Customer Support

How to Reply When the Bug Is in a Third-Party Service

A third-party bug is still your customer's problem. Here's how to confirm the cause, take responsibility for the impact, explain it plainly, offer workarounds, and keep customers updated until the fix arrives.

SupportMe••10 min read

When a bug comes from a service you depend on, like a payment processor, an email provider, a cloud host or an app store, your reply should still sound like you're responsible. Take ownership of what the customer is going through. Explain the cause in plain language. Say what you're doing about it, offer a workaround if there is one, and tell them when they'll hear from you next. Don't make "it's not our fault" the main point.

The rest of this post covers how to check that the bug really is a third-party problem, how much to say about the vendor, a reply structure you can reuse, and what to set up before the next outage.

Why "it's their bug" doesn't work as a reply

Your customer didn't sign up with your vendor. They signed up with you. Atlassian's Statuspage documentation says it directly: "While an incident may technically be caused by another provider, in your customers' eyes, it's a problem with your service." The same guide advises you to "own the problem and show understanding for how your customers may feel," and to "show empathy, and apologize when necessary."

Owning the problem doesn't mean taking the blame for someone else's code. It means owning the customer's experience: their checkout failed, their export never arrived, their login is broken. All of that happened inside your product, and you're the one who can help them.

Step 1: Confirm it's really a third-party issue

Check before you name a vendor. If you blame a provider and later find the bug was in your own integration, you lose credibility with the customer and probably with the vendor too.

A quick check:

  • Look at the vendor's status page. Most major providers publish one. Stripe, for example, posts incidents at status.stripe.com. AWS has a public Service health page that you can view without signing in. It only shows public events, though, not issues specific to your account.
  • Check your own logs and error responses. A timeout, a 5xx response or a specific error code from the vendor's API is good evidence. A vague failure on your side is not.
  • Try to reproduce it. If you need more detail from the customer first, these ways to ask for better bug reproduction steps can help.
  • Look for a pattern. Several reports at once that all touch the same integration point to a dependency.

A status page that says nothing is wrong doesn't prove the vendor is fine. It may just mean they haven't posted yet, or that the problem only affects some accounts. If your evidence points to the vendor but they haven't acknowledged anything, say you're investigating instead of naming them.

Step 2: Decide how much to say about the vendor

This is a judgment call, not a rule. Some practical guidelines:

Name the vendor when:

  • The vendor has publicly confirmed the incident (for example, on its status page).
  • Knowing the cause helps the customer, for example so they know to check their bank or card provider rather than your app.
  • The customer can follow the vendor's status page themselves.

Describe it generically ("our payment provider," "our email delivery service") when:

  • You're fairly sure but the vendor hasn't confirmed anything.
  • Naming the vendor wouldn't change what the customer does.
  • Your contract with the vendor limits what you can share.

Don't share:

  • Private details from the vendor's support conversations with you.
  • Guesses about their root cause or fix timeline.
  • Anything that comes across as venting about the vendor.

Some cases send the customer to the third party directly. Apple's guidance on responding to App Store reviews says that if a reviewer reports a downloading error or billing issue, you should direct them to Apple Support. Even then, explain why you're pointing them there so it doesn't feel like a brush-off.

Step 3: Use a reply structure that covers what customers need

A good third-party bug reply answers five questions:

  1. Do you understand what happened to me? Acknowledge the specific problem and its impact.
  2. Why is it happening? One or two sentences, without jargon.
  3. What are you doing about it? For example: reported it to the vendor, added monitoring, paused a feature, or retrying failed jobs.
  4. What can I do right now? A workaround if one exists. If there isn't one, say so honestly.
  5. When will I hear from you? A specific next step you can actually deliver.

Atlassian's incident guidance recommends updates on a regular cadence ("every 30 minutes (or whatever cadence is appropriate for the situation)") so users aren't left in the dark until the problem is resolved (source). For a single support ticket, a cadence that frequent is often unnecessary. A promise like "I'll email you as soon as it's resolved" works well, as long as you keep it.

Hypothetical example: payment provider outage

Hi Sam,

>

Sorry about this. You tried to upgrade your plan and the payment didn't go through, so you're still on the free tier when you meant to be using the Pro features today.

