Better attribution and automation drive generic search gains for Nordic Choice Hotels

September 2018

Nordic Choice Hotels is one of the largest hotel chains in the Nordics, with around 200 properties. Operating in a highly competitive marketplace against sophisticated competition, the brand used machine learning to improve their attribution and automation, increasing bookings from generic search and overall cross-channel revenue.

About Nordic Choice Hotels

Nordic hotel company with approx 200 properties in the region


Get more customers to book directly via their own website

Become more sophisticated in their digital marketing

Become more competitive in generic search auctions to win more bookings


Activated cross-channel and cross-device attribution with Google Attribution (Beta)

Analysed the real attributed value of generic campaigns

Set up Target ROAS automated bidding strategies to act on the insights in real time


The combination of machine learning in attribution and automation produced:

45% increase in revenue from generic search campaigns

22% better ROAS from generic search campaigns

Overall cross-channel revenue increased in the period, driven by strong revenue growth and investment in generic search

The hotel category in the Nordics is highly competitive, with hotel chains and online travel aggregators (OTAs) all competing for the same customers. A particular challenge for hotel chains has been getting consumers to book directly with them rather than through the aggregator sites, where consumers usually have greater choice and a more familiar customer experience. With an intermediary between them and the booking, hotels lose out on both revenue and the chance to build better relationships with their customers.

Nordic Choice Hotels realised that the best way to address this issue was to increase their share of direct bookings by becoming more competitive in paid search auctions for generic terms. However, this is no easy task. Travel aggregators have large performance marketing budgets, and tend to be some of the most sophisticated advertisers in any auction. To boost their chances of success, the brand decided to adopt the latest machine learning technology in both attribution and automation, which would allow them to maximise the profit from their investment.

Improving Attribution

Nordic Choice Hotels originally used a last click attribution model inside Google Ads, which only assigns conversion value to very last Google Ads click in the customer journey. This was problematic as their data showed that within paid search alone, 38% of their conversions involved multiple ad clicks.

To address this problem, they turned to Google Attribution (Beta), which uses a data driven model to calculate the value of each marketing interaction. By comparing the paths of customers who convert to those who don’t, Google Attribution (Beta) allocates credit across devices and channels, providing more accurate and actionable data.

Bidding Smarter

With Google Attribution (Beta) data imported into Google Ads, the brand could finally evaluate the true effect of their generic campaigns. The next step was to put the insight to work in the auction by adopting a Target ROAS automated bidding strategy.

Target ROAS bidding uses advanced machine learning to predict each user's probability of conversion and estimated conversion value, and then sets the optimal bid for each auction. The algorithm is even more effective when it has access to full attribution data rather than having to rely on last click, so Nordic Choice Hotels applied the data from Google Attribution (Beta) to the bidding algorithm and then selected a realistic ROAS target for their generic campaigns portfolio. With the attribution model and automated bidding in place, the system optimised media spend at scale and in real time.

Boosting direct revenue

The result of this project has been a significant improvement in the performance of the brand’s generic search campaigns. Looking year-on-year at the three month period after the switch reveals that across both the Norwegian and Swedish accounts, generic campaigns drove 45% more revenue with a return on ad spend (ROAS) that was 22% better. The brand also saw overall growth in cross-channel revenue, driven by the strong performance from generic search.

Google Attribution in combination with a Target ROAS strategy enables us to do more effective marketing and provide better insight. 22% higher ROAS and 45% increase in revenue is impressive

– Henrik Zahl Andersen, Digital Marketing Specialist

Attribution modelling at Nordic Choice Hotels has all of a sudden moved from being a very theoretical exercise to an integral and automated component of our always-on activities.

- Björn Alvarsson, Director of Digital Sales & Marketing

The combination of Google Attribution (Beta) and Target ROAS has allowed Nordic Choice Hotels to not only increase the value of their search marketing campaigns, but also to free up time for higher value tasks. With machine learning doing the automated heavy lifting, the team are able to give more of their attention to audience strategy, market analysis and competitive intelligence.

5 pillars of a holistic marketing measurement plan