Keywords, rankings, and SEO have gone hand-in-hand since the beginning of time. (Well not quite, but it certainly feels that way.) As an SEO, this can be a blessing and a curse, depending on circumstances…
Senior management, in particular, can sometimes be guilty of judging performance by searching for a few individual key phrases — and if they’re not happy with what they see, SEOs can find themselves backed into a corner, in spite of wider success.
Reporting and KPIs should always drive action, and so when stakeholders see keyword rankings which look unfavourable (and they don’t understand the nuance of personalisation, localisation, or the role of a given keyword within a broader strategy), they are often are often quick to react, demanding performance improvements for the term in question.
It’s the nature of our industry that there’s always pressure to improve individual keyword rankings. But just how feasible is it to manipulate improvements for individual keywords?
What’s an SEO to do?
Historically, if we wanted to improve a ranking for a certain term, we would (more than likely) start by seeking new links, with manipulated anchor text, in order to boost the relevance of the page we are linking to. However, in a post-penguin world, with potential for manual penalties, this becomes a very risky strategy.
Now, many SEO strategies do not focus on improving an individual keyword — or even a particular section of a site:
- Often, some of the biggest opportunities to improve performance come from technical optimisations which aim to make overall gains and result in improvements on every potential keyword.
- Other big wins frequently come from campaigns, content marketing initiatives and impactful media coverage which, similarly, impact broad pools of relevant keywords, or a whole website.
That’s not to say that no improvements can be made to individual keywords. We can always manipulate positive change in individual terms through tweaks to titles, headings and copy, as well as improve the flow of equity to pages and improvements in internal linking patterns, for example.
However these tactics often — and quickly — reach a point of diminishing returns, or you end up chasing your tail (pun intended), particularly with larger websites. Often, the effort doesn’t justify the return, when there are better and more effective ways to spend our time.
There are other ways we could potentially improve individual keywords through a more long-term content marketing strategy. We can attract deep links to the page that our specific keywords rank for. We can ensure that the links we get are from other sites are relevant or related to that keyword. Alternatively, we could attract links to our content piece and indirectly direct the equity via internal anchor text to our target page…
Ultimately, though, there is far less guarantee that we will improve the specific keywords we have targeted.
Realistically, we may see some improvement in individual keywords with such techniques but it’s at the risk of more worthwhile pursuits. Why focus on a single keyword, at the expense of the long tail? Why try to shoehorn deep links to improve an individual page, over improving the overall authority of a site, which assuming a solid hierarchy, will potentially improve rankings of the site as a whole?
By now, my disdain for focusing on individual keywords is clear. Focusing on a single keyword, no matter how popular, resultant traffic is likely minimal in comparison to the long tail of traffic available relating to said keywords. It’s simply not possible to try and chase improvements for each individual variation. Equally, rankings vary from hour to hour, from location to location, from user to user.
Challenges with reporting
Humans are creatures of habit and habits are difficult to break down. Rankings are easy to understand, so are easy to cling on to as a key performance indicator.
What we need, is a better way of reporting success and to educate by numbers.
In the past, I have used an average rank of a selection of keywords to individual areas of a site to report on performance, rather than a single keyword. But this approach is flawed; the average rank could well be brought down by keywords that provide no real business value. This is entirely dependent on the keywords you select, of course, but there’s a tendency to include those that rank well and allow the volume of keywords included to creep up, as you discover new keywords and equally, remove those that have no volume, or simply don’t perform. When reporting and analysing performance, consistency is key.
Instead of focusing on rankings, what should we focus on?
My preference is always to look to traffic and revenue, as this is what drives business. Traffic and revenue can be reported by pages, or groups of pages, defined by themes, depending on the size of the site in question.
Quite simply, you can’t argue with extra traffic and revenue. However, like other metrics, this one has a flaw. When businesses report on traffic and revenue, they frequently do so in the context of year-on-year comparisons. And with Google increasingly ‘shaking up’ its ecosystem, things can get messy.
I have personally seen a number of sites showing YoY growth, right up until Google’s introduction of the 4th ad spot, in late February/March of this year (resulting in an effective ‑1 drop to every major commercial keyword). Despite obvious visibility growth and success, you may not see YoY growth, as the landscape shifts around you. But that’s another story, entirely.
With any reporting, it’s difficult to isolate success. Normalisation is one method that can be employed. When making your desired changes, report the change against a normalised level of traffic and revenue. This allows you to see the change with less noise associated with it.
In addition to the ever-changing search landscape, reporting at a page level may seem a little ‘black box’ for senior managers — but to me, it’s the only way to show improvement and prove real value. This is particularly prevalent when we don’t have detailed keyword performance data in our analytics packages (in an age of “not provided” keywords).
Here’s the crux of the problem…
I fear rankings will always be the thorn in the side of SEOs, as long as we obsess over them and senior management remain uneducated to the bigger picture.
If revenues have increased by 20%, what does it matter if you’ve lost one place for a single keyword?
KPIs should help you to understand performance, and to drive action — and rankings in isolation do not.
Why try to specifically improve a single keyword, when the actions you take could improve hundreds, if not thousands? If our objective is to help the businesses which we work with, and for, to be successful, then it’s up to us to educate, and not to speculate!