Why Am I Losing Customers? Part 2 - Technical Understanding Oriented Towards End Customer Behaviour

In Part 1 of this blog series, we explored data-driven insights into churn behavior, leveraging Axiros' solutions to analyze real-world subscriber data. Through graphical analysis, correlation studies, PCA dimensionality reduction, and statistical tests, we uncovered key indicators of churn and how telecom providers can mitigate it.

Now, in Part 2, we take a deeper dive into Churn Analysis and Correlation Insights, identifying patterns that drive customer churn.

Churn Analysis And Correlation Insights
The following items were identified after the churn analysis and contrasted with the correlation analysis of the previous part, considering these as a quick guide to end-user behavior:

• Stable coverage and optimized WiFi → Improved user experience and reduced churn

• Higher utilization of 5GHz, Ethernet, and Optical traffic → Strong retention indicator

• Good network health (WIFI, Ethernet, Optical) → Lower churn probability

• Fewer anomalies in CPU, memory, and events → More satisfied users, less reboots

• Higher event severity and frequent anomalies for WIFI, Events, and Traffic → Churn risk increases

The following table has a deeper understanding per item:

Churn Analysis	Found Correlations	Insights
No-churn users show higher WIFI 2.4GHz and 5GHz downlink and uplink traffic usage, implying better network utilization. A strong correlation exists between WIFI health and user retention—good WIFI 2.4GHz and 5GHz health cases are significantly higher for no-churn users. No-churn users also show a higher number of WIFI host health good coverage cases, leading to better network experiences. Poor WIFI performance (higher anomalies, interference, or weak signals) is associated with higher churn probability. No-churn users have higher 5GHz WIFI usage, benefiting from less interference and higher throughput. Poor WIFI placement and coverage contribute to churn, as seen in lower min values of WIFI 5GHz host coverage for churn users. WIFI 2.4GHz bad coverage cases are more frequent in no-churn users, possibly due to more intelligent band-steering or proactive troubleshooting. No-churn users show higher invariability in WIFI host health and event health, indicating a stable experience over time. No-churn users have higher Optical downlink traffic usage, showing a direct relationship between higher fiber traffic demand and retention. Stable Ethernet health and higher Ethernet traffic usage correlate with lower churn rates. A higher optical health variation is linked to better retention, while poor optical health contributes to churn. No-churn users experience fewer critical event anomalies, indicating more stable device behavior or less manual reboots. They have a higher number of good event health cases, suggesting fewer unexpected connectivity issues. Churn users tend to have higher event severity cases, likely causing frustration and dissatisfaction. No-churn users show lower CPU and memory health anomalies, indicating more efficient resource utilization. Good memory health correlates with better user experience and lower churn probability. High CPU or memory anomalies indicate performance degradation, often triggering higher churn rates. Churn users experience higher anomalies in Ethernet, WIFI, and event severity, signaling unpredictable network issues.	Less interference increases traffic, improving streaming uses and general performance. 5GHz optimization reduces congestion on 2.4GHz. A strong relationship exists between interference in 2.4GHz and 5GHz—optimizing one benefits both. WIFI uplink and downlink performance must be monitored separately, both include relevant information about behavior. Strong RSSI improves coverage, but too many devices per AP degrade rates. Mesh systems outperform high-gain single APs for better performance. Interference from other networks disrupts stability and rates. Band steering improves overall experience by shifting traffic to 5GHz. Coverage should be optimized for both 2.4GHz and 5GHz to balance stability. Increased uplink/downlink usage demands more network capacity. Most WAN traffic originates from WIFI usage (both 2.4GHz and 5GHz). Ethernet health influences traffic stability and experience. Devices with stable WAN Ethernet connections reboot less often. Heavy Ethernet and WIFI users tend to avoid reboots unless interference or poor coverage occurs. WIFI 2.4GHz and 5GHz interference/noise often trigger reboots. Users experiencing stable WIFI performance typically use higher speeds. Higher traffic demands more CPU and RAM, essential for stability. RAM is critical for buffering and maintaining device performance. Reboots (BOOT event) free up memory (RAM). Hardware influences possible WIFI rates but not coverage or connection stability. WIFI 2.4GHz and 5GHz traffic require significant CPU resources. Continuous WIFI scanning and diagnostics demand additional RAM. WIFI 5GHz stability depends on RAM availability.	Improve WIFI health and use 5GHz as much as possible. Band steering improves overall network performance by shifting users to less congested 5GHz. Both 2.4GHz and 5GHz coverage should be optimized to balance stability and throughput. Channel optimization for 2.4GHz and 5GHz is highly recommended. Strong RSSI improves coverage, but overloading APs reduces efficiency. Proactive network management and auto-optimization play a role in user retention. Users with greater WAN usage are more likely to stay. Reliable WAN connections correlate with lower churn. Persistent optical degradation leads to higher churn, monitoring is crucial. Optimizing 2.4GHz and 5GHz performance is critical. Fewer Critical Event Anomalies Reduce Churn – More stable device behavior leads to better retention. Fewer unexpected connectivity issues mean fewer manual reboots. Stable WAN Reduces Reboots – Reliable WAN connections minimize manual end-user interactions. WIFI Interference Triggers Reboots – 2.4GHz/5GHz noise and instability lead to manual reboots. Heavy Users Reboot Less – They avoid reboots as much as possible. Lower CPU/Memory Anomalies Reduce Churn – Efficient resource usage leads to better retention. Good Memory Health Improves User Experience – Stable RAM performance lowers churn probability. Higher Traffic Increases CPU/RAM Demand – Stability relies on sufficient processing power. RAM is Key for Buffering and Performance – Ensures smooth operation and connectivity. Reboots Free Up RAM – BOOT events help restore device efficiency. Hardware Affects WIFI Speed, Not Stability – WIFI rates depend on processing power. WIFI 5GHz Stability Relies on RAM – Adequate memory ensures reliable performance. Continuous WIFI Scanning Uses More RAM – Diagnostics impact resource allocation.

