Skip to main content
MADIAD
← Back to Insights

Content & social media

How to resend only failed channels on MADIAD Hub

A multi-channel post can fail on just a few platforms, and resending everything duplicates the posts that already went live. MADIAD Hub resends only the failed channels through a dashboard button or the retry API, without adding another publish count.

M
MADIAD
7 min read
Cách gửi lại kênh lỗi trên MADIAD Hub

A post sent to several platforms at once usually lands on most of them right away, with only a few channels rejecting it for their own reasons: this happens to any business posting across multiple channels every day. The problem is not that a failure happens, since a platform-specific rejection is normal, it's what happens next: resending the whole post creates a duplicate on every channel that already succeeded. MADIAD addresses exactly this with an integrated system that tracks each channel's status separately within the same post, so it resends only the part that failed. The sections below walk through the mechanism and how to call the resend correctly.

Why a multi-channel post only fails on a few platforms

A post sent to several platforms is not one single action: the system splits it into separate deliveries, one per channel. That is why Instagram can accept a post at the exact moment Mastodon rejects it. Each platform sets its own character limit, file format rule, and permission scope for the connected account. One Mastodon server might allow only 500 characters, so a 724-character caption gets rejected outright. TikTok caps how many posts an account can publish directly per day; hit that cap and the post is not rejected, it lands in the app's draft inbox waiting for a person to tap publish. When some channels already succeeded and a few came back with errors, resending the whole post through the original publish endpoint creates an entirely new post. The channels that already succeeded get a duplicate, while the channels that failed may fail again for the same unresolved reason. MADIAD tracks each channel's status separately within the same post through an integrated system, so it knows exactly which channel needs a resend and which one should stay untouched. The next section covers how to call that resend correctly.

How MADIAD Hub resends only the failed channels

There are two ways to resend only the failed channels: a dashboard button or a direct API call. On the Overview page, every recent post shows a Resend failed channels button next to its result; tapping it resends only the channels that were rejected, leaving the channels that already published untouched. For automated flows or code integrations, call POST /v1/posts/retry with exactly one of request_id or job_id, taken from the original publish response or from the post history. The response returns a legs array, because a post whose caption got split per platform is actually several smaller jobs underneath, and retry re-attempts each one individually. A channel that already succeeded replies that there is nothing to retry: that is a normal response, not an error to handle. One key detail: calling retry does not add another publish count, since that count was already deducted when the post was first accepted, and an unsuccessful delivery is not a new delivery. See more technical detail on the Products page for MADIAD. The next section explains why this mechanism is built the way it is.

How MADIAD works differently

MADIAD's resend works differently because it separates the failed portion from the portion that already ran, through an integrated system that tracks each channel instead of treating the whole post as one block. When retry runs, it sends only the channels that failed; channels that already published are never touched, removing the risk of creating a duplicate where a post already succeeded. For a post whose caption got split per platform, for example trimmed to fit different character limits, the legs array in the response tracks each of those smaller jobs separately, so you know exactly which part is done and which part still needs a retry. Another design choice: the system branches on specific HTTP error codes (400, 409, 429, 5xx) instead of treating every non-success response as worth retrying. This avoids retrying forever on errors that never resolve themselves, like malformed content or an account that was never connected. Every retry call also carries an Idempotency Key, so calling retry more than once due to a slow network or a timeout does not accidentally create a new post. MADIAD keeps these mechanisms updated whenever a platform changes its limits or adds a new error code. The next section covers exactly when to call retry and when to fix the content first.

