The introduction of Google Sitelinks back in 2007 came as part of major changes in the evolution of Google’s search result pages, and fundamentally changed the way searchers discover content. But how well do they really perform for us?
Since Google’s Sitelinks first started appearing in the wild, their prevalence has spread throughout many search engine result pages, for both branded and generic queries (books, finance and amazon for example). For the uninitiated, Sitelinks are additional deep links on search engine result pages that enable searchers to drill down to specific website content from the word go.
As webmasters and SEOs, we have little control over which Sitelinks are displayed in our search results. Google’s algorithm takes responsibility for that, figuring out which pages people are likely to want to jump to. There are numerous blog posts surrounding the method by which Sitelinks are served and one that is definitely worth researching is Google Sitelinks: The Ultimate FAQ.
Even so, a modicum of control is given to webmasters in that we can block Sitelinks that we deem inappropriate/ineffective. Sitelinks may be blocked indefinitely, provided you revisit your Webmaster Tools account every three months. You can also unblock Sitelinks, though as with blocking, there will be a lag.
But how do we know what’s bad? Should we guess? Or maybe we should look to the data.
MONEY-SAVING PRICE QUOTES FROM?
How much traffic is generated by Sitelinks?
Step 1: Identify which terms result in your website displaying Sitelinks.
I’m assuming you know your site already has Sitelinks. This can be checked in Google’s Webmaster Central. If you don’t, go build some quality links and sort out your information architecture!
Once you’ve confirmed this, you should then figure out which key phrases are causing your Sitelinks to appear. This shouldn’t be too difficult. Likely keywords to check are your brand name (e.g. Moodia), URL (www.Moodia.com), and – if you’re lucky (or clever) enough to be a category killer – generic phrases (widgets). Make a list of them because we’ll need them in step 2.
Step 2: Create a Custom Segment in Google Analytics
If you don’t know how to use Advanced Segmentation in Google Analytics (hereafter referred to as GA), then this official Google video is a great primer. Beyond that, Avinash Kaushik has a superb write-up of the tool.
To sum that up:
- We need GA to only consider queries that are generating Sitelinks.
- We make sure these queries aren’t resulting in visitors landing on the homepage.
- We then limit traffic to organic visits from Google.
- Save your Segment and apply it to your report, along with some other segments for comparison (try All Traffic, all Organic Google traffic, and also all traffic from your Sitelinks keyphrases as above.)
Step 3: Analyse the data and draw conclusions
You’re now presented with an idea of the traffic being generated from Sitelinks. You can compare it against other segments and see how it performs. It’s useful to consider metrics that aren’t all about volume and we advise you to look at both ratios and averages. This makes Pages/visit, Bounce Rate and Avg. Time on Site very handy.
When gathering results it’s important to remember that you’re likely to get some data pollution. Websites with Sitelinks will often have secondary and tertiary pages ranking normally somewhere on the first five result pages. These will invariably bring additional traffic into your segment. However, many other sites will have specific goals configured, and so metrics like conversion rate, average order value and per visit value may be taken into account.
It’s down to you to judge the extent of data pollution present in your tracking. For example, if you have another page ranking in second position below the Sitelinks, the inaccuracy may be insurmountable.
Getting actionable data
Considering the only direct action you can take with Sitelinks is deletion, you need to find which specific links are bringing your site down.
Rather than simply looking at traffic landing from Sitelinks keywords on pages other than your homepage, you can break it down further and consider specific landing pages. This will allow a detailed analysis of which Sitelinks are bringing about engaged users and conversions, and which aren’t. Simply create a new Segment, but this time instead of excluding your homepage as a landing page, you must set the landing page to match exactly the Sitelink URL that you’re looking at.
Example: Quick, hit delete!
For this quick example, we’ll imagine a generic info page Sitelink that is underperforming in terms of conversions and engagement, when compared with total site traffic as well with one of the other Sitelinks. Making things even worse, the content of the generic page is reflected by the Sitelinks themselves in Google. The value added seems low. It’s time to head for Webmaster Tools and give this Sitelink the chop. After all, the other Sitelinks perform better.
So, now’s the time to hit delete, right?
Hang on, hold fire. Is that really the right approach? If you found out you had a gangrenous arm, would you hack it off with the closest available utensil, or book a trip to the doctors? It might be wiser to employ A/B or multivariate testing on the page in question to help it fare better. Run usability sessions, acquire heatmaps, and gather comments online straight from your customers.
These are all valid courses of action. But maybe, just maybe, this was never meant to be. Did fate really have this URL marked down as a landing page? The thing is, while a page may be criticised for being a poor performing landing page, it may not necessarily be there to serve this purpose.
Look at your analytics and find out how this page is performing overall, as part of your conversion and persuasion funnel.
- Is it a big exit page as well as a high bouncer?
- Do all search keywords result in a high bounce rate for this page, or just the Sitelinks?
- Is this page a winner when reached internally, but a loser when landed upon?
After all, with Sitelinks, Google is merging importance with landing page suitability. Consider whether it makse sense for visitors to be plonked down somewhere in the middle of your sales funnel, or would you rather have them enter at the top? You are the person who must hold the casting vote on that issue. And by you, I mean your visitors.
Here’s a quick recap of our arguments:
Arguments against deleting Sitelinks
- It will decrease the number of opportunities for visitors to reach your website, see your brand, and be exposed to further persuasion
- Deleting enough Sitelinks will result in your competitors inching their way up the search results page. Do you want to give your rivals more visibility?
- Ineffective landing pages should be improved by testing; they shouldn’t have their lids shut to new visits!
- Engagement metrics might appear unfavourable because Sitelinks are performing as designed – to bring users to their destination page with a single click
Arguments for deleting Sitelinks
- Increases the likelihood of a visitor landing on a preferred landing page
- Increased engagement
- Improved e Commerce performance
- Reduces clutter, helps convey key brand messaging and product/service offering
Other points to consider
- Data pollution – are your segments being contaminated by secondary/tertiary search results?
- Nature of the page – is it a one-stop-shop for quickly answering common questions? If so, engagement and conversion metrics should be used with caution
The way in which these points are weighted is up to you, and will ultimately decide on whether you delete Sitelinks, focus on testing landing pages, or simply do nothing.
Hint: you should never be doing the latter.