Guide to a Data-Driven Ecommerce Digital Marketing Strategy

Written by on 10th August 2018

Ecommerce businesses are uniquely positioned to exploit the benefits of the rich and fluid market ecosystem created by the Internet, but they also often have to deal with the challenges of competing for customers on a national or continental level.

It is true that having a wider range of potential influence than a brick-and-mortar store or a local services provider opens you up to opportunities they don’t have. However, it also means that you have to be able to respond to a much more varied list of customer demands, observe and study a much broader context than they have to, and deal with a number of other challenges unique to your situation.

There is no way to effectively work or promote your work in a system with as many connected threads and as frequent changes as ecommerce is characterized by, without constant, purpose-driven data diving.

Some of our recommendations for ways to improve your marketing efforts are not actionable without advanced Big data mining capabilities and a bit of help from AI, but most of them can be implemented with just a couple of utilities, some persistence and the right focus.

Customer insights

Getting new customers hinges on you learning from the ones you already have. Not to mention that keeping a customer can be much more important than getting a new one.

Audience segmentation is doing just that, allowing you to group people that might be interested in your offer according to different criteria, and then learn the specifics about each group. You note how they want to be addressed, which social networks guide their purchasing decisions to which extent, what they like or dislike about you, etc. While this is often reduced just to creating a Buyer Persona if you are a B2C business or an Ideal Client Profile if you’re B2B, you don’t stop there when getting to know your audience.

You also take the time to observe the behavior of these groups and individuals on your site. Where are they dropping off? Is there a stage in your funnel that could be optimized to perform even better than it already is? Are people having a hard time completing a certain process? Is there a common complaint that could be easily addressed but still significantly improve customer satisfaction?

The point is, you should gather every bit of customer info that you can, whether it’s coming from Google Analytics, direct feedback, or AI that has analyzed petabytes of data; and you should keep modifying your marketing (or even your general business) strategy according to what you learn through continuous data analysis.

Competitor and market info

Customers may be an important part of a market, but they are not all of it. While it would be nice to forget about competitors, price averages, sudden supply issues or the influence of delivery methods on a popularity of a business, doing so is not an option if you want to stay in business.

A detailed competitor analysis involves taking a look at their:

  • Backlink portfolio
  • Social networks
  • Content types and performance
  • Branding and tone of voice
  • Targeted audiences
  • Pricing

Apart from informing your SEO, PPC and overall digital marketing strategy, observing the behavior of your competitors also tells you a lot about how they perceive the market and the audience you are trying to share. In other words, you don’t have to have data mining capabilities, you just need a competitor who does and whose actions you can observe to try and determine what conclusions they reached by Big data analysis.

Testing and Optimization

You think you know your customers, but your ads are simply not leading to purchases? Your landing pages have always performed well, but the ones for your new product are simply not boosting conversions? Despite that fact that marketing cannot be successful without exact data and a systematic approach to it, the effectiveness of your promotional efforts will sometimes depend on the most tenuous and irrational quirks of the human psyche.

Even the slightest change in the copy of your PPC ad, or in the size or color nuance or your call to action button can result in dramatically different response rates. By subtly tweaking these and other elements of your marketing strategy (email subjects or body, social network posts, content on your site…) and testing the effectiveness of each of their iterations, you are ensuring that you are getting the best possible ROI from all of them, and that you are heading in the right direction.

Predictive analysis

While it is usually reserved for those with a serious Big data mining infrastructure and personnel, the potential of this particular branch of research is amazing. While testing allows you to make immediate changes based on current data, predictive analytics, ideally, allows you to anticipate future trends or issues by digging through data from the past.

It could rely on external data, like tracking market fluctuations and anticipating their effect on your business; or on your internal data, allowing you to find patterns in your past performance, along with the outcomes they lead to. If you cannot afford to handle this kind of processing internally, there is no shortage of versatile digital marketing agencies who can analyze different market segments for you, but you need to remember that, coming from the inside or the outside of your company, this information cannot always be exact, and is intended more to guide your course than to dictate it.


Finding a way to satisfy the many needs of many types of potential customers, all of whom are being pursued by many competitors is not something you can hope to do without dedicating many hours to getting familiar with all of them.

While most of the even remotely relevant data you could obtain might stand to benefit you in some way, the most important bits of information you need to focus on are the ones pertaining to you customers, their habits and preferences; your competitors and their strategies; and the effectiveness of your current marketing efforts. When you also add a bit of clairvoyance that comes with predictive analytics, the whole data analysis thing almost starts looking like an opportunity instead of a chore, so be sure to treat it that way.