Getting closer to your customer: unlocking the value of behavioural data


The challenges of building strong digital behavioural data foundations

Digital behavioural data is essential for making informed decisions and improving user experience. Discover the six main hurdles organisations face when building strong data foundations and how to overcome them effectively

Digital behavioural data can be an extremely powerful tool for any modern business, enabling deep insights into what customers and potential customers want, including how they prefer to interact with your digital platforms. 

Access to this data opens up the ability to build new offerings, using personalisation and AI to curate best-in-class user journeys. But without strong digital behavioural data foundations, it can be difficult to implement these changes successfully. Here are the six main hurdles organisations face and how to overcome them. 

01 Data collection

Before any insights can be gained from behavioural data, it must first be gathered - but what exactly is it that should be collected? Traditional approaches to web analytics involve manually tagging different elements that the user is expected to interact with, but this approach can’t always give a complete picture. Not only can it become complex to maintain, as sites and apps change over time, but it limits what can be learned from the data. 

Data collected in this way will only inform based on the assumptions made on what would be important to track, limiting insights beyond our own current expectations. In reality, it may be the case that the most insightful actions users take on a site had not been thought of ahead of time. It is key that data collection solutions cover all user interactions, with a privacy-first approach, so that decisions can be made based on the full scope of user behaviour.

Businesses should promote a data-first perspective throughout the organisation
02 Data silos

Another challenge, especially for large organisations, is that often not only is a large amount of data being collected, but it is being collected in different places, by different teams, using different systems. So, when you want to use data to answer a question, it might not be clear where the data you want sits, who has responsibility for it, or even if the data you need exists. 

In order to avoid this, firms will need to centralise the data they already have, as well as establish processes - both cultural and technological - to ensure that as new data collection methods are added they are integrated into existing systems.

03 Data quality

It’s no surprise that AI investment is a key focus, but to succeed enterprise organisations need to remember that AI technology is only as good as the data you feed it. Every part of an organisation needs to be able to trust the data they receive from any other department is of high quality. For this to happen, there should be clear processes for assessing, cleaning and validating data as necessary, and these should be consistent across the business. Implementing data governance policies is essential for this, setting out organisation-wide rules and assigning clear lines of responsibility to ensure that all data collected is consistent, accurate and complete to make the most out of AI investments. 

04 Data literacy

For a firm to effectively make decisions using digital behavioural data, the employees need to have the skills to understand, analyse and apply that data. Data can and should inform almost every aspect of a business, so data literacy should not be contained within a single specialist team. 

Businesses should promote a data-first perspective throughout the organisation, as well as identify existing weaknesses such as where decisions are being taken without the data to back them up. A business that uses data successfully will work on building the confidence of all employees across all teams in working with data and data tools.

05 Organisation design

Does every company need a specialist data team? Does every team need a data expert? Should every employee be a data expert? There isn’t a single answer here as every business has its own needs, but what is key for leadership to understand is what their business needs to get out of data. Ensuring every employee at every level of the business has access to the data they need, and that the right tools and skills are in place to make that data actionable, should be non-negotiable. At an organisational level, this must go hand in hand with efforts to ensure data quality and literacy across the business.

06 Culture & leadership

Ultimately, the first foundation any organisation needs to build to get the most out of digital behavioural data is the business culture. Leadership needs to emphasise and encourage understanding of the importance this data can have so employees at every level can point to data to back up their decisions. Leadership should explain their own decisions in these terms, demonstrating by example the potential data can unlock when used in the right way.

Spend time understanding the fundamental foundations behind collecting and using digital behavioural data, and, if any of those are missing, assess what needs to change to get them in place. This simple step will pay dividends to any business looking to take advantage of the possibilities digital behavioural data offers. The risks of not doing so, and drawing the wrong conclusions about how customers are behaving, are countless.

The strategic power of behavioural data

In the digital age, companies have a vast amount of behavioural data at their fingertips. The challenge lies in cutting through the noise to know how best to optimise this data to create a competitive digital experience for customers

Commercial Feature

Optimising digital behavioural data in the age of AI

Advanced AI solutions are only as good as the data they rely on. Businesses must prioritise flexible, scalable data collection to gain visibility into customer journeys and unlock the full potential of AI-driven insights

AI is opening up countless opportunities for businesses to change how they engage with their customers, through automation, personalisation and tools like chatbots. Every organisation will be looking at which advances they can apply to their own business - but actually implementing them is not necessarily as easy as just plugging in the latest tool and hoping for the best. In fact, almost 3-in-5 businesses report that they feel behind the times in terms of making the most of what new technology can offer. 

The underlying issue to be tackled is that of behavioural data. AI is fundamentally powered by data, so in order for AI to provide meaningful insight on, and functionality for, their customers, organisations need to make sure that they’re collecting the right data. They also need to be collecting it in a way that means it can easily be consumed by LLMs and other AI solutions that can provide insights into what the data means and turn into something that can inform decision-making and power new products.

