Looking for a PDF to take on the go? Fill out the simple form on the right, and we’ll instantly email you a copy of our complete guide to AdWords retargeting. If you have any questions, on this or any other digital marketing topics, you can always reach us at info@3QDigital.com or by filling out our handy Contact Us form.
Remarketing through AdWords has been part of every smart SEM’s toolbox since launch several years ago. With rare exceptions, it just works and brings additional value to every other channel you invest in by re-engaging your already warm prospects.
At the most elementary level, most websites can drive great ROI with basic remarketing: simply targeting all people who have been to the site with generic messaging. That said, super-savvy SEMs are taking advantage of Google’s ever more sophisticated tools, targeting options, and products that facilitate super-advanced remarketing.
In this whitepaper, we’ll discuss basic remarketing techniques using the AdWords universal remarketing pixel; we’ll then move into advanced retargeting uses for Remarketing Lists for Search Audiences (RLSA), Search Companion Remarketing, Similar Users Remarketing, YouTube Remarketing, and how to use Google Analytics for Remarketing.
Let’s kick it off.
Basic Remarketing With the Universal Pixel
In developing an advanced remarketing program through AdWords, first, shore up your core basic remarketing. The release of the universal remarketing pixel in AdWords in the summer of 2012 facilitated a huge leap in the logistics of more sophisticated remarketing. Audiences can be created on the fly now from the universal pixel without having to generate or keep track of dozens (or hundreds) of pixels, coordinate with web dev to place them, or QA and maintain them. Previous to the universal pixel, the logistical investment made super-segmented audiences too weighty to maintain – the payoff just wasn’t there for all the time investment.
Now, it’s simple. Once the pixel is placed, it’s fully in the control of the marketer (no dev wrangling!). Though too much segmentation can still become cumbersome, you can now think deeply about the main factors that show where the user is in the funnel and segment your users and messaging for each segment. You should be making tight custom audiences that align to your funnel, products, and purchase cycles.
ASK YOURSELF THE FOLLOWING QUESTIONS:
Where are the users in your funnel?
Understand your funnel and segment users by the stage they are in.
For ecommerce, the funnel may look like:
Research –> category pages –> product pages –> cart –> thank you –> current customer –>LTV
For B2B, it might be:
Research –>Download Whitepaper –> Free Trial –>Paid Segment each of these groups and align bidding and messaging to the funnel stage your users are in.
What product(s)/or services are your users interested in?
Using the ecommerce example, you can hone in on specific products or categories of interest. For lead gen, you may be looking at the content they have read that indicates their use case of your product. Travel advertisers may segment by destinations. Layering on where the user is in the funnel with the product they are interested in means tight messaging opportunities, e.g. for users interested in Caribbean Vacations who have added to cart but not checked out, a discount code may be just what they need to pull the trigger.
When did the users last visit?
Understand your funnel and segment by the cookie lengths that make sense for you. For an ecommerce site, you may find that the optimal time to target a cart abandoner is 1-3 days after abandonment. B2B advertisers may find that 90 days of slow, steady value props is the right pace.
In the travel example above, there will be a sweet spot with timing.
At some point, those users may choose another vacation or decide to get a new computer instead of a vacation this year, and you should either stop serving them ads or increase the urgency of the call to action. Your analytics will give you a good starting point; from there, continued testing of optimal cookie length and the aligned messaging stages is key.
When is the right time to kill the pixel?
If somebody buys a book online from you, they may want to buy another book the next day – especially if it is around a holiday. If they buy a sofa, you can probably kill that pixel when that person converts, since it’s very unlikely that they will be buying another sofa anytime soon. However, 14 days later when the sofa arrives, they very well may be in the market for some accent lamps, so starting new messaging cycles may make sense. Use audience exclusions in Google to ensure that you eliminate the right people when they move to the next stage in the purchase cycle.
What is the right impression cap?
