The last thing you need in a competitive broadband market is for your network to be your own worst enemy.
But that’s what can happen when persistent, unresolved network errors go unresolved – resulting in a loss of subscribers and the sunk acquisition cost it took to win them.
While most network errors never reach subscribers, and the occasional ones that do rarely will cost you a customer, the math changes when there is a steady flow of uncorrected issues. OpenVault analysis of more than 200,000 accounts at a single operator showed that subscribers with sustained data loss churned 36% more often. Patterns of more severe data loss more than doubled that figure.
Beware uncorrectable codewords
Here’s how that can happen: Data travels across broadband networks in small blocks called codewords. Error correction normally repairs damaged blocks before they impact the customer. But when a block is too damaged to fix – an uncorrectable codeword – the data is lost and must be resent, intruding as buffering, stalls, and choppy video calls that damage the subscriber experience.
In the analysis, OpenVault studied how the number of days in which 1% or more of codewords were uncorrectable impacted churn, as well as how a severe error rate of 5% or more of codewords amplified the problem. The results were unambiguous:
A bad day doesn’t drive churn. A bad month does.
| Data loss in a month | Monthly churn | vs. no data loss |
|---|---|---|
| None | 1.7% | Baseline |
| Occasional 1% loss (1 to 10 days) | 1.8% | Little or no increase |
| Persistent 1% loss (11+ days) | 2.3% | 36% higher |
| Persistent 1% loss and severe 5% loss | 3.8% | More than double |
A day of data loss means 1% or more of codewords were uncorrectable that day. Severe means 5% or more.
The longer a problem goes unaddressed, the higher the risk. This at-risk list exists on your network today, shaped by plant age, maintenance practices, and how quickly impairments get fixed. Without advanced tools that provide proper insights, operators often are unable to identify which accounts are impacted the most by persistent data loss. Left unresolved, these impairments lead to the worst call a subscriber can make: telling you about a problem you didn’t know you had.
The right questions to ask
To minimize data loss and churn requires real-time understanding of what is happening on your network, and prioritization of issues. Broadband providers should be asking themselves these three questions:
- Can I list the accounts with persistent data loss right now?
- Do my truck rolls go to the worst accounts first?
- How many of these subscribers learned about the problem before I did?
Our analysis shows that the wrong answer to any one of those questions could send subscribers streaming to the exits.
How Vantage and OV PNM help
OV PNM, part of OpenVault’s Vantage suite, can elevate your ability to give your teams the insights they need to ensure that your network is at peak performance. Software-only and vendor-neutral, Vantage combines OV PNM and OV Advanced PMA™ in one platform to:
- Track persistence, flagging accounts with sustained data loss, not those with occasional errors.
- Prioritize by impact, ranking impairments by QoE and retention impact so crews reach the worst problems first.
- Fix the root cause, identifying shared impairments on one node that point to a single fault and a single repair.
- Act early, delivering early warnings on issues like pre-equalization stress that aren’t associated with churn on their own, but could result in data loss if not addressed.
The operational payoff is proven. A Tier 2/3 operator cut truck rolls 26% in upgraded communities, and a Tier 1 operator running OV PNM continuously saw 37% to 43% fewer truck rolls and service requests.
Let us help you see the at-risk list on your network. Schedule a 30-minute demo at sales@openvault.com – before your subscribers make the issues painfully clear.
Results will vary by operator with plant condition, network technology, and local competition.