I can’t speak to large payloads within individual requests, but find that Bubble is quite “expensive” at processing inbound webhooks in high volume. Smathering of thoughts below, but I do think that if you’re processing high amounts of data, Bubble may not be the most scalable solution.
My app syncs 100-200 JSON objects every day (syncing calendar data), using a hacky webhook relay service I built. This ends up costing about 1k-1.5k Workload Units each day, which is about $0.30 to $0.45 once above the plan’s monthly limit.
If you’re running very high volume but only need simple confirmations, I would consider whether you can store that data in a traditional SQL database and use the SQL Query plugin to access the data as needed. If the transactions can be settled in bulk outside of bubble, you also may be able to construct a CSV of that data and perform bulk uploads ([New Feature] Uploading as CSV in run mode - #9 by vascolucci), though I don’t know how this would compare in terms of WU consumption.
I would generally also recommend you consider whether you need to directly store each of these records or whether it is something that could be fetched at runtime as needed, or whether you might be able to build a workaround for the “inverse case” – would you be able to assume (within bubble) that the transaction was successful after a certain amount of time unless you receive a failure notice? This would dramatically decrease the load on bubble, even if just as a holdover.