Question 1

How does it get into a private Facebook group?

Accepted Answer

Through accounts GrapeWire operates, not through yours. You never connect a Facebook account — there is no field, no form and no setting for one, and there will not be. You give us the group's URL; an account we run reads that group's feed, and where it is not already a member it queues a join request at a deliberately slow rate. Your own account is never used, never asked for, and never at risk.

Question 2

How fast does an alert actually reach me?

Accepted Answer

Two numbers, and we have measured one of them. Waiting for the next poll is the slow part: each Reddit source is polled every 420 seconds, each Facebook source every 36–54 seconds. The work after that — fetch, score, deliver — measured 170–245 ms end to end, but against a local test server rather than Facebook itself, so treat it as a floor and not a promise. We have not published a live end-to-end figure because we do not have one yet. Your own median lands on your Analytics page, computed from your deliveries, and that number is real.

Question 3

What happens when I hit my plan's group limit?

Accepted Answer

Nothing breaks — you just can't add another until you free a slot. Pause or replace any source from the dashboard and the new one starts on the next cycle. Groups are meant to be swapped as rooms go quiet.

Question 4

Will I get banned for using this?

Accepted Answer

Your account does nothing at all — it is never connected, so it cannot be flagged for anything GrapeWire does. Nothing is posted anywhere from your account or ours: no comments, no likes, no votes. The reply you send is yours, sent by you, from your own account, whenever you decide to send it. There are two automated actions in the product, both performed by operator accounts we run and never by yours: asking to join a Facebook group, and sending one message to a private subreddit's moderators asking for access. Both are paced at a few requests a day. The Reddit message says plainly that it was sent by a program, who operates the account, and what it would read — a moderator gets every fact they need to say no.

Question 5

How is this different from a keyword alert?

Accepted Answer

A keyword alert fires on a word. GrapeWire judges what the person wants. "I hate being a bookkeeper" and "can anyone recommend a bookkeeper" contain the same word and land at opposite ends of the 0–100 scale. You set the minimum score and only see what clears it — and you can check any post against a watch in the Playground before you trust it.

Question 6

What if the AI scores something wrong?

Accepted Answer

Every match shows its reason, so you can see the mistake instead of guessing at it. Mark it Ignored and it leaves your working list — nothing is learned from that, and we would rather say so than imply a model that trains on your clicks. The fix is the Playground: paste the post, adjust the watch's plain-English description and its threshold until it scores the way you'd score it.

Question 7

Which platforms are supported?

Accepted Answer

Facebook groups and Reddit. Facebook pages are not supported yet — the platform is in the product but nothing reads a real page behind it, so it is on the roadmap rather than in the plan. Alerts go to Slack, Discord, email, webhook, ntfy, Google Chat, and Microsoft Teams. The webhook payload carries the post text, source name, direct link, score, reason, and timestamp, so you can route it anywhere else yourself.

Question 8

Can I cancel, and do I get a refund?

Accepted Answer

Cancel anytime from your dashboard; you keep access until the end of the paid month. Every plan starts with a 7-day free trial, and if you're inside your first 7 days after that, request a full refund and we send it back — no lock-in contract, no retention call.

Question 9

What if Meta restricts one of the accounts you read groups with?

Accepted Answer

GrapeWire reads Facebook groups through accounts we operate, not through yours — we never ask for your social media login and there is no place in the product to enter one. Those accounts read groups the way any member does. We do not have Meta's written permission for this, so a source can stop working at any time if an account is restricted; when that happens you will see the source marked as failed, with the reason. We do not use your account for anything, ever, so nothing that happens to ours can affect yours.

Frequently Asked Questions