There is more data available about how consumers interact with digital platforms than ever before - but that doesn’t make getting insights out of that data straightforward. Andrew Fairbank, VP of EMEA at digital behavioural data analytics firm Fullstory, illustrates the challenge: “The volume of data created, captured, copied and consumed world-wide is more than 120 zettabytes - that's the equivalent of 67 billion iPhones. Harnessing the power of data, by making it actionable, has become a multi-billion dollar challenge - on average companies are spending half a million dollars on platforms to manage and mine their data for insights.”

With the right tools in place, AI-ready behavioural data can open up all sorts of possibilities: accurate modelling of user behaviour can help in forecasting and anticipating behaviour, patterns and pain points ahead of time to ensure that full advantage can be taken of opportunities and that firms can mitigate against any potential issues. Consumer experiences can be tailored and personalised based on individual behaviour, giving different kinds of support to a user displaying signals of frustration within their journey to re-engage them, or prevent them from disengaging in the first place. AI-ready behavioural data better enables LLM models resulting in timely, tailored, experiences that are revenue driving and cost efficient.  

AI-ready behavioural data can help in innumerable ways: financial services companies can use AI modelling to identify threats by looking for patterns of suspicious activity in behavioural data at the most detailed level - are certain patterns of clicks or user session lengths associated with bad actors attempting criminal behaviour? It can even help teams to diagnose technical issues - travel management firm Flight Centre used behavioural data to diagnose an issue in their booking process that led to a 22% decrease in failed bookings.

Understanding what AI-ready behavioural data has to offer is one thing, but having the right data to actually make any of it possible is another. One challenge is in collecting the data. For businesses engaging their customers through the web and other digital platforms - which is obviously, in 2024, most of them - it’s no longer as simple as sticking some analytics tags on your buy buttons, says Fairbank: “If you think about an average website, one page can have thousands of elements on it. And businesses don't have the ability to instrument every element on there, so you can have huge blind spots.”

“You're missing the information behind the clicks, in between the clicks as well because you're maybe tracking a dozen or maybe a little more elements. And our websites, our apps - they're not static beings, they're dynamic and constantly changing. So, you need a solution that is dynamic as well, that's capturing all that behavioural data and doing it in a scalable way.” This is a huge issue - according to a survey by digital behavioural data analytics firm Fullstory over two-thirds of businesses report struggling to retag their digital platforms when new questions arise so can’t easily ensure that they’re getting all the insights that consumer behaviour could be giving them.

There is more data available about how consumers interact with digital platforms than ever before - but that doesn’t make getting insights out of that data straightforward

What firms ultimately need is data collection solutions that are flexible, comprehensive and deliver data in a form that can be combined with transactional data and back-end data to get full visibility of every step of the customer journey, and that’s easily consumable by any AI tooling they want to build on top of it. The right behavioural data analytics is key to getting the right data, and the right data is key to making the most of evolving technologies. To make the most of new technologies, accurate data is essential, and accurate data comes from effective behavioural data analytics. 

Fairbank sums up what successful leaders need to understand about the relationship between data and AI: “It's hard to do AI well and leaders who are already doing it well have realised that advanced AI solutions are nothing without meaningful quality data. Because it's the fuel of the engine - you can't have AI without having the data to fuel that. That's the challenge that a lot of companies are having when they're starting on their AI journey - they're only looking at the end destination and they're missing the valuable insights and data they need to get there.”

The data is there - users are constantly giving signals as they interact with sites and apps - and the possibilities for what can be done with it are endless. Knowing how to action that data and realise those possibilities requires a comprehensive solution to ensure data is being collected and packaged in the most effective way.

Data literacy: why leaders must address the education gap

Data literacy is essential, yet many employees lack confidence in their skills. Business leaders must cultivate a data-driven culture within their organisation to ensure data-driven decision-making at every level

Most business leaders understand the importance of data. According to a 2023 DataCamp study, 78% of US leaders consider data literacy essential for day-to-day operations. But too often this has not translated to a data-literate workforce - a KPMG report from the same year found that only 21% of workers felt confident in their data skills, and only 34% felt able to make decisions based on data. If firms want to translate the theoretical possibilities that data can offer into practical benefits, their leadership needs to bridge this gap and ensure that all their employees who could be using data in their work have the skills to do so.

One solution to this is just to hire data-literate workers - a 2021 report by Forrester found that data had already become the most in-demand skill for entry-level positions, but practically speaking to tackle the issue at scale firms need to look at promoting a data-first culture in the workforce they already have. The question then becomes how to locate problem areas with using data in the business as it exists today.

The first step in identifying the parts of the business where data illiteracy is presenting challenges is to look for the symptoms that indicate teams are struggling or being let down by a poor understanding of their data and what they can do with it. Is it clear that data is informing day-to-day operations in all parts of the business? Are employees making use of the data tools available to them? Are they able to justify decisions based on data rather than intuition? Can managers explain the outputs of their processes, and make business cases using relevant numbers?