As a general rule, the shorter the period of time you are targeting a person, the more liberally you can hit them with impressions in any given day. If you are serving a user impressions for 3 months, hit them less frequently in a day. This keeps people from developing negative associations of your brand, avoids oversaturation, and won’t mess up your CTR metrics. Err on the side of conservative and test what impression caps work best for your objectives. Also note impression caps are set at the campaign level (in the “Campaign Settings” called frequency capping):
So, in architecting your campaign structure, think about how caps attach to audiences to determine what should be in its own campaign. For instance, if you have a B2B product and a B2C, product the B2B is probably a much longer sales cycle where your audiences will align with that long funnel and lower impression caps. Your B2C product may have quick periods of segmentation over a couple days and a higher frequency cap.
Remarketing Lists for Search Audiences (RLSA)
The introduction of RLSA in Feb 2012 as a beta product and the subsequent broad release in spring 2013 represents one of the biggest changes to AdWords in years. Most marketers are under-using this high powered product. The use cases are vast, and we are seeing amazing results using many different techniques, including the following:
1. Bid up on previous visitors. This is the most simple execution: if somebody has been to your site before, has not converted (use an exclusion audience for converted users to make sure), and is still searching for your keywords, then just boost the bid using an enhanced campaign % modifier on this audience.
Ensure that while they are still in the market for your products (indicated by still searching the relevant keywords), you are showing very prominently in results. We have seen ecommerce advertisers well more than double conversion rates and halve CPA on these users with this simple audience and bidding tweak.
2. Loosen match types. As a next stage to example 1 above, consider loosening up your match types for previous visitors. This requires creating new campaigns or ad groups that only target previous visits. In our agency, we heavily use exact and modified broad match to hone in tightly on our high-performance keywords. However, for this highly qualified audience of previous visitors, letting Google match queries to more loose broad match drives additional, very high-quality traffic.
3. Use Exclusions. There are several reasons not to show ads at all to user segments that have been to your site. In expensive, high-touch verticals such as EDU and real estate, you’ll want to stop ads for users in your nurturing cycle; you don’t want to pay for any more high-cost clicks once you are nurturing these leads already.
Don’t show PPC ads on brand to previous users. Don’t show PPC ads to people who have already converted (real estate, EDU, etc.). By excluding users who have filled out lead forms or opened emails, you can shift more spend to sourcing new leads only. We have seen advertisers who were spending over 30% of their investment on unwanted duplicate leads – with RLSA exclusions, those dollars can be reinvested in sourcing new leads only.
Here’s another exclusion opportunity: every SEM has heard a client say, “Why would I pay for brand clicks when I have the #1 organic result?” I would argue that most people should still pay for those clicks; however, RLSA exclusions mean that you can only target new users, or users who haven’t been to the site recently, thereby eliminating people who are using Google navigationally and were coming to the site anyway.
4. Expand your keyword list to the broad related terms that you typically can’t afford. This lets you edge into more auctions at lower CPCs. Some advertisers could go as broad as “gifts” or “business solutions” when only targeting previous visitors. For others, it might allow you to go after a top-funnel head term that is prohibitively expensive, such as toys or furniture.
If you want to maximize the volume from this method, consider using Dynamic Search Ads layered with RLSA audiences. This allows you to forgo keywords altogether and liberally target the pre-qualified users on the terms Google “scrapes” from a site crawl.
5. Tune your messaging. If you create isolated campaigns that target users via RLSA, you can tune your messaging with really strong calls to action or consider offer codes or offer extensions to get users to pull the trigger.
Search Companion Remarketing
This is the AdWords take on what is called “search retargeting” in the traditional display world. You enter in keywords that a user has searched for recently and then follow them around the internet via the GDN (Google Display Network) with your text or image ads.
We have seen for some advertisers that these users are even more valuable than their core “visitors” remarketing segment.
