Expert Opinion

Drive competitive advantage and customer excellence with customer insights

  • #Digital Experience

Patrick Simon | Director Digital Keyrus

Knowing your customer, anno 2017, remains a key challenge. Companies are still mostly ‘best guessing’. Customer expectations, on the other hand, are expanding: personalized interactions, seamless omni-channel, sales and tailored products and services. So how do companies meet these new needs?


Some companies start to centralise customer data and align information streams from various sources and departments. Others go above and beyond internal customer knowledge and use advanced analytics on external sources to understand the customer almost better than himself. Why? To anticipate the customer’s preferences and next steps, anticipate churn and detect new opportunities of course!
You probably have varied internal sources organised in silos along with plenty of potential external data. Maybe you hoped it would be easy to make the transition from traditional to more enhanced customer analytics. Such a transformation involves a radical shift towards datadriven decision making; it impacts the company culture and its organisation. Achieving customer excellence is complex.
A recent study1 showed that “66% of consumers who switched brands did so because of poor service” and also that “70% of companies that deliver best in class customer experience use customer feedback - versus industry average of 50%, and 29% for laggards.”
So service dissatisfaction drives churn, and best in class customer experience is heavily based on understanding the customer and analysing their feedback.

Customer intelligence is a key driver to:

  • Understanding dynamics in customer behaviour towards your offering, company and market
  • Developing a clear image of your ideal “customer” and “customer journey”
  • Stimulating innovation and product development to drive competitive advantage
  • Getting insight into future trends and identifying ways to increase loyalty and customer retention

The prominent question is: ‘How do I, irrespective of the level of analytics, manage the complexity of data, and derive business value from it?

Sebastian Rueling, Head of Multi Channel Operational Excellence at UCB:
“Customer engagement is not about the channel. The question to be answered first is 'what job are we getting done for our customer?' Only once that's 100% clear the channel strategy optimizes how that value is delivered. Specific and reliable customer insight is the key to both finding what's valuable and how it can best be delivered to the customer.”

4 MAIN STEPS IN CUSTOMER INSIGHTS - From centralisation of data to advanced predictive analytics

The graph below identifies 4 steps that are linked to the complexity of analytics. Based on the business challenges and needs, you can conduct insight with a limited or a high complexity of analytics. Based on the type of data, and type of sources, different insights are generated.


The first step is to create an understanding of who your customer is in one single version. It is available at all times for your organisation to make decisions. To create a 360° view of the customer, information is captured from various angles such as billing information, web portal / self-service, customer service and sales. 
Once captured, there are multiple ways to organise both the structured and unstructured data. Some companies invest in a solid platform that is centralised, and allows aggregation of customer data from various channels. Another way is to implement cloud-based solutions that allow real-time customer insights by leveraging existing CRM systems.
The next crucial step is data visualization. There are various ways to display data, based on why the data is needed. It should serve the initial business purpose. Evidently, it must be user-friendly, accessible and flexible for the users.

Elke Laeremans, Chief Digital Officer at Torfs:
“aligning our omni-channel sales and customers was a key step to improve our category management, find trends in customer groups and improve our customer interaction."


  • Have a 360° and reliable view to optimise customer interactions
  • Understand customers and target lookalike audiences
  • Find trends in customer groups and segments that result in actionable insight
  • Automate insights on client issues, product performance, customer relationships



The main objective here is to analyse action-based information about your customers. It allows you to analyse and monitor the complete customer journey. Main sources include weblogs, consumption history, operational data, purchase history, channel behaviour and campaign response.
To understand which information is essential, it’s key to have a clear vision on personae’s and link them to customer journeys. As a next step, you can map and analyse them and identify key touchpoints and events. Now it becomes possible to influence and optimise the events that are important to your customer.


  • Decrease cost of customer acquisition
  • Discover new lead indicators that can develop business
  • Stimulate a certain behaviour or next step of your customer (cross/up selling)
  • Find trends in customer groups or segments
  • Improve targeted marketing campaigns



Enriching your customer profile with external data is essential to driving product innovation, improving customer experience and efficient brand positioning. Based on the business question and objective, data can be captured from web and social media sources. 
It’s important to leverage on potential existing social media tracking tools. Many social media tracking tools allow you to have quick access to social media and web communication, however they might not ensure the right filtering and relevant content.
Another approach is to develop, train and build custom models around a specific topic. This immediately increases relevance and business value of the data.

Stef Peeters, Data Scientist at Keyrus, explains:
“For a major FMCG company, we managed to increase sentiment classification accuracy by up to 70%. Benchmarking showed that our model was even better than commercial API's. This was achieved using a combination of traditional text mining techniques and neural networks.” 


