Relayo vs building it yourself
Last updated: 2026-10-09
This one is genuinely close, and plenty of capable teams should build it. The first version is small: an endpoint that answers the handshake, checks a signature, and POSTs the body to two places. A developer can ship that in a day.
The cost is in the second year, not the first week. What follows is the list we worked through building Relayo, roughly in the order each one bit us.
Side by side
| Relayo | a DIY relay | |
|---|---|---|
| First version | Minutes to configure | About a day to write |
| Retries with backoff and dead letters | Built in | Queue, worker, backoff policy, storage |
| Duplicate suppression | Per source, 24-hour window | Meta can resend for 7 days — you dedupe |
| Batch splitting and routing | Built in | Parse envelopes, resolve Page / WABA / number |
| Re-signing per receiver | Built in | Per-tool signing keys and rotation |
| SSRF protection on delivery | DNS-pinned, private ranges blocked | Easy to get wrong, hard to notice |
| Searchable log and replay | Included | A small product of its own |
| Who gets paged at 2am | Us | You |
Pick them if a DIY relay
- Routing webhooks is close enough to your core product to be worth owning.
- You already run queues, workers and on-call, so the marginal cost is low.
- Your requirements are unusual enough that a hosted product would fight you.
Pick Relayo if
- Your product is the store, not the plumbing.
- You would rather spend the week on the thing customers pay you for.
- $19 a month is cheaper than the first incident.
Two hubs, one $19 plan, 7-day free trial.
Join the waitlist