Swarovski examined the relevance of generic search campaigns through geo-experiments

February 2019 Global Case Studies

In order to reach more potential customers, Swarovski is using geo-experiments to test the incremental value of generic campaigns, which are often relevant at the start of the customer journey. Per advertising spending, they achieved a profit of between two and three euros—a value significantly higher than last-click attribution previously had.

Goals:
Quantification of incrementality in generic search campaigns
Identification of potential for new customer acquisition
Validation of the data-driven attribution model (DDA)
Approach:
Geo-experiments in four key markets to measure the earning potential of generic search campaigns
Use of DDA for campaign control and measurement
Results:
Every euro invested in a generic search campaign yielded an incremental revenue of 2–3 euros
Increase of new customers through generic campaigns
Enhanced performance through DDA and optimization based on it

The production of the highest quality crystals has been the core of the Swarovski family business since its founding in 1895. Already in the fifth generation, Swarovski designs, produces, and sells the highest quality crystal, real gemstones, synthetic diamonds, and cubic zirconia, as well as jewelry, accessories and lighting solutions. The company sells its products around the world in 3,000 branded stores and increasingly online.

The initial question: can the additional earnings potential of generic search campaigns be determined?

Although bricks-and-mortar retailing still plays the biggest role, Swarovski is seeing a clear shift towards digital information behavior and e-commerce. While the Swarovski brand is established worldwide, according to the company, the optimization of keywords will play a key role in continuing to compete against other online retailers in the future and also in reaching new customers.


"Our market campaigns have so far been aimed at users who already know or consider Swarovski. Would it not be more effective to use generic campaigns to target users who have not yet chosen a brand? Would sales be boosted by doing generic search campaigns?"

– Philipp Dalmolin, Senior Search Engine Marketing Manager at Swarovski


The goal: To measure the added value of generic campaigns in a geo-experiment

In order to determine the added value of generic search campaigns, Swarovski, in close collaboration with Google, planned four geo-experiments in the key markets of the USA, Canada, Great Britain, and Australia, where generic campaigns are to be run in addition to brand campaigns. As a first step, each country was divided into comparable regions, which were then randomly assigned to a test and control group. Additional generic keywords and jewelry categories were added to the test group, while only brand-specific search terms were still offered to the control group.

The result: The incremental value and new customer potential of generic campaigns is significantly higher than measured by last-click attribution

The geo-experiments showed a clear result: The generic search campaign resulted in significantly higher sales and significantly higher turnover than was previously measured using last-click attribution. The calculated incremental ROAS (Return on Advertising Spend) was two to three euros.

With revenue growth and improved profitability ratios, the path to take was clear: Generic campaigns were set up and budgeted for in most key markets—today they make up a large part of the marketing budget for e-commerce.

The next step: data-driven attribution

Consumers usually use generic queries at the beginning of their research. Through the geo-experiments, Swarovski was able to measure the earnings potential of these search queries for the first time. Based on the successful tests, Swarovski decided to switch to a data-driven attribution model for the campaign evaluation and control. Unlike last-click attribution, which allocates a conversion to the last click before the purchase, the data-driven attribution model takes all touch points and devices in the conversion path and their pro-rata value contribution into consideration.


"The geo-experiments have given us the right arguments for data-driven attribution. This takes greater account of the touch points that are relevant in earlier phases of the purchase decisions and for new customer acquisitions."

– Philipp Dalmolin, Senior Search Engine Marketing Manager at Swarovski


Swarovski 2
How Kytary uses Smart Shopping campaigns to quickly expand into new markets