  • Understand consumer behaviour, preferences and sentiments
  • Improve customer experience by improving product development
  • Enhance feedback loop on product improvement
  • Find out what is being said about a brand or product, who is saying it, on which platforms, how often and who the main influencers are
  • Leverage your presence on social networks to build a strong brand and effectively manage your social media accounts and e-reputation
  • Identify key topics and FAQ that are essential to customers, competitors and/or industry



The main difference of this step is the predictive aspect. How about predicting what your customer will like in the future? Or how about building alert mechanisms that help your organisation against competitive threats?
Using next generation predictive algorithms, you can model evolutions and trends of consumer sentiment and behaviour. Not a crystal ball yet, but the real ability to spot early signs of positive or negative trends. You can see signs of a sales uptake and can capitalize by stocking up, or you can see a bad buzz coming and react pre-emptively.
Another example of predictive insight in your customer is the Cambridge Magic Sauce API. It allows you to predict multiple personal attributes of customers and customer groups. Imagine launching a campaign where you can enrich your customer profile with gender and age. It allows you to immediately fine-tune your message content.


  • Predict customer profile elements, trends, sentiments or behaviours
  • Understand consumer next steps in terms of loyalty (churn)
  • Identify future trends for product development
  • Build alerts that notify specific events that can jeopardize your business



The following approach is based on experiences with customers over the last 3 years, mainly in the FMCG, retail, banking and telecom. These industries are leading with initiatives around analytics in general and more specifically applied to Customer Intelligence. 

  • Understand - Identify key business questions and assess the expected value of the data insights. The aim is to shape business cases via a creative thinking process with a multi-disciplinary team.
  • Model – Move further into data discovery by applying data science models. Confirm the ability to derive insights from the available data.
  • Integrate – Deploy and integrate the data analytics solution into business processes that bring value to the organisation. Ensure visualization is based on the business needs.



  • Creative thinking process to shape and agree on a use case in order to address business needs in a pragmatic way
  • Identify key questions to address and measure the value of data insights that can be delivered
  • Team: Management, Business, CRM, Data Analysts/Scientists
  • Duration: 4-10 weeks



  • Capture and move data to an analytical platform
  • Perform data discovery to learn and refine data
  • Confirm ability to extract value from data
  • Team: Business, CRM, Data Analysts/Scientists, IT
  • Duration: 2-5 months


  • Apply analytical models to deliver insights
  • Present insights with the right visualization tool
  • Industrialise and embed the solution in the organisation
  • Team: Management, Business, CRM, Data Analysts/Scientists, IT
  • Duration: 3-12 months

Customer intelligence is not a stand-alone project; it is a process to customer excellence. It paves the way to achieving business objectives and steer strategic directions. As a response companies increasingly invest in customer intelligence due to fierce competition, innovation and rapid change.  

Main success factors of customer intelligence projects are:

  1. Identify and select a specific business area to focus on 
  2. Ensure iterative and agile approaches across the project
  3. Work with multi-disciplinary teams, creative talent (where relevant) and internal advocates (business experts)
  4. Ensure executive involvement and approval
  5. Capitalise on existing infrastructure for data discovery and analytics 
  6. Invest in transformation and culture management that supports data-driven decision making
  7. Focus and experiment with key stakeholders on various visualization options

In conclusion, regardless of which step or level in customer intelligence, there is value in data. It just depends on how clear the initial business question is, the quality of the data and the integration of it in the organisation.


[1] Source: ThinkJar study 2015



Patrick leads the Digital team at Keyrus Belgium. Additionally to his consulting background, Patrick held executive sales and marketing positions at bpost and VOO and is an active investor in Venn Telecom. His corporate and entrepreneurial experience gives him a pragmatic and result oriented mindset. Patrick’s main area of expertise are B2B Sales & Marketing strategy and digital transformation. He also lectures project management to ULB engineering students.


Keyrus, creator of value in the era of Data and Digital
An international player in consulting and technologies and a specialist in Data and Digital, Keyrus is dedicated to helping enterprises take advantage of the Data and Digital paradigm to enhance their performance, facilitate and accelerate their transformation, and generate new drivers of growth, competitiveness, and sustainability.
Placing innovation at the heart of its strategy, Keyrus is developing a value proposition that is unique in the market and centred around an innovative offering founded upon a combination of three major and convergent areas of expertise:

• Data Intelligence
Data Science - Big Data Analytics – Business Intelligence – EIM – CPM/EPM

• Digital Experience
Innovation & Digital Strategy – Digital Marketing & CRM – Digital Commerce – Digital Performance – User Experience

• Management & Transformation Consulting
Strategy & Innovation – Digital Transformation – Performance Management – Project Support
Present in 15 countries on 4 continents, the Keyrus Group has over 2500 employees.


Download PDF