When I first started working as a data analyst in marketing campaigns, I found myself explaining the difference between data analytics and data science on a weekly basis.
At that time, around a decade ago now (how time flies!), the use of “data” in marketing was still something of a black box: managers had an idea that they could achieve wonders by “working with data”, but had no idea what they were collecting, what it could be used for or the inherent limitations of this approach.
Nonetheless, I still get marketers walking up to my desk, and asking me to perform miracles “with the data”. If they want to know the average age of our customers, that’s fine. If they want “the AI” to “design some content”, I sigh and gently explain that data science is not magic.
In this article, I’ll run through that explanation for the benefit of everyone.
Data Analytics and Data Science
The first, and arguably the most important, point to make is this: that there are major inherent differences between data “analytics” and data “science”.
Data science is an incredibly broad term that essentially encompasses everything we can do with data, whether in marketing or any other field. It includes the techniques we use to collect data, as well as methodologies for working with these. At its core, data science is driven by scientific and statistical techniques that are used across the natural and computing sciences: regression analysis, hypothesis testing, and statistical significance testing. The term is so broad that even undergraduate courses in data management (like those I have completed) only touch on a narrow aspect of data science.
When most marketers talk about “using the data”, they mean using data analytics. In data analytics, we generally have a specific goal in mind. The objective here is not to push the limits of human knowledge, but to answer specific questions. Though data analytics also makes use of complex tools and techniques, these are designed (from the beginning) to reach narrowly defined objectives.
Even in this short explanation, you can see why so many marketers get confused when it comes to data. Though data science can achieve wonders, most organizations will need a more pragmatic approach to data analytics, in which they define their goals clearly. Let’s take a look at how to do that.
Data Analytics: Informing Your Strategy
As I’ve said, when most of my managers ask me to “work with the data”, what they mean is to perform some aspect of data analytics. Data analytics is essentially a retro-active tool that allows organizations to analyze past activity. It is extremely useful for marketers in (at least) four main areas:
- Data analytics can help to improve the quality of sales leads. It allows you to identify which types of customers are generating the most revenue for your organization so that you can target your lead generation on similar customers. High-quality sales prospecting tools take much of the grunt work out of this process, and the best can collect highly detailed information on your customers.
- Second, data analytics allows you to analyze the performance of your own staff. This allows you to identify which of your sales teams are generating the most profit so that you can encourage them to share their techniques with their peers.
- Third, data analytics is a powerful technique when it comes to customer retention. By building a database on the types of products that particular types of customers buy, and by incorporating this intelligence into your email marketing, you can ensure that existing customers receive information about products that are genuine of interest to them.
- Finally, data analytics can be used to find the right price point for products. Instead of setting the prices of products speculatively, analytics systems can provide you with detailed information on what customers are willing to pay for each of your products.
All of these applications of data analytics, as you can see, are essentially backward-looking. They take existing data on the way in which your organization operates, and use these to inform your decision making. Data science, as we will see in the next section, takes the opposite approach.
Data Science: Looking to the Future
When it comes to marketing, data science has one huge advantage over data analytics: it can allow you to look into the future. Predictive models – often driven by AI – can be used to assess changing market demand for particular products, and allow you to prepare for changes before they happen. Ultimately, data science can save you money by reducing waste, taking advantage of upcoming events, and targeting the right people at the right time.
That said, marketers should be aware that predictive models of this type are still in their infancy. Every course on data science training will stress that even the most advanced models are not yet able to predict customer demand and that unforeseen events (think a global pandemic!) will quickly render these models obsolete. For this reason, the predictions of data science should always be treated with caution.
Garbage in, Garbage out
As I’ve explained, there are significant differences between these approaches. However, in my experience, there is one factor that unites them. Or, more specifically, there is one mistake that typically undermines the efficacy of both approaches: not collecting the correct data.
To return to my point above, when a manager asks me to “predict customer demand” for a particular product, the difficulty with this is not (in most cases) that I can’t perform a quick regression analysis. It’s that I simply don’t have access to the relevant data, because we are not collecting it.
I understand, of course, that organizations are wary of collecting data just because they can. This is because several popular backup services such as IDrive or Backblaze only offer you around 5GB of space with the free versions, and therefore can be overwhelmed quickly. In addition, holding vast amounts of customer data is a security risk. Nevertheless, here is my plea to marketing managers: think about the kind of predictions and analysis you would like BEFORE designing a data acquisition system.
In other words, even though data analytics is largely a retro-active process, this doesn’t mean that you don’t need to think about it at the earliest stages of a marketing campaign.
Integrating data collection into marketing at the earliest possible stage will mean that the information that your data analysts are working with will be sufficiently complete – and sufficiently reliable – to perform the analysis you need.
The Bottom Line
Ultimately, developing an effective sales process requires an understanding of the data you need to collect, and what can be done with these. If in doubt, please talk to your data analysts, and get their input on which data they need to work with.
You might find, as part of this process, that your data analysts take the opportunity to lecture you on the differences between data analytics and data science. But now you’ve read this article, you won’t need that lecture.
Guest Post by Nahla Davies
Nahla Davies is a software developer and tech writer. Before devoting her work full time to technical writing, she managed—among other intriguing things—to serve as a lead programmer at an Inc. 5,000 experiential branding organization whose clients include Samsung, Time Warner, Netflix, and Sony. You can find her on LinkedIn.