It’s no secret that Paid Search is shifting more and more towards automated bidding strategies. With the improvements that have been made over the last year, it’s a great time for advertisers to try some of these strategies. Each case is different, and it’s not always going to be a home run switching every campaign over to an automated bidding strategy from a manual one. Generally, though, Metric Theory has found a lot of success switching many campaigns to automated bidding. Today, we’ll focus on four successful cases of testing tROAS bidding on Google Shopping campaigns, how we did it, and the results.

Shopping Automated Bidding

Best Practices for Testing tROAS Bidding

As a best practice, the first step in switching a campaign to tROAS is to set the tROAS threshold right around where the shopping campaign ROAS was previously. After allowing for the bid strategy to learn, you can then raise and lower the tROAS threshold based on performance.

Another of the keys to seeing success with tROAS is to allow the campaigns to fully learn before making any other changes. While it can be tempting to turn off tROAS if spend spikes without revenue immediately following, it is essential to allow the algorithm to fully learn before making adjustments.

Finally, turning on tROAS bidding is just the first step. Testing any automated bidding strategy is also about monitoring and revisiting your bids to drive more results. Just because something is automated doesn’t mean there’s no strategy or analysis work necessary to improve performance over time. It will save you time in the long run when you find the winning formula, but it doesn’t mean you’ll be able to just let all of your advertising run on autopilot – sorry to break it to you!

The following examples demonstrate the power of automated bidding when goals are set correctly, the test is given time to run, and you do the followup work to continue testing.

Growth Tests

Advertiser 1 implemented tROAS on their non-brand shopping campaigns back in January, and compared to the previous period, grew revenue 478% while also improving ROAS 17%. After seeing the early success with tROAS on non-brand shopping, we switched our branded shopping campaigns to tROAS in February and have seen over 110% increase in revenue while still maintaining a ROAS well above the tROAS target. Considering the previous period in both of these comparisons contains Q4 (where we see a lift in sales for the holiday season), we are especially pleased with the revenue growth since launching tROAS.

Advertiser 2 is also heavily growth-focused, and has been testing tROAS bidding on shopping campaigns for the better part of a year. We were able to grow our shopping revenue by over 300% while gradually improving shopping ROAS by 44%. While this growth is also combined with a shopping campaign restructure, these are some fantastic results. For both advertisers 1 and 2, when we noticed the early success of tROAS bidding, we lowered the tROAS thresholds on our shopping campaigns which allowed us to drive more volume.

Efficiency Focused Tests

Advertiser 3 is more focused on efficiency, so we chose to test tROAS in shopping campaigns in an attempt to improve our shopping efficiency. Compared to last year, ROAS increased 15% in a heavily competitive market with a 30% increase in revenue.

Advertiser 4 is also more efficiency-focused and is facing the challenge of revenue declines. As a result, we launched tROAS in the hopes of mitigating some of the sitewide revenue declines. Since launching tROAS with this advertiser, shopping revenue has grown over 150% with a 20% improvement in ROAS. These improvements have helped mitigate declines from other channels to keep the entire website as a whole steady in revenue. For both advertisers 3 and 4, we increased the tROAS thresholds after the learning period in order to continue improving efficiency.

Conclusion

Thinking of automated bidding as a “set it and forget it” feature can be harmful in the long run. While the algorithms need time to learn, it is essential to consistently monitor and re-evaluate the performance of any automated bidding strategy you’re using. With the combination of automated bidding, allowing the tests to run, and closely monitoring performance, we were able to surpass these advertisers’ goals.

Looking to begin or expand your automated bidding strategy? Our team is happy to help.

Share: