We’re No Longer Simply “Data Consumers”Gone are the days when we wanted data for the sole purpose of Marketing outreach or email campaigns. In the early stages of data collection, we, as marketers, soaked it up. Data, data, data. “How can we get more data?” Marketing teams were enthralled by the ability to capture data at such magnitude. And once that data was collected, it was a fairly straightforward process – load the data into your systems, send an email campaign, capture more data from said email campaign … rinse, repeat. But today, the purpose of data collection has evolved. Marketers are shifting from a data collection approach to a true data strategy, where the main goal is bigger than just having data to work with. Today, marketers care more about being able to actually read, analyze, and understand their data, in a way that allows them to gain insight into how their customers are behaving, what they are looking for, and when (which, sometimes, is before they even raise their hand – hello, Intent Data).
The Market is Focusing More and More on Advanced Data ManagementSounds easy, right? Unfortunately, it’s not. With the mass amounts of data sources available to marketers these days, data management and standardization can sometimes be a real challenge. Especially when people have the option to write in their own values when filling out forms, such as Job Title or Company Name. To begin tackling this challenge at Cloudera, Sara and the team recently began an initiative to apply Convolutional Neural Network (CNN) to their data in order to standardize the Job Title field in their database. According to Sara, it took essentially their entire Marketing and Data Science teams working together to determine where to start and how to develop the model. Sara touches on just how important data standardization is, pointing out, “it’s an important field to understand how we want to talk to that person. We don’t want to talk to a developer with the same message as a CIO. It’s just not going to resonate as much.” Sara explains further how Cloudera’s Marketing team is focusing their resources on not just consuming the raw data, but actually fashioning it in a way that can be meaningfully understood across the organization. “You have people like, ‘Chief Awesome Officer’. Well, what does that mean? How do we figure out how to manage all those people and what messages they should get? … How do we create a model that will intelligently, using natural language processing, look at the data that has been input and then not only standardize it, but then also create meaningful information for our Demand Generation team?” The Cloudera’s Demand Generation team uses Levels (Platinum, Gold, and Silver) to determine qualified and prioritized prospect accounts. Their natural language processing engine has been designed to read an input job title, standardize it with other related titles in their database (e.g. CIO standardized to Chief Information Officer), and convert that into a Level that the Demand Generation team then uses to prioritize that lead/account. Pretty snazzy, right? Standardization of data, and more importantly, meaningful recommendations from said data, is becoming increasingly critical to successful Marketing strategies. It all comes down to one thing, really: without being able to understand your data and pull useful recommendations from it, what’s the point of having it? It doesn’t happen overnight, and the amount of work required is nothing to scoff at. It’s easy to get overwhelmed when trying to tackle advanced data management, but start small, and be patient. Sara tells us from experience that all you really need to do is start with an idea. Figure out what you need to learn in order to drive the business forward and what data will enable you to gain that knowledge, and go from there.
Lead Scoring is No Longer the “Be All End All” for Propensity to BuyIt’s no secret that one of the first ways marketers began using data to understand their customers’ behaviors was through lead scoring models. And while lead scoring was certainly, in a way, the first iteration of Machine Learning, and has provided helpful insight into which leads are valuable to a Marketing or Sales team, it’s no longer the most useful way to truly determine a prospect or customer’s value or propensity to buy. Cloudera doesn’t use a lead scoring model, but instead uses patterns amongst accounts who have converted in the past and firmographic, business, and intent data to build models that determine the likelihood of new prospects to convert. They also created a model for existing customers, using leading indicators such as opening support tickets or having support conversations (among other data points) to identify how likely a customer is to expand or find further use cases with Cloudera. “It’s actually kind of a funny thing … because we feel like scoring is, we’re just pulling out a number, right? How do we decide for the customer what is meaningful?” says Sara, and it’s a valid point. Buying cycles aren’t limited to one specific individual or one specific touchpoint – more often than not, the cycle is complex and involves multiple contributors, decision makers, and purchase authorities. This requires marketers to look deeper than just one lead’s demographic or behavior score, at data points such as engagement scores and intent data to understand how customers engage along their journey and which touch points are crucial to conversion.
Image: Buyer journey data and tech example from CaliberMindSara has observed a “huge lift in terms of more direct, effective spending on target accounts” after the business made a strategic discovery roughly a year ago that they were spending “a lot” of money on prospects that didn’t fit their Ideal Customer Profile (ICP). Using their account data and scoring model, they developed strategies for new customer acquisition and customer expansion, and shifted their marketing spend to focus on those accounts most likely to convert. Following the adoption of a data and machine learning-backed target account strategy, they’ve reached 86% of their target account list, with 77.5% of the list engaging with the website. Company-wide results of the effort include a 23% increase in total revenue year-over-year, 26% year-over-year growth in subscription software, and an additional 30 customers spending $100,00+ per year gained in one quarter. Armed with this type of information, marketing and sales teams can confidently focus their efforts on the accounts, channels, and activities that are predicted to deliver the strongest ROI for the business, as well as provide a better experience for prospects and customers. And, luckily for us marketers, we don’t have to be technical or eyeballs-deep in complicated data to do so. “There are definitely tools out there that can help you, even if you aren’t a data scientist yourself or you have limited in-house resources.”