(Table 3 - Churn Analysis And Correlation Insights Per Item)

Conclusions
Hardware Performance and Resource Utilization
Hardware performance is crucial for network stability, as increased traffic demands more CPU and RAM. RAM plays a key role in buffering and maintaining CPE functionality, while WIFI 2.4GHz and 5GHz traffic require substantial CPU power. Continuous diagnostics and scans add RAM demand, especially in high-interference environments. Devices with stable WAN Ethernet connections and high Ethernet or WIFI usage tend to reboot less, whereas users experiencing WiFi interference, poor coverage, or unexpected behavior frequently restart their CPEs.

No-churn users show lower CPU and memory health anomalies, indicating more efficient resource utilization. Good memory health correlates with better user experience and lower churn probability. High CPU or memory anomalies indicate performance degradation, often triggering higher churn rates.

WiFi Performance, Coverage, and Interference Management
WiFi traffic and stability directly impact user experience and retention. No-churn users exhibit higher WiFi 2.4GHz and 5GHz downlink and uplink traffic usage, indicating better network utilization. Strong RSSI improves WiFi 2.4GHz coverage, but excessively connected devices degrade rates, making mesh systems preferable over high-gain APs.

WiFi band steering improves congestion by offloading traffic to 5GHz, enhancing the overall user experience. Optimized 5GHz coverage reduces strain on 2.4GHz, while interference affects both bands, with poor 2.4GHz conditions degrading 5GHz performance. Consequently, optimizing both frequencies, including channel selection, reduces congestion and enhances stability.

No-churn users have higher 5GHz WiFi usage, benefiting from less interference and higher throughput. Poor WiFi placement and coverage contribute to churn, as seen in lower minimum values of WiFi 5GHz host coverage for churn users. However, WiFi 2.4GHz bad coverage cases are more frequent among no-churn users, likely due to more intelligent band-steering or proactive troubleshooting.

WAN Traffic and Stability
Ethernet health, influenced by quality of service, impacts overall traffic stability. Increased uplink/downlink usage necessitates greater network capacity, as most WAN traffic originates from WiFi. No-churn users have higher Optical downlink traffic usage, showing a direct relationship between higher fiber traffic demand and retention. Stable Ethernet health and higher Ethernet traffic usage correlate with lower churn rates.

A higher optical health variation is linked to better retention, while poor optical health contributes to churn, considering variable optical is a lower problem compared to all the time optical degraded users.

Event Anomalies and Stability Indicators
Event stability is another critical factor in retention. No-churn users experience fewer critical event anomalies, suggesting more stable device behavior and fewer manual reboots. They also show higher numbers of good event health cases, indicating fewer unexpected connectivity issues. Conversely, churn users tend to have higher event severity cases, likely causing frustration and dissatisfaction due to increased reboot frequency.

Additionally, stability in network performance is crucial. No-churn users display higher invariability in WiFi host health and event health, indicating a stable experience over time. In contrast, churn users experience higher anomalies in Ethernet, WiFi, and event severity, signaling unpredictable network issues.