When to resend right away, when to fix the content first

Call retry right away when a post's status is partial: some channels already published and only a portion of the remaining channels were rejected, so only that portion needs another attempt. When the status is failed, meaning every channel rejected the post and nothing went live, fix the content first and then resend the whole post, because resending the same content just repeats the same error. Specific error codes call for different handling. A 400 means the content or request itself is wrong, for example a caption over that platform's character limit; retrying the exact same content cannot succeed, so fix it before resending. A 409 with platform_not_connected means the account for that platform was never connected in that Profile; retry does not fix this, the account needs to be reconnected first. In contrast, a 429 means a rate limit was hit and a 5xx means a temporary failure on the publishing side; both cases call for waiting the suggested amount of time before calling again, usually with an Idempotency Key to avoid creating a duplicate post. Telling these error groups apart saves time spent on retries that cannot succeed and prevents editing content that was already correct. This branching logic already lives inside MADIAD's processing system, so an integration flow only needs to read the returned error code correctly. The final section points to where to learn more about this flow.

Get started with MADIAD

Start by reading the technical documentation before wiring an automated flow into production. The posting and retry API docs cover the full request structure, the statuses you may encounter, and how to read the per-platform results array, and reading them first avoids most of the common integration mistakes. If a business is already posting to several channels daily and needs a clear way to handle failures instead of resending the whole post every time one channel breaks, this is the point to look closely at how MADIAD Hub organizes retry per channel. MADIAD works directly with each business to design a posting flow fitted to its channel count, posting frequency, and existing tools, rather than forcing one template on every case. Contact MADIAD to talk through your multi-channel posting flow, or learn more about MADIAD Hub before deciding.

Conclusion

Resending failed channels the right way avoids duplicate posts and cuts the manual cleanup after a multi-channel post goes out. Try this week: (1) check the Result column on recent posts for any channel that failed before; (2) click Resend failed channels once just to get familiar with the flow; (3) if using n8n, review the Retry On Fail setup to see if it branches by error code yet. Learn more about MADIAD Hub or contact MADIAD for advice on a multi-channel posting flow fitted to your business.

Frequently asked questions

Does MADIAD Hub work for a small business posting across several social channels?+

Yes, MADIAD Hub fits once a business is posting to several channels and needs to track publishing results per platform separately. Resending only the failed channel, instead of the whole post, matters more as the channel count grows, since the chance of a post failing in just one or two places rises along with it. MADIAD designs the channel setup and posting flow around each business's actual platform count and posting frequency, rather than applying one fixed template. Contact MADIAD to talk through the channels you currently use.

Does MADIAD support integrating with n8n to automate resending failed channels?+

Yes, MADIAD Hub exposes a POST /v1/posts/retry endpoint that an n8n workflow can call directly through an HTTP Request node. Since n8n treats every non-2xx response as worth retrying by default, the error-reading option (Never Error or Continue using error output) needs to be enabled so the workflow branches on the HTTP error code itself, instead of letting the node retry endlessly on posts that can never succeed. Each retry call in the workflow should carry an Idempotency Key so a slow network does not create a duplicate post. Check the n8n integration docs for the exact node setup, or contact MADIAD for help building this flow.

What determines the cost of MADIAD setting up a multi-channel posting flow for a business?+

The cost of MADIAD setting up a multi-channel posting flow depends on how many channels need connecting, how often posts go out each day, and how much automation the failure-handling flow needs, whether that is retry through n8n or custom code. A business posting manually through the dashboard has different needs from one that requires an automated flow tied into internal systems. MADIAD designs each flow around these specific factors rather than applying one fixed rate to every case. See details at the pricing page: https://madiad.com/en/pricing.

After deployment, what does MADIAD support, and what can the business handle on its own?+

MADIAD supports the initial channel connection setup on MADIAD Hub and walks the technical or content operations team through reading post statuses (partial, failed, draft) and branching logic by HTTP error code. When a platform changes its rules, for example a new character limit or a new error code, MADIAD updates the documentation and the sample flow accordingly. After that stage, the business runs day-to-day posting on its own and clicks resend failed channels on the dashboard whenever needed, without MADIAD stepping in on every post. Contact MADIAD if the flow needs rebuilding after a major change in channel count.

Contact MADIAD

Talk to us about your AI marketing strategy