Ultimately there are two typical approaches to building and managing websites:
- Create something that the CEO thinks looks nice, and make sure everything works as it should
- Start from business objectives, define what success will look like online, and carefully build and optimise the extended web presence to deliver that success
I’m not handing out any prizes to readers who have guessed that I prefer the second approach. If any online development exists for a reason or with a specific objective in mind, then we should be thoughtful when it comes to how we deliver that outcome.
That second approach also implies the use of metrics. If we have no way of measuring success, how do we know that we have done a good job, or where we need to improve? Just as the higher level business objective will almost always be measured using a specific metric, even if that is just revenue, so the online work we do needs to have clear, agreed metrics and associated targets.
But it is vitally important that we choose theright metrics. A cursory glance at Google Analytics tells us there are hundreds of things we could optimise for - and that’s just the start of it. The fact is there are all sorts of numbers that a business could optimise for that would actively damage their business.
So before taking my word for any of this, make sure that the metrics you adopt make sense for your own business. Ask yourself if maximising any particular number will mean an unambiguous improvement in metrics up the chain, all the way to revenue.
The big numbers you care about: growth, revenue, profitability, are a product of smaller numbers that are easier to influence. Here are five I recommend considering if you want to make a real difference. There are others of course, but these represent a decent place to start.
1. Total users
At the end of the day, traffic matters. No high level number gives you a clearer idea about how your business and online presence is growing. By looking at users (previously ‘unique visitors’) rather than total page views or similar easily-gamed metrics, we have a more realistic view around our popularity, and one that chimes with the common understanding implicit in the question “how many people visited our website this week?”
Of course volume of visitors is not everything, but it is usually a fair start. At the same time, it is worth noting that this metric is perhaps the slowest to be influenced by changes made to the site. In the shorter term changes to advertising and press campaigns are more likely to be behind fluctuations, so do keep an eye on the long term trend: over time an SEO and UX optimised site will deliver real improvements in total traffic levels
2. First time visitors from organic search
A subset of the metric above, but an important one. This number is usually a good indicator of the overall effectiveness of your search engine optimisation (SEO) efforts, both in terms of the content you are creating and the structure and build of the site itself.
As the Google search algorithm becomes more and more sophisticated in order to satisfy the various new ways that users are searching, more meaningful, high quality content is rewarded. That in turn means that the ‘long tail’ in search becomes ever more important (and it always was, of course). In that situation, the impact of SEO is best measured not by the rankings for specific keywords, or the traffic arriving at specific pages, but by the total traffic delivered from searches of all descriptions.
Of course search rankings still matter, but they are harder to influence artificially, and are becoming a more tactical instrument. The core question should be “is the totality of what I deliver online bringing results in terms of organic search?” This is the metric that answers that question.
3. Engagement
This isn’t a metric you’ll see on Google Analytics, and I use the more general term ‘engagement’ for a reason. If it helps, think of it as any combination or product of total time spent on site or page views per visit, with a little bounce rate thrown in.
That in turn explains why I use the catch all term: it covers a range of metrics each of which can be used as appropriate to measure how deeply the traffic we delivered above is engaging with our content. That is usually the next step toward conversion, and in some cases such as media sites it is pretty much all that matters.
Do note that engagement is a metric that can be viewed from either end of the telescope. In some cases lower engagement is a good thing, particularly in highly transactional cases such as purchasing of travel tickets. Take the time to define a metric that works for your business and stick to it.
4. Conversion rate
All the traffic and engagement in the world won’t pay the bills. In most cases websites exist to deliver either concrete inbound enquiries or direct sales. This metric establishes exactly how well your sites and landing pages turn visitors into those leads or revenue.
Conversion can and should be measured on individual pages and across the site as a whole. In the former case, it is vital to have a clear understanding how any one individual page performs if - for example - the desire is to have the visitor complete a form that registers interest. Without this number, any work to optimise and test such pages is done ‘in the dark’ and thus will be limited in effectiveness.
But it is also important to remember the site as a whole. This metric allows us to factor in how effectively the entire experience, from the home page down, drives people towards purchases and form completions - and helps users get to the point of conversion. We should always been looking to understand what % of total visitors complete this journey - and optimise that number to the greatest extent possible.
5. Cost per conversion
If on the one hand we wish to understand and optimise how effectively the site generates conversions, on the other we want to ensure that our business is doing so in a cost effective way.
After all, it isn’t much use to create customers with a lifetime value of €200 at a cost of €250.
For this metric to make sense, it needs to be generated and analysed at an individual source, or even campaign level. At the highest level, this enables us to see if our marketing campaigns and spend on public relations (for example) is actually effective in terms of return on investment.
But we also gain insight into the nature of the audience being delivered by these campaigns, and some guidance as to whether we are sending these audiences to the right places. If the cost per conversion numbers look challenging, we can optimise the nature of an AdWords campaign for example, or reconsider design decisions on the site and landing pages that might help that particular audience get to the point of conversion.
A final word on ‘reporting’
All metrics become more powerful and more useful to the business when they are imported on a consistent basis into a report that tracks performance over time.
This process is the last vital step towards really understanding the performance of an online business and a website specifically. It helps identify trends over time establish where challenges lie within the funnel, and sport threats and opportunities whilst there is still time to take action.
So don’t simply use these metrics on an ad hoc basis - commit to building out reports that support proper analysis over time and allow targets to be generated for the future.
