From connectivity to intelligence: where payment orchestration competes now

Louis Wapler, Senior Manager, and Rohan Shaju, Consultant at Edgar, Dunn & Company’s Paris office, discuss how payment orchestration is shifting from connectivity to intelligence, and where it competes now.

Payment orchestration is a layer that sits between a merchant and its payment providers. One integration connects to many acquirers, payment methods, fraud tools etc., and the orchestrator routes each transaction to the most appropriate provider. The merchant stops maintaining a dozen connections and starts managing one.

The orchestration segment has grown quickly. Estimates put the industry revenues at around USD 3 billion in 2026, forecast to reach roughly USD 15 billion by 2033. 100+ platforms now describe themselves as orchestrators globally, up from a handful a decade ago. The category spans venture-funded specialists and platforms offered by publicly listed groups (e.g., ACI Worldwide, Amadeus). From our work across markets, EDC has seen adoption for orchestration being the highest in five key sectors: retail and ecommerce, travel and hospitality, gaming and digital subscription, financial services, and marketplaces.

Orchestration is becoming a crowded category with an increasingly uniform product. Differentiation comes now from somewhere other than the core. This article traces that shift, from what orchestration promised and how much has been delivered, to why the economics are thinner than they look, and ends with the value-added layer that providers are building on top to stand apart.

What orchestration set out to solve

The original pitch had three parts:

• One integration in place of many, so a merchant could reach hundreds of acquirers, PSPs, and local payment methods without building and maintaining each connection itself
• Smart routing on top of that, sending each transaction to best possible route, and failover routing automatically when one goes down
• And a single view of payment data across every provider, so the whole operation could be seen and reconciled in one place

All three were real problems. The first has been delivered and is now close to table stakes. The second is where providers compete today, although what sits behind the smart routing label varies considerably across platforms. The third is the one still in progress, largely because a global business takes payments through more channels and systems than any single platform easily sees.

Connectivity, and with it routing, was the first frontier

Connectivity came first, and it is now largely settled. Every established global orchestrator now offers connections to hundred acquirers, payment methods, wallets, and fraud tools. This reach is built directly or typically through local partnerships. For e.g., a global orchestrator expanding into the GCC, will usually sign a single regional partner (e.g., Tap Payments), rather than connecting individually to each domestic card scheme and local acquirer. The same approach is common across emerging markets such as India, Latin America, and Southeast Asia, offering an efficient way to scale. However, as orchestrators rely on the same regional aggregators to enter new markets, their payment coverage increasingly converges, limiting differentiation.

Routing is where the language becomes less precise. Nearly every provider markets a smart routing engine, but what sits behind the term varies considerably from one platform to the next.

Rule-based routing is what most platforms run today. The merchant defines the logic and the engine executes it, so a transaction in euros on a debit card goes to acquirer A, or a transaction whose issuing BIN belongs to a bank the merchant holds an on-us agreement with is sent to that same acquirer (where the fees are lower). Rules of this kind can be highly granular and they work well, but they are static by nature while payment performance shifts constantly.

Dynamic routing is where the platform moves from executing a decision to making one, with AI models trained on live transaction data. It appears in two forms that differ on how much of the decision the platform holds. In the first, the rules remain the source of truth and the model works around them, reading live traffic metrics, identifying that an acquirer’s approval rate on a particular issuer has fallen, and putting that forward as a recommendation. The merchant payments team decides whether to act on it, so the model informs while the human still holds the decision.

In the second, the rules are replaced by constraints. For e.g., the merchant specifies a contractual minimum of forty percent of volume to one acquirer, or that European transactions stay within three named processors, and the model then decides each transaction itself inside those boundaries, weighing live parameters like approval rate, cost, and latency. Here the human sets the boundaries, and the model holds the decision.

The overall picture is reasonably clear. Connectivity is solved and every provider offers rule-based routing. A growing number are moving into dynamic routing, several have it on the roadmap, though only a handful can point to a live merchant case study today showing what it delivered.

Reconciliation was next, and it is mostly still open

The third promise was a consolidated view of payment operations, and most platforms now deliver it. What they deliver far less often is reconciliation, which is not the same thing. A consolidated view brings every transaction from every provider into one place, whereas reconciliation confirms that the money authorised actually arrived in the bank, net of fees, refunds, and chargebacks, and accounts for the difference when it did not.

The airline sector shows the gap most clearly because the chain is the longest. A payment has to be matched to a booking, then to the ticket issued against it, and finally to the money received, often across channels the payments team has limited visibility over. In most carriers that reconciliation is still a quarterly spreadsheet exercise.

Providers have begun to close the gap, in some cases by building reconciliation modules alongside a network airline that match payments back against booking data, and in others by developing the capability in house and selling it as a separate module rather than bundling it with orchestration.

The commercial significance is straightforward. Reconciliation is among the least visible things an orchestrator sells and increasingly among the most decisive, because the people who shape the outcome of an RFP now sit in finance as often as in payments.

