The marketer’s three-minute guide to data science

19 Feb, 2018

Categories: Data & Analytics Data Science

3 Minute Read

What is data science?

Data science is the practice of revealing hidden insight from existing data in a manner that enables businesses to make better decisions. In this article we give you a quick three-minute overview of why it’s important and relevant for marketers.

The modern consumer is hard to impress. An average person sees between 300 and 500 ad messages per day but only a small percentage of those ads “hit the target” and encourage that consumer to take action, rather than click away in dismay. Even the wittiest slogans fall flat when the offer is irrelevant at that moment. For example, I’m definitely not “lovin’ it” when reading advice about losing weight.

To influence the buying behaviour, marketers need to second-guess the customer’s next move and deliver a targeted offer at the right time, and at the right place in their journey. For that purpose, most businesses now accumulate big data – all the product preferences; past brand interactions; demographics data and so on. No business is now too small for accumulating that kind of data. Yet transforming it into meaningful action remains a common struggle. This is where data science steps in.

What data science can do for marketing

Intelligent decisions are based on accurate predictions. You don’t go and dig a hole in a random spot and expect to find oil. You study the geological data and try to predict where you have the highest chance of locating an oil trap. Data science does a similar thing in marketing. It helps you pinpoint the most profitable actions based on the data you already have. For instance, you can estimate that emailing a free shipping code to consumer group B will generate orders with an average price of £55.57, while consumer group A will likely spend under £30, as that’s how they usually roll.

Predicting better marketing outcomes isn’t the only attractive thing achievable with data science. The most promising applications of data science in marketing stretch across multiple channels.

Micro-targeting

With the help of statistical analysis, you can pin down highly specific customer groups and create more impactful offers for them. A simple example would be sending promo codes for anti-frizz products to women aged between 18-26, who previously bought similar products for curly hair, and are located in the Manchester area, where heavy downpours are expected during the month.

Automation at scale

As big data became big, a lot of marketers struggled to keep an eye on dozens of changing metrics; gathering timely reports from analytics tools and spending too much time going through all the data, rather than taking action. Analysis paralysis is another issue hampering our productivity.

So let’s outsource these chores to the more qualified executors – the algorithms. Data science enables you to automate the common routine process and streamline repetitive tasks. Some quick examples would be:

Automatically pausing all advertising campaigns spending more than £350 that haven’t converted in the past 30 days.
Launching a new social media advertising campaign in a certain region whenever condition X is met. For instance, promoting your hotel properties near the airports whenever you receive data about cancelled flights.
Scoring and grouping incoming leads based on certain parameters and forwarding the data to your sales department with a proposed action plan.

Optimised performance and budgets

Most customer journeys are no longer linear. At first glance, they may look like a tangled mess – a user hops to a third party website to compare prices, looks for discounts, or abandons the basket only to return later in the day and purchase a completely different product.

Data science enables you to understand all of these touch points and refine your message and actions accordingly. You no longer assume that it’s just male consumers aged between 35-45 who want to buy a black suit from you, because your CRM tells you that in fact most customers do.

Analytics-led marketing means that you also take additional factors into account such as the person’s past and current searches on your website, clicks, time spent on certain pages, shares, saves, likes, downloads and a number of other interactions. This data is then churned by the algorithm and transformed into a product recommendation the customer is most likely to respond to. In 2017, predictive intelligence recommendations delivered a 22.66% in conversions rate – a hefty number for a business’s bottom line.

The results of data-driven marketing progressively improve over time as more data is being collected and analysed. This new data can then be applied to tweak the campaign performance even further or reduce spending on low-ROI activities that you can spot at one glance.