Connect, Talk, Think, Act and the Connected Service Experience (CSX)

Most companies still only take action nowadays when the customer reports an issue. From a customer satisfaction perspective, reacting simply means being too late. Recent surveys show that 93% of companies fail to anticipate and act preventively. Although the technology to avoid this exists, as well as innovative and often disruptive use cases that are just waiting to be copied, many enterprises still wait for failure to happen and reports from customers before taking any action.


The situation is alarming: 99% of the data collected by companies today is latent while many opportunities are wasted. Companies apparently do not know how to use data to create value, and also fail to see that there are loads of cost-effective opportunities for them to improve and serve their customer remarkably better.

Reactive service processes that are bound to cause dissatisfied and complaining customers are typically H2H2S (Human-to-Human-to-System): manual, prone to subjectivity, and therefore extra time-consuming. If you recognize such situations, make sure to eliminate them before the competition does.

CSX: nine phases, five steps      
Although most processes aiming at a better Customer Service eXperience (CSX) are only semi-automated at present – involving sales people, field service technicians, and customer care employees – they already make a huge difference in improved customer satisfaction. Across industries, these nine main Connect-Talk-Think-Act phases are typical of many practices today:

1. A connected machine or device generates data that is transmitted using a machine data gateway.      
2. The data is aggregated and analyzed. When an anomaly is detected that can impact performance, a case is created.              
3. The knowledge base or expert system is queried to identify the “next best action”, e.g. skill set and part replacement.    
4. The integration with the knowledge base or expert system identifies the service level and response time.     
5. The integration with the HR system identifies the available co-worker with the proper skill set.            
6. The integration with the inventory or PLM system (Product Lifecycle Management) ensures that the required part or service is available or ordered.   
7. A ticket is created and a task is assigned.         
8a. A field service technician receives an alert on his mobile device. He informs a co-worker who performs the task and records the result on his mobile device, or …             
8b. sales or customer service receive a task in their CRM system (Customer Relationship Management). Then, sales sends an offer and/or customer support contacts the customer. The results are recorded in the CRM system.                
9. The resolutions are analyzed, and the knowledge base or expert system (see 3) is updated.

Digitally optimized connected services typically relate to knowledge bases, expert systems, PLM and CRM systems – inside the enterprise and beyond its boundaries, tapping into partner ecosystems. Connect-Talk-Think-Act is a colloquial translation of the corresponding 5A cycle that includes these subsequent steps of data Acquisition, data Aggregation, data Analysis, the Assignment of tasks, and the required Action to be taken:

Step A1 – Acquisition. Sensor modules extract system logs or operating data from connected machines, devices or components and transmit the data. Some of these “connected things” generate data that is not transmitted, which requires additional action. Ideally, performance and usage data is acquired and transmitted every n seconds.             
Step A2 – Aggregation. The operating data from multiple connected things is collected, aggregated and stored. The data is visualized. Most machine data stored nowadays is latent.
Step A3 – Analysis. Ideally, the data is being analyzed in real-time to identify issues. We distinguish between data visualization and analysis. When data is visualized, a person must monitor the data and decide if it is critical. This method is unreliable and subjective. Using historical and/or predictive tools, anomalies and issues can be automatically identified and trigger alerts set. Only a fully automated approach is reliable, objective, and timely.                
Step A4 – Assignment. When the data indicates an anomaly or when an issue is identified that might cause a problem or represents an opportunity to eliminate dissatisfaction beforehand, ideally a case is automatically created. The most appropriate resolution is selected from the knowledge base or expert system, and assigned to the proper actor for handling.   
Step A5 – Action. Interventions are performed to resolve the issue, and the process is recorded and analyzed. The results should be fed into the knowledge base or expert system. In many cases, the actor is currently a sales representative, field service technician, or customer care agent as in the following scenarios:

i.  Customer Care provides the operator tips and instructions.   
ii.  Sales & Marketing proactively sends the customer offers and orders, and/or …           
iii.  Technical Service is dispatched to perform maintenance or repair.   
iv.   The system sends messages, instructions and possibly software updates to the machine directly.

Consider these five A’s the core of a more detailed Data Analysis Lifecycle:          

Don’t hesitate: Connect • Talk • Think • Act, and call Sogeti

Award Winning Expertise
Latest Technology Trends
todo todo
  • Jaap Bloem
    Jaap Bloem
    Research Director at VINT
    +31 622 90 18 41
Leading Innovation
Thought Leadership
Follow us on
Follow us on