The Optimum Way To Survive Not Provided in Google Analytics

Today we’re going to talk about how to survive Not Provided. So back in October 2011, Google announced that it was going to encrypt the searches of its users when they were logged into Google products. This meant the keyword-level data was not going to be passed back to Google Analytics.

Fast-forward to today, and we’re finding that up to 70% to 90% of keyword-level data has been lost, although that is dependent on the particular industry that you’re in. For instance, in healthcare, we’ve found only around 30% of keyword-level data has been lost and [shown as] not provided. That said, we’re very soon going to be faced with a 100% encrypted search situation, so we are going to lose keywords in their entirety.

So today we’re going to look at two things: One, the data that we have and how to use it, and secondly, talking about an approach for the future.

So when we look at what data we have currently at the moment, we still have some keyword data left. It would be well worth going into Google Analytics and exporting as much of that data as you’ve got; as much of the historical data that you’ve got.

Secondly, Google Webmaster Tools contains search queries; data which is still invaluable to use. We don’t know how long that’s going to remain in Webmaster Tools. So again, it’s worth getting in there now and exporting what data you can.

Thirdly, if you advertise with Google AdWords, that data is still going to remain. So that presents an opportunity to still be able to look to see what demand certain keywords, phrases, or things still have on the marketplace and also to see how your site actually performs in that context as well.

Within Google Analytics, what we need to do now is, as well as look at what keyword analysis traditionally has been done, is start to look at the actual content the site actually has. So looking at your landing page report, if you’ve done any level of SEO, you should have really pinpointed a keyword or a key theme that a particular page is looking to target. With that in mind, you’ll still be able to analyze how that particular page is doing in your particular time period, whether it’s month on month or year on year.

The other thing that you can do is, Dan Barker has created a great website called NotProvidedKit.com, which contains lots of goodies for you to plug into your Google Analytics account, such as advanced segments, custom reports, and dashboards. These are going to help to further refine your analysis of your not provided data whilst we still have keyword data to play with.

So those are the things that you can do now, the here and now, that you’re still able to capitalise on what data there is at keyword level. But the other opportunity that not provided presents is for marketers that haven’t yet started to broaden their approach. This is a real opportunity to start looking at what actual consumers are doing on your site.

So within Google Analytics, there are multi-channel funnels and conversion attribution, which have a whole lot of insight that probably is underutilised in the marketing world. With multi-channel funnels, this is where you can actually look at how people are coming to your site. It’s going to be very unlikely that every single particular user is going to be typing a keyword into Google and coming to your site without there being any other kind of complex combination of visits to your site. It may well be that they come to your site originally from an organic search, using a keyword, but they’re not necessarily going to convert at that point. It’s really the conversion behaviour that is going to give you the insights that are going to enable you to scale any kind of improvements across your website.

So, for instance, they might well use a particular keyword in Google to come to your site, but they may well not convert at that point. They may come back later on simply by navigating directly to your domain, or they might actually click on a particular advert that you had running in AdWords, or they might have seen some content that has been shared on a third party site, and they’ve come to your site then and at that point converted as well.

This closer analysis is going to give you an understanding of where your content best sits to lead to that particular conversion. If you think of the people that convert in isolation, the behaviour of those people is likely to hold the most useful insights for you to be able to make the quick improvements and even the long-term improvements on your site.

So what is it about those converters that have led to a conversion? What content on your site are they looking at? How are they moving around your site? Comparing that to the non-converters would be an ideal situation to start to understand what barriers there are for users to actually convert, actually pay you for your particular product or service, or actually give you their email address in signing up for your particular issue or newsletter.

So these two elements, I think, are very underutilised within Google Analytics, and they hold a whole host of insight, because what you’ll then be able to do is look at a particular piece of site content and say, “Why isn’t it being shared? What can we do to make that more sharable?”

At that point, you can run tests, such as A/B tests, or even more detailed multi-variant tests, as well, to really look at what your site structure needs to be and what your site content needs to incorporate to trigger that particular behaviour, that desired behaviour that you want from your users when they hit your site.

So not provided throws up some challenges. There’s no doubt about that, but it also throws up a great opportunity for web marketers to actually get to know their site and their users in a lot more detail.