Shifting from the credit score gate to ongoing payment risk assessment and proactive payment support.
New Zealand’s electricity retailers face a difficult balancing act.
Electricity is an essential service, so there is understandable pressure to ensure consumers can access supply. At the same time, retailers carry the financial risk when customers are unable to pay. Asking them simply to relax credit standards without giving them better ways to manage that risk is not a sustainable solution.
The Electricity Authority’s recent consultation on improving consumer access and choice brought that tension into sharp focus. In 2025, around 46,900 applications for post-pay electricity were declined for credit-related reasons. But the retailer submissions show that the story is much more complicated than a simple choice between accepting and declining a customer.
For example, Mercury trialled an alternative transaction-based assessment for 107 customers who had failed its standard credit process. After 12 months, approximately three-quarters were still active Mercury customers. The cohort had higher arrears than other customers, but the trial demonstrated something important: some people who would have been rejected by traditional credit assessment were nevertheless able to maintain an electricity account.
Genesis makes the other side of the argument equally clear. Credit assessment remains an important part of prudent retailing because retailers carry substantial exposure to unpaid energy consumption. Genesis does not automatically reject every applicant who fails its automated assessment. Some are referred for specialist review, where additional context is considered before a final decision is made.
Both positions make sense.
Perhaps the opportunity is not to choose between credit assessment and greater access. It is to rethink when and how often we assess payment risk.
The best data arrives after the customer does
A credit score is necessarily a snapshot.
It is based largely on information collected before the customer joins a retailer, often from external sources and for purposes other than managing an electricity account.
Once the customer relationship begins, the information available becomes much richer.
Retailers start accumulating their own first-party information: bills, payments, payment timing, balances, usage, payment arrangements, digital activity, service interactions, support requests and responses to previous communications.
Most individual signals mean very little on their own. A late payment may be an oversight. An unusual top-up may reflect a change in routine. A customer logging into an app several times may simply be curious about their usage.
The value comes from changes and patterns in context.
Contact Energy makes this point particularly well in its submission. It cautions that indicators such as prepay disconnections should not automatically be interpreted as financial difficulty. A disconnection can reflect choice, absence from the property or delayed top-up behaviour. But repeated or extended disconnections, particularly when combined with other information, may indicate affordability pressure. Contact notes that retailers can consider factors including disconnection frequency and duration, top-up behaviour, customer contact history and vulnerability indicators, and says it monitors customers with repeated or extended disconnections for potential follow-up.
That principle should apply well beyond prepay.
Rather than asking once, at acquisition, “Is this customer a credit risk?”, we could increasingly ask:
“Is this customer’s payment risk changing?”
That is a very different question.
From screening risk to managing risk
This matters because customers’ circumstances are not static.
A customer who looks risky at acquisition may prove perfectly capable of maintaining an account. Another who passes every credit test may experience a job loss, relationship breakdown, rent increase, mortgage reset or unexpected expense six months later.
Neither outcome is particularly surprising. Yet much of the industry’s approach to credit risk remains heavily weighted towards what we know before the relationship begins and what becomes obvious after a payment is missed.
There is a substantial space between those two moments.
Modern analytics can monitor patterns across first-party operational data and identify customers whose behaviour is beginning to change. That does not mean diagnosing hardship or making assumptions about why someone may be struggling. The underlying cause may remain unknown.
It means recognising that the probability of a poor payment outcome is changing, early enough for the retailer to respond constructively.
The response matters just as much as the prediction. Different customers will need different forms of engagement. Some may simply need a timely reminder. Others may need a payment arrangement, a conversation about available assistance or referral to more specialised support.
Contact’s submission highlights the challenge with existing high-touch support pathways. Advocacy and referral arrangements can produce good outcomes, but they can also be highly manual, relationship-based, resource-intensive and difficult to scale.
Better identification could help us reserve that expensive human support for the customers most likely to benefit from it.
Could this change the economics of acquisition too?
There is also a slightly more provocative question worth asking.
If we become materially better at monitoring payment risk after onboarding, does every prospective customer need to face the same level of upfront credit screening?
Not tomorrow, and certainly not without evidence.
But Mercury’s transaction-score experiment suggests that traditional screening can exclude some customers who ultimately prove capable of maintaining supply. At the same time, the higher arrears in that group show why simply removing the screening barrier would be naïve.
A more interesting model may be lighter barriers at the front door, combined with much closer and smarter monitoring after the customer comes in.
If that approach could be demonstrated safely for selected cohorts, the benefits might extend beyond consumer access. Less friction in acquisition and less dependence on external credit assessment could potentially reduce acquisition cost as well.
That should be treated as a hypothesis to test, not an assumption.
The shift is from a gate to a relationship
Credit scores will continue to have a role. They solve a real problem for retailers and provide useful information at a point when comparatively little is known about a prospective customer.
But they should not be expected to carry the full burden of managing payment risk.
Once a customer joins a retailer, the information environment becomes richer, fresher and more relevant every week.
The bigger opportunity for our sector may therefore be to shift our thinking from credit risk as a gate at acquisition to payment risk as something managed dynamically throughout the customer relationship.
That could allow retailers to protect themselves financially while giving more customers a sustainable pathway to electricity supply.
Those objectives do not have to be in conflict.
We’d be interested to hear how others in the sector are approaching this, and whether payment risk is beginning to be managed more dynamically after onboarding.