Obviously this depends on the product and media mix. This is worth trying for any advertiser who has a strong head term, and it can sometimes be a way to circumvent super-expensive auctions that the advertiser can’t otherwise afford. For instance, maybe you can’t afford to bid $50 “MBA” keywords, but through the GDN, you can follow somebody whose searches showed intent to get an MBA. Similarly, competitor name keywords or tangential keywords are great to test here, as quality score metrics often make the search auction unsustainably expensive even though they can be great indicators of interest.
Similar Users Remarketing
This Google functionality is called “lookalike targeting” in the display world. It requires that you have a conversion pixel installed and a fair volume of conversions. (The more conversions, the better; for low-conversion-volume B2B sites, for instance, this may not be a good fit.)
This targeting algorithmically looks at the demographic, psychographics, and behavioral patterns that your converted users have in common and serves your ad to similar users. We have also seen this targeting perform better than standard remarketing lists of visitors who didn’t convert:
Combining this target with CPA-based bidding leans even more on algorithmic matching for the impressions you get. For advertisers with sufficient conversion volume, this is typically a great performer. That said, there is rarely huge volume associated with this method.
This is also called product retargeting in the display world. This targeting reaches previous site visitors with dynamically generated creative that shows products they have looked at (or similar products).
Typically this would only be used by ecommerce sites – although as the product matures, there may emerge ways to dynamically update the ads based on content areas of a site uses have visited.
Activation of dynamic retargeting requires help from a Google rep (as of Q3 2013). Additionally, it requires an active and healthy Google merchant feed. For advertisers getting good results from PLAs, this is a great way to extend the exposure for a highly valuable user segment to product images and info on the GDN. For advertisers with larger budgets, this is a great way to test dynamic retargeting without large investment or commitment – and, if it works, extend the reach via channels outside of AdWords (such as Criteo, Tellapart, and Dotomi).
Dynamic Remarketing has delivered not only data-driven results, but has also been a great branding tool for Rent the Runway,” said Rent the Runway online marketing manager Jordan Deodato. “Our remarketing program has accounted for 7% of our total paid marketing revenue in its first month since launch. Furthermore, it has engaged some of our habitual browsing customers with relevant, attractive ads, making them more likely to convert.”
YouTube remarketing allows you to remarket to users who have completed a number of actions within your YouTube channel. Most commonly this would be used to target a user who has viewed one of your videos, but for advertisers with very active YouTube channels, segmenting the targets and messaging even further makes sense. For instance, you could target users who have shared one of your videos, or those who have not yet viewed a certain video but have viewed another.
Set-up for YouTube remarketing is pretty easy:
Google Analytics for Remarketing
Remarketing via Google Analytics was released at the same time the Google universal remarketing pixel was released. This form of remarketing does require a code update to Analytics; this allows Google Analytics data collection and AdWords remarketing to run off the same universal pixel.
The AdWords remarketing pixel buckets users via inclusion and exclusion only based on what pages or sections of a site a user has visited. The Analytics pixel goes deeper, allowing retargeting based on metrics including things like time-on-site, geography, referring domain, order values, etc. This is most useful on sites with a high amount of traffic and multiple conversion funnels where the granularity can really pay off.
Remarketing is becoming a more complex and powerful tool in an advertiser’s arsenal, and Google is releasing more and more tools to help advertisers take advantage of this in the most efficient manner possible. Cutting-edge digital marketers are utilizing these new features to stay ahead of the competition and get the most out of their advertising dollars.
About the Author
Susan Waldes has worked in the search engine marketing industry since 1999; she joined 3Q Digital in August 2011 after serving as the in-house SEM manager for ecommerce retailer ivgStores. Susan has handled a multitude of lead generation, branding, and eCommerce clients in her previous roles at ROI Revolution and Rockett Interactive and as an independent SEM consultant. Susan has a BFA from Savannah College of Art and Design and has contributed insights about SEM and client relationships to our in-house blog and other highly regarded outlets, including Techipedia.com and SearchEngineWatch.com.