Cloud technology allows marketers to access data more easily and uncover insights more quickly. Alison Wagonfeld, vice president of marketing at Google Cloud, shares three strategies to successfully integrate data, analytics, and machine learning—and increase your marketing productivity.
Over the course of my career in various entrepreneurial and marketing roles—in venture capital, research, and technology—I’ve been passionate about helping people use technology to accomplish tasks they couldn’t before. As the marketing lead at Google Cloud, that means seizing the opportunity to use data science in new ways, through access to sophisticated cloud analytics tools, including machine learning.
This technology is fundamentally changing the way we view and interact with data. It’s had an enormous impact on how we:
- Identify new customers by using data to reach prospects in strategic, scaled ways.
- Approach customers by highlighting products that solve specific customer pain points.
- Grow customers by helping them to take full advantage of products that enable their success.
- Optimize spend by investing efficiently and tracking intelligently.
Machine learning, data science, and predictive analytics are the new, increasingly crucial complements to traditional marketing best practices. But they may require some new talent on the team, including people with specialized knowledge of analytics and experience working with engineers who have built machine learning models. In addition, bringing science into your marketing process might require a mindset shift.
Here are three lessons I’ve learned about blending science and marketing:
Live, breathe, and eat data integrity
Capturing and managing data in a consistent manner throughout the organization sounds simple. But it can quickly become overwhelming, especially when you’re working across different countries and regions. One team may input data that’s incompatible with another team’s data, making it difficult to analyze both sets in consistent ways. By focusing on data integrity in how you collect, manage, and store information, you’ll avoid spending time figuring out how to compare apples to oranges.
At Google Cloud, we’re standardizing our processes for integrating new data and systems, and investing time in making our tools scalable for the future. This also involves working with our sales team to ensure our data sets are aligned. Ultimately, starting with clean and accurate data has allowed us to get to insights faster.
Establish an experimental mindset
A critical part of our mindset shift with using new technologies includes fully embracing testing and experimentation. We put this approach to work as part of our efforts to promote G Suite through paid advertising. We rely on paid advertising to raise awareness, which means we’re always thinking about ROI and testing which channels work best. Given the huge volume of performance data and the need for real-time optimization, cloud-based data analytics and machine learning have been essential to our success. These tools enabled us to test hypotheses and make targeting decisions at the campaign level rather than the channel level—something we couldn’t have accomplished manually.
A critical part of our mindset shift with using new technologies includes fully embracing testing and experimentation.
For example, take the challenge of knowing who, of the people who trial a product, will convert to paid users. Typically you would need to wait weeks for that trial to conclude in order to understand which campaign was successful. With advanced analytics and machine learning, we’re able to know the likelihood of a channel bringing in the right people to trial a product in two days, significantly improving ROI.
Select specific business problems to solve
Cloud tools help you make better decisions by uncovering relevant information at a level of granularity that wasn’t possible before. However, this only works if you approach business problems incrementally—breaking your business needs into smaller, discrete questions. Rather than trying to drive more overall traffic to your website, for instance, focus on driving repeat visits within a particularly valuable customer segment.
In our case, success depends on selling our products. But rather than targeting all potential buyers, we decided to focus on a specific segment: existing customers who were likely to want additional services. Applying Google Cloud Platform technology and open-source cloud tools—BigQuery and TensorFlow chief among them—to our customer’s product usage data enabled us to improve product recommendation relevance. It’s only one piece of our marketing efforts, but by focusing on a small, particular business problem, we achieved results that made our campaigns more effective.
So, are cloud computing, data analytics, and machine learning the new holy grail for marketing? Not quite. The data that comes out of these advanced models is just one piece of the equation. The real magic comes from combining deep customer insights with thoughtful creative and messaging.
It’s in this combination of creative thinking and advanced analytics—the marketing and the science—where I see the next wave of marketing innovation.
To learn more about using big data analytics and cloud technology in your organization, download my presentation, “Marketing Meets Science.”