Payments
Account updater
Account updater is a card network service that pushes reissued or renumbered card details to enrolled merchants so stored credentials keep working without customer action.
When it works the failure never occurs: a card that expired last month simply charges successfully. That makes it the single highest-leverage involuntary-churn intervention, because it costs no customer goodwill at all.
Coverage is not universal — it varies by network, by issuer and by how the credential was stored. Enrolment is nonetheless close to free relative to the recovery it produces, and it acts before the failure rather than after it.
What the network services actually are
Each network runs its own service: Visa Account Updater, Mastercard Automatic Billing Updater, American Express Cardrefresh and the Discover equivalent. They are not one system and they do not behave identically, which is the first thing that surprises teams who enable "account updater" as a single switch in a processor dashboard.
The mechanism is a batch exchange rather than a lookup. Your acquirer or processor submits the stored credentials you have enrolled, the network matches them against updates that participating issuers have supplied, and a response comes back for each one. That last part is where most of the value sits, because the response is not only "here is a new number".
A submission comes back as one of a small set of outcomes: new account number, new expiry date, account closed, contact the cardholder, or no change. The first two save a charge. The third and fourth save you from a retry sequence that was never going to work, and they arrive before the failure rather than after it. Systems that only act on new numbers throw away half of what they paid for.
What it can and cannot fix
The service exists to solve one problem: the stored credential is stale because the bank changed it. Every failure mode outside that description is untouched by it, and reading the boundary correctly is what stops an updater from being treated as a general recovery tool.
The boundary is worth drawing explicitly, because the failure modes it does not cover are the ones that generate the most support contact, and a team that believes the updater has them covered will not notice until the cohort has already churned.
- Fixes: expiry roll on reissue, renumbering after a compromise, portfolio conversions where an issuer migrates its cards to new BINs, and brand changes when a portfolio is sold.
- Does not fix: insufficient funds, credit limits, issuer risk declines, authentication requirements — none of these are credential problems.
- Does not fix: a customer who deliberately cancelled the card to stop being billed. The updater may even return a replacement, and charging it is a dispute waiting to happen.
- Cannot reach: issuers that do not participate, which skews toward smaller banks and credit unions, and coverage that is materially thinner outside the networks' core markets.
- Largely redundant for: credentials already stored as network tokens, where the network maintains the mapping continuously and the token you hold never goes stale.
- Irrelevant for: bank debit mandates, wallet-funded payments and any method that is not a card in a network updater programme.
Latency is the reason it belongs before the charge, not inside dunning
An updater is a batch, and there are two queues in front of your answer. The issuer supplies its updates to the network on its own cycle, and the network returns matches to your acquirer on theirs. A card reissued this week may not surface in your response file for some time, and no amount of urgency on your side compresses either queue.
The design consequence is unambiguous. Querying the updater at the moment a charge fails is close to useless, because the answer is not ready and the customer is already in a failed state. The query has to run ahead of the renewal on its own cadence, so that the credential is fresh before it is charged.
A worked example makes the volume concrete. Consider a book of 20,000 active subscriptions where cards are issued on a three-year cycle and expiry dates are spread evenly. Roughly one thirty-sixth of the book reaches its expiry each month — about 555 credentials. Every one of those is a failure you can see coming weeks ahead, and the only reason to let it fail is that nobody scheduled the query.
The same cadence produces the pre-dunning list for free. Credentials the updater cannot fix, and credentials whose stored expiry precedes the next billing date with no update returned, are exactly the customers worth contacting before the charge rather than after it.
Where it sits relative to network tokens and dunning
Recovery interventions form a ladder, and every step down it costs more and returns less. The top of the ladder prevents the failure invisibly. The bottom asks a customer to go and find their wallet, which is where most of the churn actually happens.
Account updater sits second, above every reactive measure and below network tokens, which solve the same problem continuously rather than in batches. A merchant already tokenising through the networks gets much of the updater's benefit automatically; a merchant storing raw credentials in a processor vault gets none of it without enrolling.
| Intervention | When it acts | What it fixes | What it costs the customer |
|---|---|---|---|
| Network token | Continuously, at authorisation | Reissue and renumbering, without you storing anything | Nothing — the customer never knows |
| Account updater | On a batch cadence before the charge | Expiry, reissue, closure signals on stored cards | Nothing, when the query runs early enough |
| Pre-dunning outreach | Days before the renewal | Credentials no service can refresh | One email the customer has to act on |
| Decline-aware retries | After the charge fails | Soft declines, mainly funding timing | Nothing directly, unless attempts are visible |
| Payment-method update request | After retries are exhausted | Everything left over | The full cost: they must re-enter a card |
Frequently asked questions
- What is an account updater?
- An account updater is a card network service that supplies merchants with updated details for cards they have stored. When an issuer reissues a card with a new number or expiry date, participating merchants receive the change through their acquirer and can keep billing without contacting the customer. Visa, Mastercard, American Express and Discover each operate their own version.
- How does an account updater work?
- The merchant's acquirer submits enrolled stored credentials to the network in a batch. The network matches them against updates that participating issuers have provided and returns a result for each: a new account number, a new expiry date, an account-closed notice, a request to contact the cardholder, or no change. The merchant applies the updates before the next scheduled charge.
- Does account updater work for every card?
- No. Coverage depends on the issuer choosing to participate, and participation is uneven — thinner among smaller banks and credit unions, and thinner outside each network's core markets. It also only addresses stale credentials. Insufficient funds, credit limits, issuer risk declines and authentication requirements are untouched by it, because none of those are credential problems.
- How long does an account updater take to return a new card number?
- It is a batch process with two queues in front of the answer: the issuer supplies updates to the network on its own cycle, and the network returns matches to the acquirer on theirs. There is no real-time lookup, so querying at the moment a payment fails is too late. The query must run on a schedule ahead of the renewal.
- Do I still need dunning if I use an account updater?
- Yes. An updater prevents failures caused by stale credentials, which is a large share but not all of them. Insufficient funds, credit limits, issuer risk declines and customers who deliberately cancelled a card all still fail, and those need decline-aware retries and outreach. The updater reduces the volume that reaches dunning rather than replacing it.