It also points to where the market may fragment. Reconciliation is not the same problem across verticals. A marketplace matches it to an order, then to a seller payout and a commission split. A subscription business matches it to a billing cycle and a dunning attempt. The data model is different in each case, so this is rarely a single build. It is one reason vertical specialisation may hold up better here than elsewhere in the stack.

The commercial model is also shifting

Most independent orchestrators remain venture funded (recent 2026 examples - Juspay, USD 50m in January; APEXX, USD 10m in February; Primer, USD 100m in May; BR-DGE, GBP 10m in June), and profitability has not yet been widely demonstrated across the category. A few cents per transaction, on a service that is itself commoditising, is a getting to be narrow base for a business of scale.

So, the model is moving. Three directions are visible.

First, down into the stack. Some orchestrators are extending into gateway and acceptance capabilities, which captures more of the economics on each transaction. Juspay’s payment aggregator authorisation in India, is one example of that direction.

Second, sideways into value-added services. Tokenisation, 3DS, fraud, DCC/MCP and reconciliation are usually priced separately, not bundled. Alternative payment methods are more lucrative still, because the orchestrator can front the commercial relationship with the method and take a margin (esp. in BNPL), rather than passing the merchant through.

Third, and most interesting, toward paying for outcomes. A handful of providers now contract on a success fee alongside the per transaction rate, taking a share of the uplift they can quantify. What has held this back is attribution rather than appetite. A provider cannot bill for a three-point gain in acceptance unless it can show that the gain came from its own decisions, and few have built measurement the merchant can verify independently.

For merchants, they should also look closely at what counts as a billable event in during contracting. Some providers charge on every payment attempt/API call, which means declines, refunds, and each leg of a failover cascade are chargeable. Others charge only on a successful payment, at a higher unit rate. The two are not comparable without adjusting for the merchant’s own decline rate, and the gap between them widens exactly where performance is worst.

The intelligence layer is where the advantage now sits

Some orchestrators are repositioning as payment intelligence providers. CellPoint Digital announced USD 34 million of fresh investment in October 2026 to build Zenith, an AI intelligence platform sitting above its orchestration layer. DEUNA has rebuilt its product around Athia, which normalises data across every provider, answers questions in plain language through a LLM chat interface, and runs named AI agents covering acceptance, cost, reconciliation and disputes. Yuno describes itself as an AI-native intelligence system, with Payments Concierge and Nova, its voice agent. When a category’s own players begin to drop the category’s name, it usually signals that the value is moving.

The underlying logic holds. An orchestrator sees every transaction across every provider, from checkout through to settlement, which no single acquirer can. The advantage is not simply access to more transactions, since a large merchant can already consolidate its own data across providers. It lies in being able to compare that performance against a wider set of merchants and providers, identify patterns that would be difficult to detect within one merchant’s data alone, and feed those insights back into the decisions the platform makes. That is the real fuel, and the value it holds for merchants is only expected to grow.

Conclusion

Payment orchestration is becoming infrastructure, and infrastructure competes on reliability and price alone. That can make growth harder to sustain, encouraging providers to position themselves as more than orchestrators.

For merchants, the evaluation questions have changed accordingly. Not how many connections a provider offers, but how intelligently it routes transactions and how transparently it explains those decisions. Not whether there is a dashboard, but whether everything reconciles back to the bank. And not whether the intelligence layer is worth paying for, but whether a provider without one is worth shortlisting.

The providers best placed for the next phase are likely to be those moving up the stack, since the connection layer beneath them has already become standard.

About the authors

Louis Wapler is a Manager at the Edgar, Dunn & Company Paris office. He joined the firm in March 2021 and graduated with an MBA from ESCP Business School in 2020. Prior to joining EDC, Louis had combined seven years of experience in the banking and aviation industries. Louis enjoys witnessing the modernisation of payments and payment-related processes in B2B ecosystems that drive the ongoing change of the financial institutions’ landscape.

 

 

Rohan Shaju is a Consultant in the Paris office. Before joining Edgar, Dunn & Company (EDC), Rohan worked for three years in India and France across technology strategy and transformation projects. Since joining EDC in 2023, he has gained valuable payment strategy expertise working with global card networks, issuers, fintechs, and travel merchants. Rohan holds a Master's in Management from ESSEC Business School in Paris, alongside an Electronics Engineering degree from India. Outside the realm of consultancy, Rohan follows cricket, football, and Formula 1. Additionally, he finds solace in activities like cooking, yoga, and skateboarding.

 

About Edgar, Dunn & Company

Edgar, Dunn & Company (EDC) is an independent global payments consultancy. The company is widely regarded as a trusted adviser, providing a full range of strategy consulting services, expertise, and market insights. EDC’s expertise includes M&A due diligence, legal and regulatory support across the payment ecosystem, fintech, mobile payments, digitalisation of retail and corporate payments, and financial services.