>

The cause is an ongoing issue with our payment provider, Stripe, which is affecting some card payments right now. They've confirmed it on their status page. Your card wasn't charged.

>

In the meantime, I've turned on Pro for your account manually, so you can keep working. Once payments are working again, I'll send you a link to finish the upgrade. You won't need to do anything before then.

>

I'll email you as soon as it's resolved.

>

Thanks for your patience,
Alex

This is an illustrative example, not a real exchange. Notice that the reply leads with the impact, names the vendor only because the vendor has confirmed the incident, removes the customer's immediate blocker, and ends with a clear next step. The customer has nothing to chase.

Hypothetical example: cause suspected but not confirmed

Hi Priya,

>

Thanks for flagging this. You're right that the password reset emails aren't arriving, and that's blocking you from getting into your account.

>

From what we can see, the emails are leaving our system but not being delivered. That points to our email delivery provider, and we've reported it to them with the details. We don't have a confirmed cause yet.

>

While we wait, I've sent a one-time login link to the address on your account using a different sending method. If that doesn't arrive within a few minutes, reply here and I'll verify your account another way.

>

I'll follow up when we hear back from the provider.

>

Best,
Alex

Here the vendor isn't named, because the cause is suspected but not confirmed. The reply is honest about what's known and what isn't.

Step 4: What to avoid

  • Leading with blame. "This is a Stripe issue, not ours" may be accurate, but it tells the customer you care more about being in the clear than about helping them.
  • Passing on the vendor's ETA as a promise. If the vendor estimates a fix by 3 p.m. and misses it, your customer remembers that you missed it. Say "the provider expects…" or skip the estimate.
  • Telling the customer to contact the vendor when they don't have a relationship with them. Only redirect when the third party really is the right contact, as in Apple's billing example above.
  • Over-technical explanations. "Upstream 502s from the webhook endpoint" means nothing to most customers. "Our payment provider is having trouble confirming payments" does.
  • Going quiet. Atlassian's team playbook gives an example of a common failure: not updating customers for over four hours because there was no update from the dev team. "No news yet, still on it" is a perfectly good update.

For more on tone, see these ways to make bug report replies more reassuring.

Step 5: Report it to the vendor and track who's affected

When a vendor is down, you're both a support agent and a customer. Report the problem to the vendor with timestamps, request IDs, error messages and the number of affected users. This is the same kind of clear report you'd want from your own customers.

On your side, keep a list of everyone who's been affected. If several tickets come from the same vendor problem, consolidate them into one bug issue so you can update everyone at once. When the vendor fixes it, close the loop with each affected customer. This workflow for notifying customers when fixes ship works for third-party fixes too.

Step 6: Handle public replies more carefully

Public replies, like app store review responses, community forum posts or social media, can be read by anyone. That makes blaming the vendor look worse and makes accuracy more important. Apple's App Store Connect documentation notes that only one response per review appears on your product page, that responses may take up to 24 hours to appear, and that the reviewer is notified and can update their review.

Because of that delay, an update like "it's fixed now" may not be true by the time people read it. Write public replies that will still make sense in a day: acknowledge the issue, give a short cause, offer a way to get direct help, and edit the response once the issue is resolved.

Prepare before the next outage

Most of the work can happen before anything breaks:

If you use an AI drafting tool, read every draft carefully during a vendor incident. Details like whether the vendor has confirmed the issue or whether a workaround is safe change from hour to hour. SupportMe, for example, is built so nothing sends without your approval, which leaves you room to check those details before a reply goes out.

Conclusion

A third-party bug changes who fixes the problem, not who looks after the customer. Confirm the cause before you name anyone. Lead with the customer's impact, explain the cause in plain words, and only name the vendor when it's confirmed and useful. Offer a workaround where you can, promise an update you can keep, and follow up with everyone once the fix lands. Templates, status subscriptions and known workarounds make all of this much easier during the next outage.

References

Tags

third-party service bugcustomer support replyvendor outage communicationincident communicationbug report replystatus pageSaaS customer support

Related posts