Merchant of Record vs PayFac vs White Label: Choosing a Model for Your Platform

  • Alexandra Potapska, Head of Customer Success at Corefy

  • 10.11.2026 08:45 am
  • #payments

Most payment teams can open a dashboard in seconds. Fewer can say which number on it would make them change a routing rule today. That gap separates payment analytics that describes the business from payment analytics that runs it.

The role has already moved towards the second kind. In our analysis of 112 Payment Manager job descriptions, reporting and analytics were the most frequently mentioned responsibility, appearing 69% more often than PSP management. The same research shows how much of that analysis still happens by hand: approval rates and declines reconciled across PSPs in spreadsheets, because each provider reports them differently.

In customer success, the dashboard conversation comes up early with almost every team, usually about which charts deserve a place on the screen people open every morning. This column is my answer: the short list of metrics that drive payment decisions, the groundwork that makes payment data comparable, and how often each number deserves a look.

What belongs on payment analytics dashboard (and what doesn't)

Gross merchandise value, transaction count, and total revenue belong in every board pack. On an operational payment analytics dashboard, they mostly add noise. They move with demand, marketing, and seasonality, and a drop in them rarely tells the payment team what to change.

A practical test for every chart: can you name the decision it informs, the person who owns that decision, and the threshold at which they act? If any answer is missing, the metric belongs in a monthly report, not on the screen your team works from.

Two more conditions turn a number into a signal:

  • Baseline. An 87% approval rate means little on its own, but against the same window yesterday and last week, it shows whether something is moving.

  • Right cut. A total across all providers hides the provider that caused the problem, so most metrics below are worth reading per provider and per payment method before anything else.

7 payment metrics to track, and the decision each one drives

The metrics below move from whether payments get through to what each one costs and risks.

Approval rate by provider and payment method

This is still the anchor metric, and approval rate is the KPI employers name most often in Payment Manager job descriptions. I covered how to define and lift it in a separate column on improving approval rates across several PSPs. For analytics, one rule matters most: track first-attempt and final approval separately, so cascading and retries do not mask a weakening primary route. The decision it drives is routing priority.

Decline mix: soft, hard, technical, and rule-based

Approval rate shows how much volume failed. Decline mix shows what kind of failure it was, which determines who fixes it. Group declines into four buckets: issuer soft declines, which are recoverable through a retry or a second route; hard declines, which no retry will fix; technical errors on the provider or integration side; and blocks from your own fraud rules or failed authentication. A growing share of technical errors on one provider is an escalation. A growing share of rule-based blocks is a conversation with your risk team. The anatomy of a declined transaction explains why these buckets need different fixes.

Payment method conversion rate

This is the share of attempts on each method that end in a completed payment. For alternative payment methods, much of the loss happens before any provider responds: a redirect the customer abandons, a banking app that times out, a QR code nobody scans. Method conversion catches problems that approval rate never sees. The decision it drives is which methods to show, in what order, and through which provider.

Provider health: failed, pending, and slow transactions

Failure share is the obvious signal. Pending share is the underrated one. A transaction stuck without a final status is neither revenue nor a decline. It generates support tickets, and a queue of pending payments often exposes a provider-side problem before failures do. Track how many transactions are pending, how long the oldest have been waiting, and how provider response times compare with their usual level. The decision here is fast: shift traffic and open an incident with the provider.

Payout success and provider balances

For marketplaces, trading platforms, and any business that pays out, a failed payout is a support case and a trust problem. A payout that cannot leave because a provider balance ran dry is worse, because it was predictable. Track payout success rate per provider and how each balance moves compared with the previous day and week. Then rebalance funds or reroute payouts before the gap appears.

Chargeback and fraud ratios

Chargeback rate and fraud rate both rank among the most frequently named KPIs in the job descriptions we analyzed. Card scheme monitoring programs set ratio limits per merchant, and acquirers often apply stricter internal limits. The useful view is each ratio per merchant account, plotted against the nearest threshold, with a trend line. Decide whether to tighten rules or move risky segments before a monitoring program makes that decision for you.

Cost per successful payment

Rate cards compare providers on transaction price. Payment analytics should compare them on what you pay for each payment that actually completes: all fees, including fixed fees charged on declined attempts, FX margins, and scheme pass-throughs, divided by successful payments. Then read that figure next to the approval rate, because the cheapest route per success is not always the most profitable.

A simplified example shows why. Take 1,000 attempts of $50 split between two providers with the same fixed fee per attempt:

 

Provider A

Provider B

Difference

Pricing

$0.30 + 1.5%

$0.30 + 1.7%

 

Approval rate

85%

92%

+7 pp

Successful payments

850

920

+70

Total fees

$937.50

$1,082.00

+$144.50

Cost per successful payment

$1.10

$1.18

+$0.08

Revenue per attempt, net of fees

$41.56

$44.92

+$3.36

Provider B costs eight cents more per successful payment, yet it completes 70 more payments and returns $3.36 more per attempt after fees. Judged on its rate card, B loses. Judged on revenue per attempt, it wins clearly. That pairing of metrics is what turns a provider review from a pricing debate into a business case.

How to compare payment data across multiple PSPs

Analytics is only as reliable as the payment data behind it, and data from several providers isn't comparable by default. Our research points to inconsistent error codes between providers as one reason cross-PSP comparison still lives in spreadsheets. Four pieces of groundwork fix most of it:

  • One status model. Map every provider's statuses into one set, so 'successful', 'failed', and 'pending' mean the same thing across your stack.

  • One decline taxonomy. Translate provider-specific response codes into the decline buckets above. Without this step, you can't compare decline mix across providers.

  • One clock. Calculate daily and weekly figures in your organization's time zone. Otherwise, a 'bad day' can be a side effect of a provider's UTC cut-off.

  • One definition per metric. Decide whether approval rate counts attempts or payments, which currency totals report in, and keep those definitions fixed.

This is the least visible part of payment analytics and the one most often skipped. It is also what lets a team trust a chart enough to act on it without first checking the provider's back office.

How often to review each payment metric

Not every metric needs the same attention. Provider failures and pending queues change within minutes and call for alerts. Approval rate, balances, and payout success suit a daily check. Decline mix, method conversion, and dispute ratios show their patterns over a week. Cost per successful payment and provider scorecards need a month of data to support a renegotiation or a change in the provider mix. Each rhythm also has a different owner, which is why a single payment analytics dashboard rarely serves everyone well.

A fair objection: a merchant with one provider in one market does not need seven metrics on day one. Approval rate, decline mix, and cost per successful payment cover most of the decisions this setup enables. The rest earns its place as providers and methods are added, which is when the cost of not seeing a problem early grows fastest. Our payment maturity research found that only 23.9% of businesses use complex operational functionality, including refunds, chargeback management, and analytics, so many teams are still building this layer.

Final thought

Start by removing. List every chart your team looks at and write next to each one the decision it informs. Anything without an answer moves to a report. Then fix comparability, so the remaining numbers mean the same thing across providers, and agree on who checks what, and how often.

Related Blogs

Other Blogs