Why am I losing customers?

Probably some churn-related elements mentioned in this article are new to you, and you are ready to deal with them for better churn management. But you must consider the most important thing: You need to monitor, search, and understand end users with greater intelligence.

This example will be helpful:

• John has a high-speed Internet service since 5 months

• John is experimenting degradation or intermittence in the WiFi (or WAN access) and he is considering moving to another provider

• He feels an experience degradation and other providers offer better prices, you are close to losing a client

• Due experience degradation he performs reboots and even try with a factory reset, this has a placebo effect:

  • A lot of the reboots are related to experience and are not related to hardware issues

  • Sometimes will help to free RAM, but why not do it automatically?

• When he experiences a WiFi degradation will perform also some reboots:

  • Instead, why not perform an automatic channel optimization?

  • Or suggest other APs/mesh before this point?

• After all the reboots and bad experience he will generate less traffic usage:

  • Check it, he is close to leaving you but there is still time

As you can see, a basic churn management can be created with specific use cases but you require some tools for this.

A Churn Management Process
Considerations

The detected correlation and churn behavior in this article enable the following suggestions, these can help your end users:

1. Enhance WiFi 5GHz Optimization → Encourage users to adopt 5GHz and deploy band steering for better load distribution

2. Monitor CPU and Memory Utilization → Offer hardware upgrades for heavy users experiencing high resource anomalies

3. Improve User Experience via Proactive Reboots → Identify patterns in Hardware Usage and trigger automated corrective actions, especially for RAM usage

4. Reduce WiFi Interference → Perform massive WIFI 2.4GHz and 5GHz channel optimization to get a better channel distribution reducing interference

5. Optimize Coverage and Reduce Interference → Deploy mesh systems in high-density areas to improve experience, and try to educate the users about optimal device placement

6. Prioritize Proactive Monitoring → Detect Ethernet, Optical, and WiFi anomalies early to prevent service degradation, considering traffic reduction is usually an indicator of a possible churn

Going into action
The following actions can be taken to reduce churn, identify if these actions can be performed by your ACS or ask Axiros to obtain all these solutions with no limitations:

• Monitor hardware and check the firmware and model behavior periodically, considering:

  • Less RAM and CPU usage are necessary for higher speed Internet and WiFi utilization enabling the full speed and experience for Internet

  • Less RAM usage impacts to less reboots due WiFi and WAN tasks

• Disable the auto-channel included in the CPE firmware and enable a powerful channel optimization mechanism:

  • Proactive tasks are suggested for a better WiFi experience (2.4GHz and 5GHz network)

  • This will free some RAM with better channel detection to reduce interference

  • This will improve WiFi 2.4GHz and 5GHz health ensuring less churn

• An automatic reboot considering the WAN traffic can be triggered when the RAM usage exceeds some limits:

  • The relation between RAM usage and churn behavior is clear

• For end users with poor WiFi experience, the two rules are:

  • A WiFi channel optimization must be used, it will help as a churn management process

  • Rather than relying on high-gain single APs, deploying mesh systems reduces device congestion per AP and improves network distribution

• The following monitoring is required to reduce churn:

  • Medium Access Health:

    • Proactive in case of medium access errors (like an optical degradation or fiber cut)

  • WiFi Traffic Health, Traffic and Hosts:

    • Proactive to offer mesh solutions and enable WiFi channel optimization

  • Hardware Health:

    • Proactive to offer a more capable device for the cases it is used for

    • Proactive to request a better firmware for specific issues

• Mesh solutions must be installed in important places to customers, considering the coverage for 5GHz must be the same for 2.4GHz:

  • This reduces the interference from other networks and improves the speed

  • It is the basic process for churn management, making clear the relation between WiFi experience and churn

• Users should be informed about WiFi placement, the benefits of 5GHz, and how interference affects performance:

  • Raising awareness of the use of 5GHz and how to improve WiFi according to the knowledge mentioned in this document is ideal

Clarification
For all the monitoring and management necessities mentioned in this article, Axiros has a solution, considering a proper churn management needs high-end technology, data management, and proper knowledge.

Written by Cesar Quirga, M.Sc
Cesar Quiroga is a Project Definition Engineer at Axiros, specializing in defining projects scopes based on ISP necessities and ensuring that technologies align with evolving industry demands. He combines technical expertise with analytical skills to drive impactful solutions. Currently, his focus is on machine learning, statistical analysis, and data-driven decision-making, enabling him to design innovative solutions that enhance the end-user experience.