Share this content
Small data

Why the next era in customer engagement will be defined by small and wide data - and the right mindset


Big data was big news. But organisations are now shifting their focus to small and wide data. So what are they, and why are they so important to customer engagement?

16th Aug 2021
Share this content

In a recent article, I wrote that a smart way to go about business model design is to view it through the eyes of the customers because they are the best equipped to talk about their needs and wants. In turn, analysing those needs and wants is one of the very first steps in providing a better customer experience.

This is where the adjacent concept of “customer engagement” comes into picture as an all-encompassing concept to reflect the full set of activities through which organisations can build direct relationships with their customers.

AI and ML for enhanced customer engagement?

Customer engagement is the emotional connection between a customer and a brand and, as such, it is focused on understanding and managing all the touchpoints with the customer, including advertising and marketing, sales transactions and post-sales customer service, among others.

Customers that are highly engaged buy more and then buy more often, promote more, refer to others more, and show more loyalty. In building an effective customer engagement strategy, organisations must remember that providing a high-quality customer experience is critical. Understanding customers and their behaviours in this process is a must.

Advancements in artificial intelligence (AI) and machine learning (ML) have facilitated this better understanding of the customer, strengthening the decision-making capabilities of organisations.

AI and ML are able to analyse large amounts of data in a very short amount of time. Moreover, they have an excellent predictive power that can produce meaningful and actionable insights that can guide the next interactions between a customer and a brand. But is this all there is to the story?

Remember the David and Goliath story?

AI and ML have been slowly, but steadily thriving on the idea that the more the data, the better the analysis performed. This view has been facilitated by the emergence of big data, which has been hailed as the holy grail of modern decision-making. As a result, AI and ML have become overly reliant on data hungry approaches.

However, most companies are still limited to small data. Not only that, but today, many large organisations have actually started shifting from big data to small and wide data, in an attempt to use the available data more effectively to extract more value and derive more meaningful analytics.

What are small and wide data? Small data are data in a volume and format that make them accessible, informative, and actionable. Wide data, on the other hand, are data that are tied together from a variety of small and large, structured and unstructured data sources; wide data, in this sense, are more about complexity and depth.

According to Gartner, 70% of organisations will actually shift they focus from big to small and wide data analytics by the year 2025. This is even more the case if we consider that many AI and ML models suffer because of disruptions such as the one generated by the COVID-19 pandemic, which has caused historical data to quickly become obsolete.

It is in this context, then, that we are witnessing the dawn of a new data era, one powered by AI and ML on small and wide data. The good news is that this era is auspicious for all types of organisations.

Let’s rephrase that! An organisation does not need to have mountains of data to leverage the power of AI and ML analytics for better customer engagement. As the famous Bible story goes, even a small man can defeat a giant. Why, then, couldn’t small and wide data contain the potential for solving big problems to eventually generate massive impact?

How to maximise value with small and wide data

So, yes, massive progress can be made with small and wide data. Because while big data are about finding associations and correlations in enormous amounts of data, small and wide data are about extracting more context and establishing causation, for more insightful business results. In other words, it is about depth of analysis over quantity of data.

For example, research has shown that marketing is one of the areas that benefit the most from small and wide data analytics. Marketers can extract intelligence from small and wide data in order to map in a more accurate manner the people with the highest propensity to become customers and highly personalise messages.

In his book titled “Small Data: The Tiny Clues That Uncover Huge Trends”, Martin Lindstrom writes that “if one takes the top 100 biggest innovations of our time, perhaps around 60% to 65% are really based on small data”. What a wonderful insight!

Bigger data are not always better data. Depending on application, smaller data may actually be more appropriate for intensive, in-depth examination to identify patterns and phenomena, and as mentioned, establish causation. It all depends on what the end goal is.

This brings us to the second point of this article, which is that before embracing AI and ML tools and analytics, organisations must adopt a problem-centric thinking and ask themselves: “What is the problem that I am trying to solve and what added value am I hoping to achieve”?

Answering these questions will help clarify the data analytics strategy to be embraced, including the types of data analytics and the type of data needed.

A problem-centric approach puts the customer first!

Do not poke around the data! That is amateurish. You will end up identifying correlations that simply do not make sense. Take, for example, the notable work by Leinweber (2007), who demonstrated that data mining techniques could show a strong but silly, spurious correlation between the changes in the S&P 500 stock index and butter production in Bangladesh. You do not want that!

Also, just because you can show that two things correlate, it does not necessarily mean that one causes the other. So, yes, there is a very real danger of getting yourself entangled in muddy waters.

Independent of the size of the data and the type of data analytics strategy adopted, interpretation remains at the heart of great data analytics. And that interpretation is context dependent. In other words, if you want the correct insights, then understand the problem you are trying to solve first. That frames great big, small, and wide data analytics!

In conclusion…

While organisations can certainly benefit from big data, extracting knowledge from such data does not discount the knowledge that can be derived from the small and wide data.

The smartest strategy, therefore, is to understand when and how to combine and integrate big data with small and wide data analytics and when to employ each on their own.

And if you truly want to revolutionise customer engagement, and thus, experience, then remember to adopt a problem-centric approach. Do it right and the gains will soon become obvious.


Replies (2)

Please login or register to join the discussion.

Sampson Lee, founder of Global CEM and creator of PIG Strategy
By Sampson Lee
17th Aug 2021 03:13

Hi Charles: Thank you for your article and I agree with you that “Do not poke around the data! That is amateurish. You will end up identifying correlations that simply do not make sense” and “if you want the correct insights, then understand the problem you are trying to solve first. That frames great big, small, and wide data analytics!”

However, the security of data storage is so fragile. For example, 3 billion Yahoo accounts leaked in 2013, 700 million LinkedIn users leaked in 2021, and 533 million Facebook users leaked in 2019. All the victims received were only apologies from these digital giants.

Even if the data stored in the world's most secure technology giant may be stolen, how can customers ensure that their data is safe?

In 2006, Clive Humby the famous UK mathematician coined the phrase “Data is the new oil.” Business people now generally use the phrase "data is the new currency."

There is no doubt that data has its value, especially customer data. When a customer allows the company to use his/her data, the ownership of the data is still the customer, not the company.

When a company leaks customer data, should the company directly compensate the customer instead of paying a fine to any other organization or entity? Considering that the customer data of so many large companies and digital giants has been leaked, should a mechanism/pricing structure be established to directly compensate for the loss of customer data?

If there is no compensation to the customer, how can the customer comfortably let any company do any personalized things for them? This is unfair and unjust. Remember: customers own their data and the data has value, not to mention the huge value of big, small and wide data for personalization.

I would love to hear your thoughts.

Thanks (1)
Steven Walden
By Steven Walden
19th Aug 2021 19:46

Interesting article. I would also extend to better consideration of how we embed cognitive science in understanding customers.

Thanks (1)