Google Shopping: Increased eCommerce ROAS by 73%
Our client, located in the United States, maintained a successful eCommerce business with a Google Shopping component. The Google Shopping campaigns had recently stalled in performance as competition and products were changing. Search keyword campaigns were not producing the volume and opportunity they once did as the searches were more relevant to exact part numbers and manufacturer combinations. We saw an opportunity to potentially drive increased performance from the Google Shopping campaigns while growing Return On Advertising Spend (ROAS) at the same time. The Google Shopping campaigns had the natural position above the search results and could be made to reflect the search intent almost exactly.
After careful analysis and client consultation, Tribbute formed a strategy for implementation and growth with the Google Shopping campaigns to drive revenue production and increase ROAS without any significant changes to the Google Shopping feed.
The required baseline for ROAS performance from the client had increased, along with a declining volume of traffic in search campaigns that traditionally produced much of the campaign value for the client. Google Shopping traffic volume had remained flat and did not present an immediate opportunity for growth in revenue or ROAS.
How Tribbute Helped
Data generated by a revenue-attribution solution, developed by Tribbute for the client, identified a significant pool of mobile visitors that had not completed a purchase. The data also revealed that many non-purchasing visitors were repeating searches for the same or related products around which, their intents has produced their initial visits. Tribbute assessed the size of these audiences and tested AdWords’ shopping solutions focused on re-acquiring these newly identified audiences. Specifically, RLSA and audience bidding strategies were focused on.
Tribbute integrated the audiences through RLSA and Google Shopping, initially without any bidding changes for the purposes of collecting data. Once enough audience data was generated, Tribbute implemented a custom bidding strategy, composed specifically for this experiment. Campaign results changed immediately. The data illustrated that most product groups were impacted by the changes. Some groups showed ROAS and revenue volume improvements. Other product groups were negatively impacted. Tribbute identified a series of signals that were integrated into updated bidding formulas. The resulting active bidding adjustments implemented at the product group level produced led to the development of a new algorithm to determine the exact changes required to optimize outcomes on a regular cadence.
Since those changes have been made, our client continues to see ROAS results over 2x higher than the pre-experiment baselines with increased volume from the impacted product groups in aggregate. Tribbute continues to capture and analyze the data while executing experiments focused on additional opportunities.
The initial, baseline ROAS for the Google Shopping campaigns resulted in marginal profitability for our client after their other costs were factored in. After implementing and establishing the changes found through the performance analysis Tribbute completed, gains in revenue were being realized while ROAS improved 50% after the first round of experiments and adjustments. Further testing, data segmentation, and analysis showed additional signals of potential performance increases, so a secondary strategy was implemented using data AdWords, Analytics and Tribbute-generated sales-funnel data.
The next round of experiment iterations produced data showing greater increases in revenue production and another ROAS improvement 23% relative to pre-experiment baselines. The campaign ROAS stabilized at these levels which are exponentially profitable for the client considering the break-even baseline status relative to all non-media hard costs associated with fulfilling their eCommerce orders. During the year since these improvements, Tribbute has successfully optimized for volume improvements that are in excess of 100% at the time of this case study.
Increased Conversion Volume by 135% Through Bidding Strategies
Our client provides services in Canada and the United States and wanted to increase the volume of transactions at a target cost per acquisition (CPA). When Tribbute’s engagement began, the client was acquiring transactions at a CPA lower than their target, suggesting opportunities to increase volume at effective per-unit costs existed.
After an analysis of the current data was completed, Tribbute confirmed the client’s belief and presented a plan to take advantage of the opportunities. Using specialized campaign structures focused on individual markets and unique bidding tactics developed specifically for this client, Tribbute-managed digital campaigns increase in transaction volume by 135% while remaining 24% under the target CPA. We continue to find volume increases with existing and new campaigns for this client that invests seven figures annually into Tribbute-managed digital marketing and acquisition initiatives.
Experiments illustrated that simply increasing bids to produce a higher cost per acquisition (CPA) would result in small volume gains. But when managing to an aggregate CPA, many outlying campaigns exceeded the client’s defined limit. We knew the volume opportunity existed, however, the client-defined solution would require more than bid increases. Each geographic market/location served by the client was segmented by a campaign according to several targeting dimensions (geo and day/time parting, intent matching and others) making each volume increase opportunity unique by geo-location and statistically significant data sets across dimensions rare.
How Tribbute Helped
We wanted to employ Google data to predict conversions from clicks and acquire to a target cost per acquisition (CPA) when the data we have access to is light or aggregated across diverse dimensions as was the case with this client. In a select few campaigns, Tribbute launched an experiment to verify the fitness of this methodology with this client’s account. It was quickly revealed that the solution was going to work for some, but not all geographic markets.
Our analysis led to a new hypothesis that by applying some custom rules around how and when Google bidding strategies were used, performance would improve and stabilize across all markets and segments. The experiment results did not disprove the hypotheses and the new process was launched account wide after sufficient testing was completed.
After account-wide implementation of the bidding strategy, and the development of campaign expansion processes meant to take the best advantage of the strategy, certain geo-locations were judged to have only limited incremental volume as an opportunity. However, this group was in the minority. Most geo-markets were able to grow materially when the strategic increases to the target cost per acquisition (CPA) were implemented with Tribbute’s custom methodology and process.
The results is a 135% (conversion volume) increase while maintaining allowable cost per acquisition (CPA) at a rate 24% (in aggregate) below the target set out by our client. All of is being accomplished every day as of the time of this writing without any material segment of this multi-million dollar investment exceeding client defined CPA limits. Years in to this engagement, Tribbute continues to find new areas of transaction growth for our client, but this process is a standard that we use to capitalize on ongoing volume increases.
For more examples of how Tribbute can help, please contact us today!