It never fails to astound me how few companies actually demand any sort of KPI or return from their website investment. If the same client came to me with £20k for a marketing campaign their primary focus would be the likely ROI they are going to see for their money – understandably so. But so many companies will spend the same amount on a website simply ‘because we need a website’ without any real expectation as to its role in sales and/or marketing.
So here are a few key ideas as to how to monitor website performance.
Make sure you have the tools to evaluate
To support you in evaluating your website, install Google Analytics and set up Google Search Console. As well as providing great FOC insight, it will help you monitor performance in terms of traffic, rankings and page quality to name just a few. Yes, you can spend money on better analysis tools but for most the Google suite will do all you require.
What should your site be doing?
All too often we forget about the primary objective for our site. It may be sales or leads, either way before you dig into content or use any third-party analysis, sit back and consider whether your business is well supported by your website. All the data in the world cannot replace good old gut feel. Are there enough sales coming through, are there particular products that sell more/less than others, are you getting enough leads/downloads? If you have this clear in your mind it will help to direct your investigations.
Don’t just look at traffic levels
With the best will in the world and using as many filters as possible, you simply cannot ignore/remove bot traffic or spam from your reporting entirely. So, whilst the number of users, visits and breakdown of new Vs old should form part of your analysis, it should be placed alongside search rankings, traffic sources (compared to effort applied), page quality, broken links and search terms – the core ingredients for a successful site.
Do Adwords offer value for money?
Quickest way to check this one is to turn it off! Yes, that’s right, like a light switch, turn it off and see if it goes dark. Don’t simply look at traffic, which will of course drop off – but consider other metrics. Do sales immediately stop (if ecommerce)?, does the phone stop ringing?, do literature/downloads stop?. The nice thing about AdWords is you can switch them on/off or pause individual campaigns and so if the world does end, you can very quickly reinstate the campaigns and know that your job is about refinement and targeting to achieve a better ROI. If traffic drops off but everything else remains broadly the same then the answer is obvious – no your AdWords do not offer you value for money. Sadly, for an increasing number of clients, the cost of the AdWords drug is greater than the benefit, but like a drug, they become hooked on it for fear of cutting it off.
When is a high bounce rate a good thing?
Whilst I am a fan of Google web stats you need to add context when reviewing data. The bounce rate is literally that, visitors bouncing off that page/site. Normally, a low bounce rate is desirable as it means your site is engaging and keeps users on it. But you have to consider the context of content! As an example, we once had a destination management organisation as a client. They maintained a relatively good (low) bounce rate for the majority of their site for most of the year. But every November their bounce rate would shoot up, all spurred by one single page – the Xmas lights switch on event info. This page would have a near 100% bounce rate but this never bothered me, even though it would attract thousands of visitors. The reason – it was an event listing that included the day, time and details of the lights switch on. Visitors would use a direct link that was published openly on a number of other sites and in the press, they would arrive on the page, take down the details they needed and then jump off the site. So that page did everything it was supposed to do and therefore a 100% bounce rate was validation of our content – not the opposite. So don’t consider the site wide bounce rate, look at each page and consider what your next action would be if you arrived on the page.
How do I improve my website content?
We’re used to the concept of marginal gains, British cycling did well from it as a strategy, and websites are no different. Using search console and analytics data, look at each page in turn and in particular focus on those with high exit rates and bounce rates. As noted above, you need to add context here, but using this data look at the ‘worst’ pages and consider why those pages may be a bit of a turn off to visitors and make changes accordingly. It may be design, content or navigation but make minor (or major) adjustments to improve the visitor experience. Do this regularly, picking a few pages each month and constantly chase the highest/worst pages. All being well, after a few months the percentages that identify these pages will be lower and you’ll be making fewer refinements.
And finally…
Don’t stop. Dipping into analytics and your website once in a blue moon, shaking it up and then forgetting it about it for another few months is not going to improve your site. Stay on top of the data, check in every month and make regular changes or pose a different question each time. You’ll pick out embryonic issues faster and be able to nip them in the bud or equally capitalise on quality content or a bump up the rankings for a specific search term.
Ben Cooper, August 2021
