How do I know If my SEO is improving or getting worse?
To know for sure if your SEO is improving or degrading you need to select the right metrics to track and understand how SEO affects them.
First we'll look at some things that aren't good indicators of your SEO success.
A number one ranking isn't definitive
If you sell jewellery who doesn't want to rank #1 in Google for jewellery! The problem lies in defining exactly what a #1 ranking is and how people use search engines.
Search engines vary search results based on many different factors including those specific to who is doing the search, including where they are geographically and their search history. So even if I'm sitting next to you and do the same search as you for 'jewellery' we may not get the same set of results.
Positions are also far from static. Depending on the server that services your search you may get different results as new algorithms are tested and indexes updated.
So #1 in Google varies on all kinds of factors, making it impossible to say that everyone who searches will see you at #1.
Manipulating keyword rankings
Rankings are also open to abuse by shady SEO techniques. Queries that consist of three or more words (long-tail queries) can be manipulated by a shady SEO person focussing 100% on ranking for an obscure phrase.
For example, 'lincolnshire jewellery forged steel' may look like a worthwhile phrase to rank for, but no-one is searching for it. So an SEO that offers to rank you for it is offering a chance for you to appear for something that no-one searches for. They can easily guarantee a #1 ranking by placing this keyword everywhere in their SEO efforts.
This may look good on the surface but will not bring you visitors and sales. This kind of SEO is also very temporary, as search engines look broader than just a single obscure keyword.
How people search
There are many different words that can be used to express a search for the same thing. The last time Google released data on this, they showed that 80% of searches they see are completely unique- the search hadn't been performed by anyone previously.
If you look at the keywords that a site with successful SEO is ranking for, you'll see a broad range of searches all expressing roughly the same thing. So focussing on a single keyword phrase is ignoring how people search by disregarding the many variations on the phrase.
Alexa rank, PageRank and others
PageRank is still a part of how Google decides to list sites (its algorithm). However they've stopped making the PageRank number public. Even before then, the PageRank number suffered from a big data lag (waiting for the public number to update) making it very hard to use it as anything more than a vanity metric over the longer term.
All of this, coupled with the fact that it is only one piece of the Google algorithm, means PageRank is dead for serious SEOs.
Other services like Alexa provide numbers based on data they've gathered, or publically available data. These can be useful to benchmark yourself against competitors or to research what they're doing well at.
However, as an accurate measurement of your own SEO progress they are flawed. First, they tend to roughly measure traffic rather than things directly influenced by your SEO. Also due to their data collection methods there are problems with taking the numbers as absolutes.
A good SEO checker helps you to fix SEO problems on your site and to tune your SEO. But they are all opinionated: checking different things in different ways. It isn't possible to achieve a perfect score on all of the SEO checkers there are.
An improving score in your chosen SEO checker can indicate that your SEO may improve, but it's up to the search engines to actually react to your changes. An SEO checker is a diagnostic tool rather than a way of analysing how your SEO is performing.
And there's one way of measuring that can be either really useful for measuring SEO or misused...
Total visits and other big pots
Traffic comes to your shop from all kinds of sources. To look at total visits, total conversions or other big pots of unsegmented data is rolling the things that your SEO can influence in with other actions like email campaigns, social media and changes you're making to your site (affecting conversions for instance).
So it's important to work out what the SEO funnel actually looks like and how to measure it.
Know your SEO funnel
SEO is about taking actions that you hope will make your shop appear more prominently in search engines and making sure that the traffic coming from search engines is relevant.
This means that to measure your SEO success we want to measure:
- how much of our site search engines have added to their index
- how often we're appearing in search results
- how often users are clicking on our site in the results
- when those users visit what actions they're performing
Improving SEO will show an increase in these numbers and/or an improvement in their quality (like appearing for less irrelevant and more relevant keywords).
When to measure
Before getting stuck into measuring each of these things it's important to note that the time taken from you making an SEO change (say, editing your page titles) to that change being picked up by search engines and reflected in their results can vary from minutes (often when you publish something to your blog) to weeks or even never! (perhaps when the change isn't viewed as significant enough by search engines)
Search engine traffic can also be volatile. Google especially will test out how you perform for certain searches on a small sample of users before dropping you or boosting you up based on clickthrough and bounce rate.
But don't despair: we've found a decent rule-of-thumb to measure your SEO. After making a bunch of SEO changes leave them in place for at least one month. During the month try to avoid tweaking the same things you've just worked on.
When the month is complete, compare the whole month's data with the previous month. This allows you to compare a before and after with usually enough data gathered. SEO tends to favour a gradual, steady, long-term improvement rather than instant swings.
Which brings us to how to measure...
These tools all have different ways of presenting the information, but you'll be able to get everything from the big one (Google) and most from the others.
You want search engines to have visited and indexed all of the pages on your shop. The number of indexed pages shown should be flat or increasing and closely correspond to the number of pages on your site.
Of course, this depends heavily on how your site is evolving too. If you're shrinking your product lines for example, these product pages may no longer exist and be removed from the index, causing the number of indexed pages to fall.
The usual, healthy SEO profile for 'indexed pages' is a steadily growing number without big drops.
Search engine impressions
When a search engine decides to show you in search results this is a 'search engine impression'. Note that you won't see these in your analytics data because unless your result is pressed they never visit your shop.
Impressions are key because it means that the search engine has viewed you as relevant for the search term and decided to show you.
When you appear in search results and a user presses your result this is a clickthrough. When they clickthrough you'll obviously see this as a visit in your analytics. But we need to know impressions too to see the clickthrough rate: one of the most crucial for healthy, growing SEO.
Improving SEO will show growing 'impressions' over time and/or 'clickthroughs'. If your impressions are falling but the number of clickthroughs is growing this is not generally a need to worry. Search engines adapt to how well you do at keeping users on your site once they click a result (the 'bounce rate'). They'll stop showing you for searches with a high bounce rate or a poor clickthrough rate, and show you more for searches that have a low bounce rate and good clickthrough rate.
However if both your impressions and clickthroughs are falling you should analyse why. Often a poor clickthrough rate is the heart of the problem. Look at the search terms you're appearing for, how your shop appears in search results for them and the user's experience when they visit your shop.
If a user searches, sees your result, presses it to visit your shop but doesn't stay long and goes back to the search results they're a bounce. Search engines view this as a failure to direct the user to the best site for their search, so it is crucial to keep bounce rate for relevant searches low.
Improving SEO will show a falling bounce rate in your analytics for visitors from search engines.
Sometimes your shop will be irrelevant to a user's search, so this traffic is no use to you. Be sure to look at bounce rate in your analytics package by what they searched for.
A fly in the ointment here is that many search engines now do not provide the search term to your analytics for you to be able to analyse this easily. You can instead use the landing page the user arrived on to make a rough guess about the keywords they used. At the very least you can optimise the landing page with a poor search engine bounce rate for better conversions.
It goes without saying that you want to convert any visitor to your site into performing an action that's valuable to you. Something like buying, subscribing to your newsletter, sharing on social media, viewing five or more pages etc.
Good SEO will bring more good visitors to your shop: those who convert into performing an action vs. those who don't do anything.
Part of this is making sure that your shop appears for relevant keywords (SEO) but a lot is optimising the conversion rate of your shop itself (CRO). There's overlap here.
Any questions about how to measure your SEO progress? Send us a mail to firstname.lastname@example.org: we're happy to help.