By asking these kinds of questions, not only can any necessary training be offered to aid individual data skills gaps, but management can look towards employees who are working in a data-first way already as evangelists of the approach, providing time and resources to share knowledge and success stories. Beyond this, the whole relationship of that part of the business has to data can be examined - if employees at the bottom end of the org chart aren’t using data appropriately, the issue might be as much with how management is communicating the importance of data to their work as it is an issue with skills as such. There needs to be an understanding that data is not just a tangential piece of admin, but a valuable asset that drives returns in and of itself when the time is spent getting the right insights to make the right decisions. 

The key is not just having skills to work with data, but in understanding the importance data has to all parts of the business, and if leaders don’t take the time to understand that themselves then they’ll struggle to build a data-literate culture underneath them. Recruiting people with, and training people in, data skills is not enough in itself- it has to become such a key part of the business culture that is taken for granted, so that every part of the business is motivated to retain and build on that knowledge rather than it becoming a box-ticking exercise.

Data-driven companies have higher profits and productivity than their competitors

Ultimately, leaders need to practise what they preach - making it clear that the decisions they make are based on data, demonstrating the links between data-based decision-making and positive business outcomes, and spelling out that they expect those working underneath them to inform their own decisions in turn. As much as there is an education gap in the workforce as a whole, there’s one at senior levels too, with a 2022 survey finding that only 52% of C-level executives are confident in their data literacy skills and 45% admitting to frequently relying on gut decisions.

Studies have repeatedly shown that data-driven companies have higher profits and productivity than their competitors, and technological progress only offers more opportunities for data to add significant value - any business leader not prioritising data literacy issues is leaving money on the table.

What are ‘quiet critics’ and how can you hear them?

In today's digital landscape, not all dissatisfied customers voice their concerns. Learn how 'quiet critics' can impact your business and how to detect their unspoken frustrations through behavioural data analysis

Every business needs to listen to its customers to understand what they’re doing right, what they’re doing wrong, and where there are opportunities to change and improve. And on the face of it, doing so has never been easier with all the different ways customers can leave feedback through online reviews and social media posts. But not everyone who interacts with your business is going to give you that kind of response.

In the same way that the world of HR has begun to talk about ‘quiet quitting’ - employees who don’t complain about their work, but strip back the effort they put in to the bare minimum - businesses trying to understand their customers need to think about ‘quiet critics’. A quiet critic is someone who interacts with your business, has a bad experience and then simply disappears.


If there is even a slight delay, glitch or error in a digital experience, it can quickly have a huge impact on business. As many as 60% of shoppers have abandoned a purchase due to a poor user experience, with 88% saying they would abandon a payment process if they experienced significant friction. According to digital behavioural data analytics firm Fullstory, around 43% of shoppers who have a bad experience trying to buy something online won’t bother to complain about it. So how can firms understand what’s going wrong and change accordingly?

The key is to collect behavioural data on how users are interacting with a website or app - they might not leave a review complaining about it, but if the right data is collected, their frustrations can be detected in certain signals through the way they interact with the site.

For example, a customer might leave ‘rage clicks’, clicking or tapping repeatedly on a button or other component of a site that isn’t doing what they expected: a ‘buy now’ button that hasn’t been active because the form above is missing some information, a decorative feature of the site that looks like it should perform a function, or a broken link. 

For businesses that want to learn and adapt, listening to customers is vital

Similarly, if users are thrashing around their mouse cursor erratically, it may indicate they’ve reached a point in the process where the next step isn’t clear or are experiencing high load times leaving them feeling stuck. While this behaviour won’t necessarily indicate what’s going wrong, it is the first step in identifying areas that need improvement, because most users aren’t going to take the time to fill in a contact form to leave that kind of feedback themselves.

There are other signals that digital content has issues: do a lot of mobile customers use the pinch-to-zoom gesture? That could be a sign that font sizes are rendering at too small a size on most phones. Though the user can zoom in it’s still a more awkward experience than it needs to be, and throws another obstacle between them and the checkout process. The more obstacles, the fewer users will arrive where you want them to. 

As well as problems with the overall design of a digital user experience, behavioural data can also surface technical issues like ‘error clicks’ - where customers interact with elements on a site that have functionality issues. The way they react can help identify priority issues for developers to fix, for example, a bug in the underlying JavaScript code. 

For businesses that want to learn and adapt, listening to customers is vital. A huge number of consumers won’t explain their issues ‘out loud’, so investing in behavioural data is key to discovering, and fixing, the pain points that are causing users to leave a site and never return. Understanding as much as possible about how users interact with sites and apps is vital for businesses to create better digital experiences that keep potential customers on track to becoming actual customers.

Ed Jefferson
Ed Jefferson Freelance journalist and creative technologist, his writing has been published in The Guardian, the New Statesman and CityMetric.