Big Data -vs- Small Data: What Big Data Means for Small E-Commerce Businesses

In the past decade, Big Data has gone from being little more than a buzzword to something that has very tangible real-world uses across many industries. With access to vast amounts of data – and the means to actually use it – the big players in eCommerce now appear to have a considerable advantage over their small to medium-sized competitors. What can be done to compete with such resources?  

Big Data Has Transformed the Digital Landscape, Not the World

It’s important to remember that this type of problem is as old as the free market. Large and established businesses have always had certain competitive advantages due to their size and resources, but small enterprises have been able to thrive nonetheless. Small market participants will always have an edge in other areas, such as the opportunity to be agile and inventive. Big Data hasn’t changed any of this.

What is Big Data Anyway?

But first let’s take a step back and try to sort out what Big Data really is and how it’s used within the frame of eCommerce. After all, it’s a rather convoluted topic that spans many different fields of IT and comes with a similarly varied set of definitions. The concept obviously refers to managing large amounts of data, but that is still a very vague description.

In eCommerce, it is mainly about collecting large amounts of customer/demographic data, analyzing it and – most importantly – converting the data into valuable marketing and business intelligence. Due to the size and complexity of the data, the processes must be automated.

However, there is no clearly defined boundary where a certain data set suddenly becomes “Big” – this is open to interpretation. So essentially any amount of data that is sufficiently large to require (also not clearly defined) sophisticated data analytics methods to extract value from it could be considered Big Data.

This is not to say that the size of the data sets is not important. As with everything related to statistics and analytics, a large sample size yields more accurate results than a small one (everything else being equal). With sufficient data sets and the means to process them correctly, Big Data has many benefits in eCommerce.

Some of the Key Benefits of Big Data

Many of the major big-data benefits are derived from predictive analytics. It can, assisted by machine learning, identify patterns based on historical data. For example, it might predict seasonal sales volumes to help you optimise pricing and inventory levels. It could also find patterns in previous customer activity to improve cross-sales, customer retention, and much more. Other, possibly related benefits include:

• Personalised shopping – Using snippets of information about each of your customers, such as their site-search history, shopping patterns and browsing habits, you can tailor their shopping experience with personal site-search results, product recommendations, upsells or targeted ads.
• Reduced shopping cart abandonment – The shopping cart abandonment rate is a vital metric for all eCommerce sites. With big-data-methods such as retargeting (potentially with cross-device tracking) and automated e-mail reminders, it can be reduced.
• Data-driven decisions – Access to large amounts of business data will help you take informed decisions, based on facts rather than educated guesswork.
• Informed customer service – With every conceivable bit of information on each and every customer available at a glance, you could (at least in theory) greatly enhance your customer service and make it much more effective in the process.

The above examples really only scratch the surface of what can be done, and is being done, using the massive amounts of information harvested by retailers such as Amazon and eBay.

Similar techniques are of course used by a wide variety of eCommerce businesses, albeit on a lesser scale than the industry’s giants. But although there are certainly lessons to be learned from data crunchers like Amazon, there are some things that Big Data can’t do or provide.

The Limitations of Big Data

• Business intuition – Having data to back up your business decisions is great, but data alone will make you aware of all fundamental factors that may affect your business, such as politics and other human concerns.

• Predict the future (with certainty) – Predictive analytics offers a comprehensive and useful toolset for finding patterns based on previous behavior, i.e. existing data, which doesn’t account for outliers. 

• Find new, creative solutions – Anything resembling creativity in a machine is at best a reflection of its human designer. A Big Data application will only find what it’s programmed to find.

• Great customer service – Only real, helpful people can inspire brand confidence. Going too far with automated responses to customer concerns may have a negative impact on the bottom line.

The list could go on, but it’s of course just pointing out the obvious: Big Data is very useful but not essential. If you run a large multi-national eCommerce business it is probably worthwhile to invest in proprietary technology that puts your data to good use.

Even for a startup, it’s a sensible strategy to start collecting as much information as possible early on, using economically viable off-the-shelf solutions. It will be useful eventually, but in the meantime, other factors will likely play a much bigger part.

Potential Strategies for Small and Medium-Sized Businesses

The apparent threat from Big Data – that it creates a huge divide between those who have it and those who don’t – isn’t necessarily that serious. While there are obvious advantages to having access to massive, business-related data sets (if you can extract value from them), a small enterprise usually has more immediate concerns, such as refining its business model and building brand awareness.

As a startup or a medium-sized eCommerce business you may want to start by simply learning from your main competitors’ Big Data strategies and adapt them to your own goals. What works for the big-name competition may not even be suitable for your business model, but perhaps there are other ways to identify and solve problems – using a Big Data approach or something else entirely. Some readily available solutions don’t require large data sets at all, but rely on Deep Learning to solve related problems.

For most businesses, scalable solutions are normally preferable to starting out with proprietary alternatives, which may require large initial investments as well as continuous, dedicated development resources. A scalable model also gives you the much-needed flexibility to grow at your own pace.

Contrary to Amazon, eBay and other huge eCommerce sites (with their own big-data programs in place), your business likely caters to a particular niche, which is also a key competitive advantage.

The largest retailers cater to a very large part of the market, and thanks to Big Data they are reasonably successful on the whole. However, their lack of specialization is also a weakness for niched retailers to exploit. Retailers with a very wide range of products can’t follow up on specific trends and opportunities within smaller subsets of goods.

Sourcing, merchandising and marketing niche products at the right time, using intuition and conventional analytics, will be a very difficult trick for broad-range retailers to imitate.