💡🧩 A few months ago, in the aftermath of major PA's: Razorpay, Cashfree, & PhonePe pulling away from Juspay's payment orchestrator, the Hyperswitch, and preferring to pursue direct integrations with the merchant, I had written about what the future of payment orchestration would look like. And my view was:
1️⃣ It'll have to be self hosted to remove the "possibility of biased routing"
2️⃣ This will impact smaller merchants, which prefer a standardized, shared cloud type model, which allows faster go-live. Bigger merchants (Amazon, Flipkart) would be the ones to build out self hosting - if they haven't already.
3️⃣ Hyperswitch is not the only orchestrator in the market: each PA has its own. (Razorpay -> Optimizer, Cashfree -> Flowwise, PayU -> Maximizer, Paytm -> AI Router). So in the short term, there are "plug & play options". But eventually as each of these scale, if what has happened with Juspay plays out, then every PA will choose to be only on its own orchestrator. So self hosting may become inevitable.
❗And this week, two new pieces of news dropped:
✅ Paytm is following the footsteps of its compatriots, and going the direct integration route.
✅ Juspay announced that its Hyperswitch is now open source, and merchants can self host.
💡🧩 So it does seem that eventually the "to each his own orchestrator" strategy will play out, leaving the smaller merchant high & dry. While larger merchants will self host their own orchestration solutions, smaller merchants will struggle.
So, in line with my earlier thoughts, this is how I see this playing out further:
💡🔄 There could be smaller, "orchestrator only" players which don't get into the PA side of things. This may be unlikely, margins on orchestrators are low (atleast on domestic payments), so you need to diversify revenues, and the payments infra side of things is getting saturated in any case.
💡🔄 For smaller (and smaller here is relative) merchants, operating only domestically, the effort required to self host vs the benefit of orchestration may not be worth it. Its possible that a "merchant owned, TSP managed" type model may work (and is cheaper than a self hosted model). It's still deployed in merchant infra, but in a “managed services model.” The cost to merchant, operational overhead, and if this is even something the TSP wants to get into will have to be figured out. But TLDR: the smaller merchants will take the hit over these “orchestrator wars”
💡🔄 There could be an opportunity to build region specific / global orchestrators as merchants expand into specific corridors. This is something I talked about previously and is another piece of news that Juspay announced - it's expanding Hyperswitch to the US and Europe.
🧠 For the deep dive, check the link in the comments
Note: All views are my own, and do not reflect the views of my employer
This post was originally shared by on Linkedin.