ZeroBounce says its newest catch-all email verification update cut wrong guesses on Microsoft 365 addresses by over 99.98 percent. That’s real progress on one of email’s oldest headaches. But a day of Microsoft 365 downtime on November 12 showed a flawlessly verified list can still take a reputation hit when the mail infrastructure itself breaks.
- ZeroBounce’s Microsoft 365 catch-all update, announced in July 2026, claims a 99.98 percent drop in wrong catch-all guesses, based on the company’s own internal testing, not an independent audit.
- Catch-all domains still make up more than 9 percent of the 11 billion-plus addresses ZeroBounce checked in 2025, and the company says about 23 percent of a list decays within a year regardless of how clean it started.
- Microsoft 365’s November 12 outage left Outlook and Exchange Online digging out from retried messages for hours to days, a reminder that reputation depends on infrastructure nobody on your team controls.
- Truelist, a rival verifier, makes a similar claim (up to 2x more accurate catch-all detection). That says more about the category than either vendor: everyone’s racing to fix the same blind spot at once.
What Did ZeroBounce Actually Fix?
Catch-all is the setting a mail server uses when it accepts every address thrown at it, real or not, instead of bouncing the fake ones up front. Verification tools used to hit a catch-all domain and just shrug: unknown, could go either way.
ZeroBounce’s July 2026 update targets that exact gap on Microsoft 365 and Google Workspace domains. Its internal testing claims the update cut wrong catch-all calls on Microsoft 365 by more than 99.98 percent. ZeroBounce calls catch-all domains “one of the biggest sources of uncertainty in email validation,” and it’s hard to argue otherwise.
Keep the number in perspective: it’s the vendor’s own testing, not an independent benchmark, and it only covers Microsoft 365.
Why Does Catch-All Email Verification Still Trip Up Teams?
Because the server won’t tell you the truth. A catch-all domain says yes to every address during an SMTP check, so the usual ping-the-server method can’t tell a real inbox from a typo.
ZeroBounce’s own analysis of over 11 billion addresses checked in 2025 found more than 9 percent were catch-all, and over 1 billion of the addresses it verified fell into that bucket. The same analysis found at least 23 percent of a typical list decays within a single year.
Verification accuracy and list decay are two different problems, and fixing one doesn’t touch the other. You can resolve every catch-all address perfectly today and still be sending to a quarter dead contacts by next November.
Can an Outage Hurt Your Sender Reputation Even With a Clean List?
Yes, and that’s the part most deliverability advice skips. Microsoft 365 and Outlook went down on November 12, 2026, disrupting Outlook, Exchange Online, Teams, and Copilot for businesses, governments, and schools.
When the service came back, Microsoft and every other mail server on the internet faced what one report called a tidal wave of pent-up messages and retry attempts. Clearing the backlog took anywhere from hours to days, depending on the sender.
If your queue backed up and started retrying hard during that window, some receiving servers could read the spike as suspicious, no matter how clean your list was going in. We covered a related mail problem in our breakdown of the Outlook DMARC rejection bug, and the pattern’s the same: the failure sits upstream of anything your verification vendor can touch.
| Number | What it measures | Source |
|---|---|---|
| 9%+ | Share of checked addresses that are catch-all domains | ZeroBounce, 2025 data |
| 23%+ | Share of a typical list that decays within one year | ZeroBounce, 2025 data |
| 99.98% | Claimed drop in wrong Microsoft 365 catch-all guesses | ZeroBounce internal testing, July 2026 |
| Hours to days | Time to clear the backlog after the outage | Reporting on the Nov. 12, 2026 outage |
What Should You Actually Do About It?
None of this means verification’s a waste of money. It just means it’s solving a narrower problem than most teams think.
- Re-verify your list on a schedule, not once. A list that passed in January can already be a quarter stale by November.
- Treat catch-all results as a confidence score, not a flat yes or no. A tool that forces every address into “valid” or “invalid” is hiding the real uncertainty from you.
- Plan for provider outages separately from data quality. They’re not the same risk, and a clean list won’t save you from a backlog-driven reputation dip.
- Watch your own send volume and retry behavior during any major outage. Pulling back instead of hammering retry is the simplest thing you control.
- Ask any vendor which providers its catch-all claims actually cover. A Microsoft 365 number tells you nothing about Yahoo or a corporate gateway.
If you want to see how a smaller player approaches the same problem, EmailAddress.ai, a related company of ours, resolves catch-all domains through deep SMTP analysis and what it calls active inbox signals, with separate gateway detection for things like Mimecast and Proofpoint, and scores each address from 0 to 100 instead of forcing a flat answer. It’s one more sign of where the verification category is spending its effort this year.
Our look at falling email response rates found that warm, well-maintained lists already beat cold ones by a wide margin. Catch-all accuracy and outage resilience are just two more variables stacked on that same advantage.
Frequently Asked Questions
Is a 99.98 percent accuracy claim realistic? It’s plausible for the specific case ZeroBounce tested, Microsoft 365 catch-all domains, under its own internal testing. Treat it as a strong directional claim, not an independently audited industry number.
What is a catch-all email domain? It’s a mail server configured to accept messages sent to any address, including ones that don’t exist, instead of bouncing them right away. That makes a simple SMTP check useless for telling real inboxes from typos.
Does list decay happen even to verified contacts? Yes. People change jobs, companies merge domains, and inboxes go stale. ZeroBounce’s own data puts typical annual decay at 23 percent or more, which is why a one-time pass goes out of date fast.
Should I pause sending during a major provider outage? Slowing your retries down is safer than hammering a recovering server. Aggressive retries during a known outage tend to look like abuse, not patient delivery.
Sources: Fortune’s reporting on the November 2026 Microsoft 365 and Outlook outage; ZeroBounce’s analysis of catch-all rates and list decay and its July 2026 announcement of Microsoft 365 catch-all improvements; and Truelist’s Enhanced Validation claims for catch-all domains.