all of the websites I regularly work on which have speed issues, clicking around a few pages on the site would seem to contradict the information coming from Google Webmaster Tools. But for most large commercial websites, it simply isn’t possible to get a meaningful sense of typical user experiences by simply following a few likely user journeys.

It’s not about rank

Every self-respecting SEO nerd now knows that Google use page speed as a ranking factor, but they also know that the “site speed signal” is at best a minor ranking factor for a tiny fraction of sites. But working towards snappy page load times isn’t principally about ranking, it’s about creating a pleasing user experience which, assuming the rest of your SEO ducks an in a row, can have far reaching results.

Show me the money

All well and good so far, but we still haven’t quite arrived at that “compelling business case” I was lamenting the absence of. Blair’s article last week referenced one of the better known pieces of research (from Akamai) on what site speed means to a site’s users, but as one client pointed out to me, impassioned pleas for faster sites from a company that earns its crust selling speed solutions for websites have to be regarded with a degree of circumspection.

Lucky for us, then, that Shopzilla’s VP of Engineering, Phil Dixon, was kind enough to share some detailed insight into his mission to get his site’s page speed under control.

Shopzilla Google Sessions Graph

Shopzilla Sessions Data

Phil presented data showing that having reduced page load times from an average of 4-6 seconds to 1.5 seconds, conversion rates increased by between 7 and 12% (depending on product), page views increased by 25%, and SEM performance for the UK rocketed by 120%. Not only that, but costs related to serving their content was also reduced as a result of improved performance.

Compelling enough?

Granted, what Shopzilla did was a big undertaking and, as Phil Dixon concedes, it can be hard to retrofit an existing site to perform like a site built from scratch with speed high on the priority list. But what Shopzilla’s work shows is that speed most definitely makes a measurable difference to all manner of important metrics, and with ever more crowded and competitive markets,  it’s fair to suppose that any effort spent on faster pages is far from wasted.

Have you got any speed related success stories? Do you plan to slap some go-faster stripes on your site? Or do you think the speed thing is an expensive SEO cul-de-sac? Let us know what you think!


Tags: , , ,

  • Tayfusion

    While I agree that load times can indeed have a negative impact on user experience and should definitely be a key performance indicator in terms of usability, it may be difficult for smaller businesses to effectively measure any potential ROI from investment in re-design or technical implementations such as local caching. An example I would give is a client of mine operating within the travel sector invested in providing more engaging content on the homepage, in an attempt to boost engagement and ultimately conversions.

    A subsequent impact of this was increased page load times due to additional content objects and imagery, however, pageviews were up 227% and bounce rate fell by 27.5%.

    The end objective of increased conversion was indeed achieved at +7% due to a more compelling user experience.

    After reading this post, it has me thinking that possibly investing some time andd effort in speed optimisation may indeed bolster the gains achieved through content regeneration. Hmm!

  • Glynn Davies

    Thanks for the comment Paul. You make a fair point, and I’d argue that, as with any element of a website, the issue of load speed shouldn’t be considered in isolation. As you say, compelling content is key and the fastest loading page ever seen won’t substitute for that.

    Thinking about it now, it’d be interesting if sites incorporated questions about load speed into customer surveys… maybe some already do, I’m not sure, but it certainly seems like a good way for a business to determine whether it’s an issue worth investing in.

    • Paul McDonald

      Nice idea, I may drop a KissInsights survey in to gather customer feedback on this.

      • Glynn Davies

        That”d be great! It’d be wonderful if you could share elements of what you find, too…