In the past year, the phrase “data-driven” has appeared within the context of almost every marketing discussion. But for marketers to truly realize the promises of big data, they need visibility into all customer interactions, behaviors and motivations — across every single touchpoint.
But bringing all marketing insights into one view is arguably the biggest challenge. Most are staring down mountains of data silos, with a great deal of uncertainty around the accuracy of the analytics that they are getting from separate sources such as another business unit, partner or even their own martech tools. Thereby, leaving stakeholders with segregated and often, untrusted, unusable data.
In 2019, marketers are pushing to get a clearer picture of how everything is working together and will need to connect the “pipes” to break out of the marketing silos. These silos include websites, digital, email, social, Salesforce and more. With this new, complete picture in-hand, marketers can then more effectively make predictions based on past buyer behaviors, focus on their most valuable target segments, and more efficiently personalize their digital campaigns. Peers in the industry will often use the code name Everest for these type of projects as the process of getting the data trusted and in one place feels equivalent to climbing Mount Everest. While a data warehouse is an ultimate goal, by starting with a few key data sets, proper implementation, and standard API’s immense value can be garnered very quickly and without a great deal of friction.
Take World Surf League (WSL) as an example. In March 2018, after consolidating several live events from major surfing brands onto its own video streaming platform, the WSL needed more comprehensive insights about its viewership; with the ultimate goal being able to provide a more targeted and personalized experience for their fans, while driving down acquisition cost and unnecessary marketing infrastructure spend for the brand.
For marketers like WSL, beginning a data-driven journey proves incredibly valuable, but takes a few initial, critical steps. The first — auditing and cleaning your data so it can be trusted — may take the longest, but is essential.
1. Trusted data: Sorting data into one place with a unified data model (data dictionary) seems self-explanatory, yet you’d be surprised how many different places companies store the numbers that have the power to fuel future growth. Once it’s in one place (data warehouse), the data is easily visualized in a dashboard. As a company takes this data-centric journey, they often start with siloed dashboards (media, website, CRM), then as the company’s processes mature, these datasets are merged, helping customers determine the right place, time and message for their customers (optimizing).
2. Segmentation: With some or all of a business’s data in one place, advanced data science approaches can be utilized. Using unsupervised Machine Learning techniques, segments can be created from the data. Machine Learning can pick out the most important clients based on different attributes. Through supervised Machine Learning it is then possible to predict what segment a customer will fall into before they even interact with your content; moreover, dynamically predict their lifetime values of customers in each segment. This is adjusted after every transaction, providing decades of perspective and allowing marketers to more quickly understand which prospective customers are most valuable — and more efficiently act on that knowledge.
3. Activation with Personalization: Does a high-value customer segment (or potential customer) prefer a phone call? Will they more likely respond to email targeting or social media outreach? Or be responsive to message A over B? It is now possible to target specific segments with personalized messages and channels based upon the predictive modeling of new clients. Having predictive data greatly increases the odds of getting communication right the first time, creating large ROI swings.
After a series of activations and implementations using the aforementioned tactics, the WSL received results that would further prove the effectiveness of data-driven marketing. By getting smarter about their ad spend, they reduced their cost per conversion by more than 50%, doubling their ad spend efficiency aimed at driving fan engagement and content viewership. By focusing on higher ROI customers, they drove a 30% increase in concurrent viewers to their live events and a 31% increase in sessions per user to their website and app. By moving its infrastructure to Google Cloud, WSL was able to scale their use of resources to match the demand of traffic to their website, mobile, and connected device apps. This means that during their offseason, they reduced infrastructure costs by 70% by scaling resources down.
This mix of Machine Learning and personalization isn’t a new idea overall, but in some ways, it’s capabilities are still barely understood by those who need it most. By connecting the data pipelines and implementing the right Machine Learning tools and platforms such as Google Cloud, marketers can create predictable customer growth now — and target content and even products to the businesses most valuable segments w in the future.
Break down the silos, build trust in your data, and empower your decision makers to dynamically adjust the marketing mix based upon data. These are the keys to seeing large returns and changing and transforming your organization’s marketing forever, and allowing a company to truly become “data-driven.”