Meaningful SEO Metrics: Going Beyond the Numbers – SES London 2011 (#SESUK)
Next up are Pete Young from Brilliant Media (and State of Search blogger!) and Matthew Bailey from Site Logic Marketing talking on how to look beyond the basic/obvious metrics such as PageRank, indexed pages etc. and further into metrics that can make big changes to your company and your ROI. Moderating the panel is Brian Clifton, author of Advanced Web Metrics with Google Analytics.
So what does meaningful metrics mean? Brian Clifton gives a little background – firstly we have to consider two different types. Firstly there’s the nicely controlled on-site metric – what visitors are doing on your site, what they buy, where they come from, how long they stay. The majority of these are all within our control. Off-site metrics are quite different. People talking about your brand, tweets, mentions – how do we compare these metrics alongside one another and gain meaningful insight from them?
First up is Pete discussing these off-site metrics. As SEOs we’re obsessed with metrics, from number of links to pagerank and effectively what rankings we get. However a client is focused on ROI and customers. Already there is the hint of how clients and SEOs see things differently.
Traditionally SEO metrics are not aligned with commercial KPIs such as traffic, leads and sales. Tehcnology hasn’t been available to provide a 360 degree perspective of the integrated picture. And SEO often only has limited understanding from a client side so there is a division between the two sides.
However more and more people are investing in SEO, it has grown hugely over the last few years. More people are buying online, or researching online then buying in-store. Or even looking instore then returning online to buy. The lines are being blurred.
Therefore the lines are being blurred, there is increased accountability. Pete focuses on the four Rs:
What sales/leads were generated?
How much traffic has the campaign produced? Was it incremental or seasonal?
How many more searchers have been exposed to my product?
Why? Why are those searchers now finding my product?
Incremental increases are what counts. Fundamental for Pete is ensuring you start with a baseline for your existing activity – especially important.
This then provides a baseline for forecasting and ROI analysis. Click-through-rate data is key to forecasting, so you must understand this to be able to use it well. However there are huge caveats involved. It’s not 100% accurate and considerations such as search queries, searcher intent, sitelinks and blended search results have a significant impact on where people click.
Forecasting directly impacts on potential campaign profitability – looking at position, click-through-rate, seasonlity, average order values etc.
There is a halo effect from this. Metrics such as share-of-voice provide a wider picture beyond the initial campaign focus. Tools such as Raven or your own bespoke system can give you a broader perspective on share-of-voice and share-of-traffic.
Integration and de-cuplication can be crucial and is a big growth area for larger companies. Integrated tracking to allow you de-duplicate figures between PPC and SEO campaign will open a number of new opportunities including:
- Actual channel lead blends
- Blended search ROI
- Impact of one channel on another
- Understanding of client online marketing behaviour
Offline tracking is another growth area. New tools on the market such as AdInsight and Infinity allow you to tracking call centre activity on a granular level, tying it back to search engine and keyword level metrics.
Econometrics can bridge the gap between web-based tracking and measurement of off-line sales. This uses mathematical models to measure the number of in-store sales driven by online leads.
- Report metrics that are relevant to your client not just convenient for you
- Simple forecasting at the start of a campaign can act as a benchmark for potential campaign profitiabiilty. Increamental impact is key.
- Try to de-duplicate your conversions. This gives you much more accurate and realistic data.
- Don’t look at hero rankings in isolation. Remember the bigger picture.
Now speaking is Matt Bailey, speaking on why he thinks data needs to be made more complicated in order to make it easier to understand. If you can communicate properly to a CEO/business owner exactly what those metrics mean, you can get more investment or interest.
Edward Tufte quote: “To make charts simpler, add more information”
Matt’s 6 keys to integrating:
- Show multi-variate data
Traditionally data points are single, the more we add, the more we understand what is happening with each person coming to a website.
- Show comparisons
This allows us to compare different groups of people which allows us to learn what is and isn’t working.
- Show causality
Ask WHY? What caused this to happen? Why did this visitor buy and this visitor not?
- Show different reports for different sorts
CEOs and product managers and IT staff are all interested in completely different things. The CEO only wants to know one number, the product manager is interested in in-depth analysis.
- Show the money
Where is the money going? Where is it coming from?
It’s all about multi-variate data. Why? Firstly, more data points mean better comparison. Love your pivot tables. This gives you a number of data points within one chart.
Secondly, better comparisons means better causality analysis. You can’t look at a table of numbers and remember what the page looks like. You can segment by traffic acquisition for example – where did your visitors come from? Determine why your visitors came to your site from that source. Context is huge – consider the difference in a person coming from a contextual link within a piece of editorial versus a visitor from a twitter link. Those from twitter are less likely to stay as they are not already drawn into that particular content area, unlike the editorial reader.
A high degree of content and low competition – people will stay longer and convert at a higher rate. A low level of context and high competition will generate transitory visitors, unlikely to stay long or convert.
Learn to show causality. Metrics mean nothing if you don’t look into why they areas they are. Always look at the site at the same time you are looking at the data!
Thirdly, better causality will lead to better decisions. Consider what is happening with to your visitors, not just Googlebot. For example, implement 404 page monitoring in Google Analytics so you can see what 404 pages people are actually landing on.
Finally – follow the money! Report based on the revenue you are making. If you can show more money as a direct result of your efforts, you are winning. Consider per visit goal values – give them an actual value. Think of it as actual goals rather than simply shots on goal.