A recent internal survey across all levels of Tinuiti’s Paid Search team has revealed what experts believe will most impact Search advertising in the year ahead.
With over fifty responses spanning a range of roles (from Strategists all the way up to Director), the results show that over 80% of Tinuiti’s Search team consider automation and audience targeting features their highest priorities.
Automation & Audience Targeting Top Priority for Majority of Tinuiti Search Experts
Google & Microsoft have made significant investments in both automation and audience targeting features over the past few years.
The first step Google took as a dominant force of automation has been the launch of automated bidding.
Google’s automated bidding solutions have actually been around for nearly ten years. However, Google announced the evolution of automated bidding as “smart bidding” in 2016 and has since made significant advancements in its machine learning capabilities which include real-time auction-based signals.
If you asked a search marketer a few years ago if they recommend testing Google’s smart bidding solutions, the response would often be discouraging based on early results, and instead, a third-party bid tool like Kenshoo or Marin would be suggested.
Now the tables have turned, with 80% of respondents saying automation is their top priority and a full 100% of respondents planning on using at least one of Google’s smart bidding solutions.
Why Automation Doesn’t Mean Autopilot
Applying automation to search campaigns is by no means a flip of the switch or quick and easy solution, but instead is an extra layer of complexity in search optimization that needs to be carefully navigated with human intuition and fine-tuned for optimal results.
“While smart bidding has its place somewhere in most accounts, the new-age recommendation of using smart bidding across all campaigns has its flaws. While eCPC is able to use auction time insights for bidding, sometimes this results in higher CPC’s which doesn’t always generate a meaningful amount of impression share for the cost increase it incurs.”
— Steve George, Senior Search Specialist at Tinuiti
Although all bid strategies from impression share-based to conversion-based options are expected to be adopted, tCPA (target cost per acquisition) is currently recognized as the most effective option with 60% of survey respondents indicating this bid strategy as their first choice preference.
“In recent tests, we’ve actually seen manual bidding out-perform eCPC for certain verticals when looking strictly at CPC’s and impression share. This has allowed us to generate incremental impressions due to lower CPC’s allowing us to stretch our dollar further,” says George.
“This is not always the case as we have also seen tremendous results for clients who utilize smart bidding. Just like anything else, you need to test to determine which strategies are best for your account and campaigns as there is no one size fit all bidding approach for any client.”
Nearly All Tinuiti Search Specialists Plan on Testing Responsive Search Ads
The latest wave of automation features that Google has made available is through its new ad format, Responsive Search Ads.
While only officially rolled out to all advertisers this past October, Responsive Search Ads have already become widely adopted. Google turned quite a few heads when Responsive Search Ads were announced and disrupted the industry. How can you possibly automate creative assets in an effective way?
Instead of traditional A/B ad copy testing, advertisers supply Google with up to 15 distinct headlines and four description lines in one ad unit. Google will then mix and match the assets to the ideal audience and learn which combinations perform the best. Rather than running only between two to five ads, there are now hundreds or even thousands of ad copy combinations for the same brand.
45% of Survey Respondents See Stronger Results from RSAs than Expanded Text Ads; 55% See Same or Mixed Results
Responsive Search Ads should definitely be a feature that you consider testing, but not to serve as an all-encompassing ad copy solution. The best results come from a myriad of automation and audience techniques.
Remember, automated bidding has been around for almost ten years and has only more recently shown encouraging results. Enhancements to machine learning of Responsive Search Ads are likely to follow.
“In most cases, Responsive Search Ads (RSAs) are driving improved CTRs across a variety of client verticals. While conversion rate increases aren’t always quite as strong, RSAs can be used as a great “research” tool to help us better understand what combination of messaging is expected to engage our audience the most (per the platform’s algorithm).”
— Lauren Mahon, Senior Search Strategist at Tinuiti
“These learnings can then be used in subsequent RSA, ETA and landing page tests. We continue to refine how we leverage RSA results and look forward to enhanced reporting and perhaps the ability to tell the system that a particular HL or description position is off-limits (in addition to the current pinning options),” explains Mahon.
