How to quickly check if your SEO work was effective

by Tim Hill (aka Doodled)

August 16th 2015
Your website is an oil tanker in an ocean of rankings. Selecting reverse gear right now will change nothing for quite some time
Your website is an oil tanker in an ocean of rankings. Selecting reverse gear right now will change nothing for quite some time.

So you hired a company to improve your rankings or you beavered away learning about on-Page SEO and link building and put your knowledge to work.

The changes have been made but your rankings haven't changed yet so how do you know if the right edits have been carried out? How do you know if you are on the right path?

N=50

Before you go ploughing through Google Analytics it is important to understand a basic statistical rule. No data is believable unless it has 50 points.

So, for example, if a page on your website (or your website as a whole) had 20 visitors yesterday and a bounce rate of 70%, 15 visitors today and a bounce rate of 45% this does not mean your bounce rate has improved. 20 < 50 and 15 < 50. Not enough data to draw conclusions.

In an example like the above you would need to compare two weeks, rather than two days. For small businesses I generally wait until a month after the changes have been made but larger sites with an already established level of traffic can be done sooner - it all depends on data volume.

Where to go and what to look at

Its Google Webmaster Tools that most people get glued on. Are their rankings going up? Forget this, for reasons that I will explain later, and head over to Google Analytics.

Once you log in you'll be shown the overview screen with a line graph about visitor numbers on a daily basis. In the top right of the screen choose a date range which includes the month before and the month after you made your changes. If you have a high level of visitors already do it by weeks.

To the top right of the line graph choose 'months' or 'weeks'. This gets rid of the daily noise making it easier to see what is actually happening over time.

Now above the graph choose 'Add Segment' and select 'Organic Search'. This is what your SEO work is supposed to affect so we want to strip out data relating to referrals, people who come to your site directly or those that visit because of PPC.

Depending on who you are aiming at you might also want to add some additional filters. Common ones would be:

  • Only visitors from the country you are aiming at.
  • Only visitors using a particular device (e.g. tablet users).

Check again, does N=50?

By the time you have done all your filtering do you still have at least 50 pieces of data (50 visitors) on both sides of the date where you implemented your SEO work? If not, wait longer and then expand the time scales - e.g. two months on both sides of the date the SEO work was implemented.

Its all about user experience

Google wants to give users the best experience which means the obvious, offering up the right pages in searches. How does it know the pages are right? It has many signals at its disposal:

  • CTR - how often users click through from its search page to your site
  • Bounce rate - how often users who visit your site hit the back button
  • Dwell time - how long users stay on your site
  • Links to your site from other websites

I'm going to leave the Links to one side. You can have all the back links in the world but if CTR, Bounce and Dwell Time are awful for a particular keyword/keyword phrase Google isn't going to rank you.

CTR can be found in Webmaster Tools but it is too early to look at this, I'll explain why in a moment. Bounce rate and Dwell Time are dutifully reported in Google Analytics.

So here is the question: Have you made users coming from Google Search happier? Do they bounce less and stay longer? Do they visit more pages on your site or less?

Oh No, its all got worse! Should I undo my SEO changes?

Don't Panic and don't be tempted to reverse everything you have done just yet. You may end up throwing the baby out with the bathwater.

Start looking at who doesn't like your changes. Mobile Device Users? Internet Explorer Users? Visitors to a particular page? And so on.

In other words can you refine your changes rather than reverse them.

As an example I recently carried out major site changes for a client and their bounce rate increased from 45% to 65%. On investigation it was iPad users whose bounce rate had increased to nearly 80% and they represented over half the websites traffic.

The changes we had made had not gone down well with iPad users at all while the bounce rate for desk top users had dropped to 35%. We had made one group happier, and another group unhappier.

The root cause was offering tablet users the mobile version of the site. Once we changed this so all Tablet users saw the desktop version their bounce rate immediately reduced to 35% as well and so overall we had taken the website from having a 45% bounce rate to a 35% bounce rate.

Everything got better but my rankings went no where or got worse

This is where many people make their greatest mistake, they rush to undo changes because they don't seem to affect rankings or they seem to make rankings worse.

Painful though it is you have to wait it out, not bottle out. If analytics shows improvements then Google knows about them but that doesn't mean it is ready to act. Google also knows about N=50 and it will want a solid track record before it goes changing its search results for any particular keyword phrase.

Depending on your volume of visitors this may take months. For small to medium sized websites I generally give it six months. If you don't want to wait that long use PPC to accelerate the amount of data that you give to Google.

You should also take into account three factors:

  • declining and increasing rankings have momentums
  • Google's tests and the rankings mess
  • its all relative

Declining and increasing rankings have momentums

For example, if you were losing your rankings before you made changes this decline will be slow to turn around. Like an oil tanker making a U-turn its not going to happen straight away.

This is where most webmasters make mistakes. Either they make bad changes to their site, see their rankings rise and are happy but later cannot understand why their rankings drop or they make good changes to their site, see their rankings fall and undo the changes before they can have their real effect.

It is also why some Webmasters are positive about the work of bad SEO companies and negative about the work of good ones, unless the content of this post has been explained (and understood) by them.

If the data in Google Analytics is positive then stay the course!

Google tests and the rankings mess

After you've made your SEO changes Google will crawl your site. You can see how often it is doing this in Google Webmaster Tools. If it isn't upload a sitemap or request a crawl.

So it knows you have made changes but it doesn't know if they are for the better. The only people who can answer this are users and it has to try you out. Not just for the keyword you are aiming for (Google doesn't know what this is) but for all keywords and keyword phrases it thinks the site and pages are related to.

It will try ranking you for [this] or for [that] every now and again and it will make mistakes. This is why your overall CTR figure in Google Webmaster Tools might actually get worse. Ignore the headline figure and only look at the phrases you are targeting.

Google will sort out its mistakes over time.

Once again, when you are looking at CTR for a particular keyword phrase remember N=50. You will need to have had at least 50 click throughs to be confident the CTR change you are seeing is real.

Its all relative

You may have decreased your bounce rate, increased your CTR and Dwell Time and let the months pass but still no change. Remember these measures (along with some 500 others which Google considers) are all relative.

If your competitors have been doing the same, or if your improvements still haven't reached those of your competitors, you're not going to see any changes in your rankings.

Your competitors may have been working on some of Google's other signals such as link building or that may still, unfortunately, be better liked by users than you.

This is why I can be pretty pleased when, for some clients, I decrease their bounce rate from 75% to 65% because in that particular industry 80% is the norm.

If your rankings drop look for new competitors who have entered the market or older competitors who were ranking lower than you but have now moved upwards - have they been making changes?

When I turn round to my long time clients in highly competitive markets and announce - Good news, you are still in the top ten - they know it is good news because they understand the work I have done has held the boat steady and without it their rankings would be lower.

Unfortunately you can't know your competitors stats in the areas of CTR, Bounce and Dwell Time so simply focus on improving yours to the absolute best of your ability.

What about links?

Now all I've talked about is improving user experience when most SEO companies bang on about links, links and more links.

Well if visitors click through more often and don't leave so often but stay around what do you think some of them will do? That's right! They're going to link to you.

Make content that users love and they will take care of your link building work for you.

Summary

So I've outlined the basic checks you can carry out at a fairly early stage to see if there is a good chance your rankings will improve.

Remember your website is an oil tanker in an ocean of rankings. Selecting reverse gear right now will change nothing for quite some time.