Today’s marketers face new pressures to prove the value of their work. With the advent of big data and advanced marketing analytics tools, there’s no excuse for failing to demonstrate how your work is tied to tangible bottom line business growth.
But forward-thinking marketers are already realizing the need to move beyond ROI and into the even more complex territory of marketing predictability. They’re focused not only on demonstrating how their marketing efforts are driving business growth today, but also on predicting the impact their campaigns will have in the future. In fact, 66 percent of the global CMOs surveyed in IBM’s 2015 CMO Study said they plan to increase their use of predictive analytics in the year ahead.
This is no easy task. Developing and acting on predictive analytics requires a high level of data proficiency, and 71 percent of those same CMOs say they’re underprepared to deal with the “data explosion” that has taken place in the last few years.
In order to avoid becoming overwhelmed by the complexity of predicting the future through data, use these three straightforward steps to improve your marketing predictability efforts.
66 percent of the global CMOs surveyed in IBM’s 2015 CMO Study said they plan to increase their use of predictive analytics in the year ahead.
Assess past performance
Historical data from past campaigns must serve as the foundation for every marketing prediction. If you’ve been utilizing a CRM and a marketing automation tool, you should have historical data about the source of each of your recent leads and the tactics that nurtured them down the sales funnel. By assessing which tactics generated the greatest quantity of leads, which tactics impacted the conversion of those leads into customers and how much you invested to make each sale, you should be able to gain a relatively clear understanding of the potential impact of the various components of your marketing strategy. Based on data from past campaigns, develop an estimate of the number and types of leads your current campaigns will bring into the funnel, compare against strategic plans and goals and optimize your marketing mix accordingly.
Develop scoring models to predict buyer behavior
As you bring your marketing mix to life, how will your leads and prospects react, and what does their behavior tell you about what you can expect in the future? This is the next question you must ask yourself in your effort to make marketing more predictable. In order to begin predicting which leads and prospects will react in what way, it’s essential to develop scoring models that will help you separate your strongest leads from the weaker ones and identify the criteria that makes them most likely to buy.
What content does a prospect find most relevant? How might they have answered specific questions on specific forms? What role do they play at their company?
Again, historical data can help you do this. Analyze data from your marketing and CRM systems to identify behavior point patterns that indicate intent to buy. What content does a prospect find most relevant? How might they have answered specific questions on specific forms? What role do they play at their company? These insights can inform the development of a scoring model that attributes numerical scores to each action a lead takes based on your assessment of its potential to impact a future sale.
For example, you might observe that 60 percent of your recently closed leads visited the pricing page on your website, and 25 percent downloaded at least one of your case studies. Because of this, you’d assign a high number of points to those who visited pricing, and a not-quite-so high number to those who downloaded a case study. Conversely, you may notice that people who visited your careers page are (unsurprisingly) very unlikely to become customers: these people might get assigned negative points and be filtered out of the funnel so they never make it to your sales team.
Top of the funnel actions like visiting your homepage may receive a single point, while actions that typically take place further down the sales funnel, like downloading an in-depth white paper, viewing a case study or completing a form, might receive 10 points. Based on this scoring model, you can begin to determine which leads are most likely to buy and predict future sales weeks or even months in advance. With this created, you can identify opportunities not just for further marketing outreach, but for the sales team to reach in and help lead a particularly engaged prospect over the finish line.
You don’t have to be a data scientist to develop a lead scoring model, but enlisting one can certainly help. Consider working with a data consultant or analytics specialist, like the ones on our team at Movéo, when you’re ready to take your lead scoring beyond the basics. They’re the ones who can help you wrangle the many complex variables that typically come into play and improve the accuracy of your scoring models.
Measure, analyze and improve
Improving marketing predictability through data requires multiple trials (and often many errors) to determine what works. Be sure you’ve set your marketing and CRM systems up to deliver as much real-time data as you possibly can, and then analyze it carefully.
Data is no longer an optional add-on for your campaign — it’s a given. To demonstrate your ROI as a marketer and track your results, set every campaign strategy to include KPIs, measurements and metrics, and continually refer to these metrics to determine whether or not you’re moving in the right direction.
Ensure you’re tracking both tactics and strategy; keeping account of whether your tactics are working, whether your models are correctly calibrated and the success of your targeting is all essential to understanding the current and future actions of your campaign. Are the tactics you’ve employed in the past performing as you predicted they would? Is it time to modify your predictions or work new tactics into the mix to bring more leads into the funnel? Are your lead scoring methods accurately predicting a lead’s likelihood to close? Or do you need to modify your models based on new data and insights? These are some of the many questions we need to constantly ask ourselves in what must be a never-ending quest to optimize both the predictability and the performance of our campaigns.
Rather than making big, sweeping shifts to your campaigns and processes, try tweaking just one thing at a time, and testing that change during a period of focused analysis. As you become more comfortable predicting campaign performance and analyzing it in real time, this will begin to feel more like a science than an art. Stay at it long enough, keep making improvements, and soon you’ll be able to determine the impact of nearly every element of your marketing program with a fair level of accuracy. And isn’t that every marketer’s dream?
Movéo is a fully integrated communications company uniquely built to help its clients measurably improve marketing performance. We use data-driven insights to inform strategy and guide our use of creativity and technology. This results in integrated communications solutions that attract, secure and retain profitable customers for our clients. For more information on Movéo, visit moveo.com