TUI France: Building a data-driven media strategy on Google Marketing Platform and Google Cloud
With CRMint, Google Marketing Platform, and Google Cloud AI, TUI consolidates its data into one place and analyses it for more effective ad targeting and reduced media spend.
Google Cloud results
– Drives insight from customer data with CRMint, Google Marketing Platform, and BigQuery
– Identifies which customers are most interested in making a purchase with 91% accuracy using Google Cloud AI
– Enables more relevant ads with automated, data-driven decision-making for lower media costs.
– CRMint helps reduce cost per sale on display by 7.5x
A subsidiary of the multinational TUI Group, TUI France is a leading tour operator in France, offering a wide and diverse range of travel and accommodation for individuals as well as businesses in 100 destinations. TUI’s offering is available in 4,000 travel agencies and stores and on its showcase website, which was among the top 20 most visited travel websites in France in 2019, according to Médiamétrie.
But online travel is a competitive market, and TUI France has needed to continually adapt and build on its successes. In late 2018, TUI France sat down with marketing specialists and Google Cloud partner Artefact to hammer out a digital road map. Artefact had already helped TUI France fine-tune its marketing campaigns using Google Marketing Platform and products such as Google Analytics and Google Ads. The goal now was to make use of all the marketing data available to optimise the company’s strategy. To make that happen, TUI France turned to Google Cloud.
“We wanted to bring data-driven decision-making to our media strategy to keep our current customers happy as well as attract new ones,” says Hamis Badarou, Head of Digital Analytics at TUI France. “We’re already big users of Google Marketing Platform, so using Google Cloud to power a data platform was the obvious choice.”
360 analytics with Google Cloud and Google Marketing Platform
Effective marketing campaigns are great drivers of business, but in order to last, they have to be value for money. TUI France wanted to reduce the cost of its ad spend and media investments.
“The travel industry is very competitive, so we wanted to be sure that TUI France was investing in the right channels, at the right time, and targeting the right prospects with high potential for the business,” says Natacha Kocupyr, Head of Media Consulting at Artefact.
To achieve that, TUI France would have to take its data analytics capabilities to a new level. The first step was simply to consolidate the data all in one place. Using Google Marketing Platform for its campaigns had already led to a wealth of data that the company was investigating with Google Cloud products like BigQuery.
However, there was also a significant amount of useful data on customer behavior in TUI France’s Customer Relationship Management (CRM) platform. Together, the two datasets would form a powerfully compelling picture of TUI France’s customers and what they needed. Unfortunately the CRM data and the data on Google Marketing Platform were stored separately and in different formats. Reconciling them would take significant resources.
In 2019, while TUI France was deep in the planning of its digital transformation, Google Marketing Platform released CRMint, an open source orchestration tool. Hosted on App Engine, CRMint allows companies to link their CRMs with a host of Google products. TUI France’s situation presented a great challenge for the new tool, so representatives from Google Marketing Platform worked closely with data scientists from the company and Artefact to help integrate the CRM data with what had already been collected in BigQuery.
“Getting the CRM data, Google Marketing Platform, and Google Cloud to talk to each other seemed like a difficult task, but with an orchestration tool like CRMint, it wasn’t hard at all,” says Hamis.
With the two main datasets harmonized and collated in BigQuery, TUI France and Artefact could concentrate on the next phase: predictions. Data scientists from both companies analysed the data using BigQuery and fed the results through a machine learning algorithm built with AI Platform. The algorithm helps TUI France figure out which customers visiting its website are just browsing and which ones are looking to make a purchase.
It was important for Artefact that the platform was adaptable. “We didn’t just want to come up with a solution for one day. It needed to be able to detect changes in the data and respond accordingly,” says Natacha.
For that reason, the algorithm is retrained with fresh data every week. Using a cloud-hosted service like AI Platform meant that TUI France didn’t have to sink costs on hardware that would only be used once a week.
“The beauty of AI Platform is that it works locally as well as in the cloud,” says Iliass Benryanne, Data Scientist at Artefact. “That meant we could test it locally, and then when we went into production, we could host it on App Engine and not have to worry about scale.”
Based on the AI scores, TUI is able to not only prioritise bids, but work on lookalike strategies for its display campaigns, identifying key similarities between people from its original custom audience and finding similar audiences to reach in a new campaign. This enables TUI to target its messages to people who are more likely to book its offers.
Greater precision, more effective targeting
After six months of planning and testing, and a further period of building and refining, the new data platform went into production. By early 2020, TUI France was already seeing dramatic improvements. Consolidating all of its customer data onto one platform provided a number of powerful insights about customer behavior that simply weren’t available before. The AI predicted the customers most interested in making a purchase with 91% accuracy, leading to a much more effective media strategy.
“Google Cloud has helped us tighten the focus on our media channel targeting and make our ad spend work harder,” says Hamis. “On our display channel, the cost per sale is 7.5 times lower than before while the cost of bringing customers to the website has gone down by 29%.”
But for TUI France, the work doesn’t stop with lowering the cost of customer acquisition. The insights gained have opened the door to all sorts of other projects. In the short term, TUI France is building a recommendation engine that will feed emails and the website with relevant content. Customers who visit the site will see a personalised homepage with banners and offerings tailored to their needs.
“Combining Google Marketing Platform and Google Cloud in this way has led to some really exciting results,” says Hamis. “I’m keen to expand this type of initiative across the entire marketing department so we can have some truly data-driven decision-making.”