Blog Series

Takeaways from Google's Hangout on RLSA for Analytics

Get the inside scoop from Google's hangout on RLSA for Google Analytics.

Last week, Google hosted a live hangout all around RLSA for Google Analytics (GA). Below are some of the key takeaways from this, and how they can be applied to search accounts. This is a feature that is being rolled out across accounts at the moment, so may not be live if you check your own accounts yet. This guide will help you get started once it is enabled in your account.

What is RLSA?

Firstly, lets (very broadly) cover what RLSA is. It’s short for Remarketing Lists for Search Ads (RLSA), and it allows advertisers to tailor search campaigns around a specific set of conditions that you choose – some of these being if a user has previously been onto your website, or if they’ve viewed a specific page and not converted. For example, if you have a retail website and someone has been on to your purchase page and not converted, you can show that user (or similar users) a specific ad designed to get them to return to your website, or also bid higher/lower for that traffic.

The Relationship With Google Analytics

Google Analytics (GA) allows you to use remarketing audiences using a GA tag, which will allow you to create very granular lists of users for you to segment. One of the benefits of this is that it is set up very easily (it can be done in under a minute) – and the bonus is that there is no need to update the tag once it’s live.

Checklist

In order for you to use RLSA with GA, there are a few things that you must ensure are done:

  1. Have a GA account with administrative/editing access permissions.
  2. Have at least one active AdWords account that is linked to your GA account.
  3. Agree to GA terms of service – this is important as if you try to set this up you may encounter a stumbling block if you haven’t accepted the most recent terms.
  4. Agree to adhere to the GA advertising features policy.
  5. Enable ‘Advertising Features’ in your analytics property settings. (You do this by clicking the ‘Admin’ tab, clicking ‘Tracking Info’ under the ‘Property’ tab, and switching both ‘Remarketing’ and ‘Advertising Reporting Features’ to ON (see diagram below)

rlsa

Building Your Audience Lists

So, that’s done – now on to the fun stuff. The ‘Audience Builder’ feature allows you to build a list of users based on a set number of categories – with the inclusions of ‘conditions’ and ‘sequences’ that let you tie them into one another. You can segment audiences by several factors allowing for smarter remarketing, such as:

  • Demographic (Age, sex, language, etc)
  • Technology (Device, OS, browser, screen size, etc)
  • Behaviour (Site actions, pageviews, etc)
  • Date of first session (often users can return and convert anywhere between a few days to a few weeks, months based on vertical)
  • Traffic source (self-explanatory – how they arrived at the site)
  • Enhanced e-commerce (When customers added items to cart, started checkout process, and completed purchase – you can see where users drop out of shopping funnel).

rlsa

For example – a very popular RLSA method is to segment and target those users who visit a website but do not convert. To set this audience up in GA with RLSA, simply click the ‘Behaviour’ tab and select the following conditions:

  • Sessions => 1 (Sessions that are equal to, or greater than one)
  • Transactions (per user) = 0.

Conditions and Sequences

‘Conditions’ allow you to segment users and/or their sessions according to a single or multi-session condition that you choose. For example, you could create an audience list with the following conditions:

  • Session date is between 01/01/2015 and 01/05/2015, AND
  • Transactions per user => 1.

This list will allow you to target customers who have made at least one purchase from the website between a specific time period – useful if there is a newer, or updated version of a specific product that is released, which you want to drive highly targeted awareness of.

Other conditions allow you to target by session duration per user, which you can use to tie into an audience list as you please. For example, if a user spends over 180 seconds (3 minutes) on a product page but doesn’t convert – you could build such an audience list to target those users specifically.

‘Sequences’ on the other hand, allow you to segment users and/or their sessions based on sequential conditions. For example, you could create an audience list with the following conditions:

  • Landing page contains ‘winter’, AND
  • Transactions per user => 1.

This list will allow you to target customers who brought a product from a page containing ‘winter’ in the URL – which is useful in certain verticals if you want to drive awareness of another seasonal range of product.

Smart Lists

This looks like a really cool feature, if used correctly. Smart Lists are made up of anonymous data that will only trigger when there are a minimum of 500 transactions and 10,000 daily page-views.  If this is the case, then your smart list will be generated anonymously around whichever factors that ultimately led your users to convert. If your site doesn’t meet these minimum requirements, then a smart list is generated around conversion data from businesses similar to yours.

To enable Smart Lists – when selecting the remarketing type, you must check the box labelled ‘Allow Google to manage my list for me [MY SMART LIST]’.

On top of this, the GA Solutions Gallery has a wealth of pre-defined lists that have been created by other users, and members of the GA team which you can apply to your own account.

Things You Should Avoid

Given the wealth of tools at your disposal to segment website users, it can be very easy to get carried away. Below are a few things you should definitely avoid:

  • Referencing a user, or how they behave in your ad messaging. This will likely not get past approval, and is pretty eerie. Avoid saying things like ‘You bought a t shirt from us, now get the matching shorts!’ No, just, no.
  • Over-segmentation – don’t get TOO granular with your lists. Segmenting users based on several conditions that may not be necessary can completely contradict the benefits of an RLSA strategy.
  • Finally, using the same RLSA strategy for hybrid (search and display) campaigns is a no-go. It’s likely that the targeting goals are different for each.
What To Remember

You can expand cookie length to 180 days, so depending on your vertical – it may be wise to start very broad lists, and then optimise down to your core list(s). Don’t forget, a list must have at least 1,000 users to be eligible to run – this is to protect the privacy of its users – so don’t start small, then scale up. Work the other way around and you’re good to go!