100% of Respondents Plan to Prioritize Audience-Specific Approach in 2020
Much newer than automation is the rise of audience data targeting features. Whether it be Google’s first-party data such as in-market audiences (e.g. people who are actively looking to buy a car) or a brand’s first-party data (e.g. remarketing or CRM lists), 100% of respondents are focusing on at least one audience-based approach to search.
When Google’s first-party audience targeting data was initially released a few years ago, the majority of traffic would fall into an “Unknown” bucket while attempting to target the audience rather than data being attributed to an actual audience type.
Audience data is becoming much more prevalent and sophisticated. Google just released Custom Audiences this February which gives brands the ability to further refine who you are able to target in search by creating “personas” that represent segments of your audience. For example, you can now target a combination of audience types such as parents who are fashion enthusiasts and in-market for buying shoes. Survey respondents indicated Custom Audiences as their first-choice response in terms of their approach to audience targeting this year.
Microsoft has also enhanced its audience offering with Linkedin demographic targeting released in early 2019. 66% of survey respondents indicated they plan on testing Linkedin targeting in 2020. There is now a plethora of audience data available through both Google and Microsoft that can be combined with automation to achieve goals.
Over 90% of Respondents Plan on Restructuring Campaigns to Adopt Automation & Audience Features
How should brands tackle the challenge of adopting both Audiences and Automation in tandem for 2020?
The answer is in the strategic approach of how advertisers target consumers and design campaign structures.
Only a couple of years ago, Tinuiti’s best practice was to segment campaigns based on device and keyword match type. However, that strategy has begun to shift recently as more and more search marketers are consolidating campaign components such as keywords and devices into a smaller subset of campaigns and ad groups. The more “signals” that Google has at its disposal will help enhance machine learning to show the right ad at the right bid to the right consumer at the right time.
“For my ecommerce clients, I’ve seen the best results from segmenting campaigns by audience, theme, and keyword match type. The granularity in the audience segmentation has allowed me to prioritize new versus returning visitors based on my client goals. For example, one of my retail clients is goaled on new customer growth.”
— Leslie Gzehoviak, Senior Search Specialist at Tinuiti
“Their campaign structure allows us to prioritize new customers, and we’ve seen new buyers increase 33% YTD YoY. In addition to the campaign audience segmentation, we have been leveraging smart-bidding aimed at driving more orders at a more efficient ROAS. YTD, we have seen overall orders increase by 36% without a decline in our ROAS.”
However, as there is a growing increase in consolidation, the number one response from Tinuiti’s survey regarding restructuring campaigns was actually an audience segmentation approach. Audiences except for CRM and remarketing lists are not considered in smart bidding algorithms, so it makes sense that a growing number of search marketers are turning to audience segmentation to deliver an effective performance-based search strategy. Being able to identify and segment your target audience while simultaneously leveraging machine learning gives advertisers enhanced control over delivering a relevant consumer experience and achieving performance goals.
“By consolidating campaigns from device segmentation to a unified campaign by all devices and targeting campaigns by first and third party audiences we were able to successfully build a consumer framework to take consumers from prospect to repeat purchasers.”
— Alex Turner, Search Strategist at Tinuiti
“This structure was fueled by SA360 bid strategies across search and shopping campaigns across one unified conversion metric which resulted in a drastic increase in customer lifetime value and full portfolio CTR while reducing cost per order drastically.”
The rise of audiences and automation does not mean putting all of your keywords and ad copy in one campaign, turning on a switch and crossing your fingers for positive results. Not all brands have the same quantity or quality of data, and not all components of audience targeting and automation are treated equally.
The balance of applying both segmentation and consolidation is an emerging art that needs to be carefully and consistently curated to ensure that we are not blindly adopting automation but instead taking a human approach to delivering optimal results.
Want to learn more?