Caller ID Reputation: Why Your Numbers Get Flagged and How to Fix It

Your answer rate dropped 20 percent overnight and nothing changed on your end. Same leads. Same dialer settings. Same agents. Same script. The only thing that changed is something you cannot see: the reputation score attached to your outbound phone numbers. Somewhere between your SIP trunk and the recipient’s handset, an analytics company decided your numbers look like spam, and now T-Mobile is displaying “Scam Likely” next to your business name on 40 percent of the phones you call.

This is the caller ID reputation problem, and it is the single largest uncontrolled variable in outbound calling operations. You can optimize your dialer, train your agents, buy the best leads in the industry, and lose it all because a third-party company you have never heard of flagged your numbers based on an algorithm you cannot see. Understanding how the system works is the first step toward managing it.

This guide is written by SIPNEX, an FCC-licensed carrier that holds its own STIR/SHAKEN Service Provider certificate. We provision DIDs, manage attestation, and work with operators daily on CID reputation strategy. What follows is the carrier-side view of how reputation scoring works, why your numbers get flagged, and what you can actually do about it.

What caller ID reputation is and why it exists

Phone carriers and analytics companies maintain reputation databases that score individual phone numbers based on their calling behavior. Every phone number that places outbound calls accumulates a reputation profile over time — similar in concept to a credit score but for phone numbers. A number with a good reputation passes through to the recipient’s handset normally. A number with a bad reputation gets labeled as “Scam Likely,” “Spam Risk,” “Potential Spam,” or similar warnings that cause the recipient to ignore the call.

The reputation ecosystem involves three layers:

Analytics companies are the primary scoring engines. The major players are Hiya (headquartered in Seattle, serves over 450 million users), Transaction Network Services (TNS, owned by Somos), First Orion (founded in Little Rock, partners with multiple carriers), and Nomorobo (primarily consumer-facing but contributes data to carrier filtering). Each company maintains its own proprietary database of phone number reputations, using its own algorithms and data sources.

Wireless carriers consume the analytics data and apply it to their call-filtering systems. T-Mobile uses Hiya’s data to power its Scam Shield service — when a call arrives for a T-Mobile subscriber, the network queries Hiya’s database and applies a label based on the reputation score. AT&T uses TNS data for its AT&T Call Protect service. Verizon uses a combination of sources for its Call Filter service. Each carrier may also apply its own internal scoring on top of the third-party data.

Consumer apps add another layer. Apps like Hiya (consumer version), Truecaller, Nomorobo, and RoboKiller maintain their own databases, often crowdsourced from user reports. When a user marks a call as spam in one of these apps, that data feeds back into the reputation ecosystem and may eventually influence the carrier-level scores.

The result is that there is no single “caller ID reputation score.” There are multiple competing scores from multiple companies, consumed by multiple carriers, supplemented by multiple consumer apps. A number can be clean on T-Mobile but flagged on AT&T. It can be fine for Verizon subscribers but blocked by Truecaller users. This fragmentation makes reputation management complex and explains why fixing a flagged number requires engaging with multiple parties.

Why your numbers are getting flagged

Understanding the triggering mechanisms helps you avoid them proactively rather than remediating reactively.

High call volume from a single number. This is the most common trigger. Analytics algorithms flag numbers that exhibit calling patterns inconsistent with human behavior. The threshold varies by provider, but general guidance is that more than 100 outbound calls per day from a single number attracts scrutiny, and more than 200 per day almost guarantees flagging. The algorithm does not know whether those calls are legitimate collections, insurance verifications, or scam attempts — it sees high volume and assigns negative weight.

Low answer rates. If 80 percent of your calls go unanswered, the analytics interpretation is that the called parties are intentionally avoiding your number — which correlates with unwanted calling. This is a frustrating catch-22: your answer rate is low because your number is getting flagged, and your number is getting flagged because your answer rate is low. Breaking this cycle requires proactive number management rather than reactive remediation.

Short call durations. A pattern of calls that last under 10 seconds — answered and immediately hung up — signals robocall behavior to analytics algorithms. Legitimate calls typically last 30 seconds to several minutes. If your operation generates a high percentage of very short calls (common with AMD false positives in predictive dialing, where the system connects an agent to what turns out to be a voicemail), those short-duration events accumulate against your number’s reputation.

Consumer complaints. Even a small number of spam reports from consumers can tank a number’s reputation. When a T-Mobile subscriber marks your call as spam through Scam Shield, that report feeds directly into Hiya’s database. Five to ten spam reports on a single number can trigger flagging. The consumer does not need to be right — the report alone is sufficient to influence the score. This is particularly problematic for operations calling consumers who did not expect the call or do not remember opting in.

B or C level attestation. This is the factor most operators overlook. Analytics companies now incorporate STIR/SHAKEN attestation level into their reputation scoring. A-level attestation (the carrier verified the caller’s authority over the number) is treated as a positive signal. B-level attestation (the carrier knows the customer but did not verify the specific number) is neutral at best and negative in some algorithms. C-level or unsigned calls receive significant negative weight. If your carrier provides B-level because it is a reseller inheriting attestation from upstream, every call you place starts with a reputation handicap that your calling behavior must overcome.

New numbers with sudden high volume. A brand-new DID that has never been used before and suddenly places 300 calls on its first day looks like a freshly provisioned robocall number. Analytics algorithms are tuned to detect this pattern because it is exactly what robocall operations do — provision disposable numbers, blast calls for a day or two, then discard the numbers. Legitimate operations that fail to warm up new numbers trigger the same detection.

Recycled numbers with bad history. Phone numbers are recycled. When your carrier provisions a DID to your account, that number may have been previously assigned to a robocall operation, a debt collector with aggressive practices, or any other entity that accumulated negative reputation. The reputation history travels with the number, not with the owner. You inherit whatever reputation the number carries, and you may not know about it until your answer rates tell you something is wrong.

How the analytics companies and carriers work together

Understanding the specific relationships helps you target your remediation efforts.

Hiya is the dominant player in the carrier reputation space. Hiya powers T-Mobile’s Scam Shield (the largest deployment in the US by subscriber count), Samsung’s Smart Call (embedded in Samsung Galaxy phones globally), and provides data to several other carriers and device manufacturers. If your number is flagged on T-Mobile, it is almost certainly a Hiya score. Hiya operates a business portal (hiya.com) where you can register your business, claim your numbers, and check their reputation status.

TNS (Transaction Network Services) powers AT&T’s Call Protect service and provides data to other carriers. TNS operates the TNS Call Guardian platform. If your numbers are flagged specifically on AT&T but clean on T-Mobile, TNS is likely the scoring engine. TNS has a caller registration portal for businesses.

First Orion has partnerships with multiple carriers and powers branded calling features for T-Mobile (in addition to Hiya’s spam scoring, T-Mobile uses First Orion for branded business caller ID). First Orion operates the First Orion Portal for business registration.

Free Caller Registry (freecallerregistry.com) is an industry initiative that allows businesses to register their identity and phone numbers across multiple analytics platforms simultaneously. It does not guarantee your numbers will not be flagged, but it establishes a legitimate business identity that analytics companies can reference when evaluating your numbers.

The key insight: there is no single authority to appeal to. If your number is flagged on T-Mobile, you need to work with Hiya. If it is flagged on AT&T, you need TNS. If it is flagged on Samsung devices regardless of carrier, it is Hiya again. If consumers are marking you in Truecaller, that is a separate database entirely. Comprehensive reputation management requires monitoring and engaging with multiple platforms simultaneously.

How to fix flagged numbers and restore reputation

When a number is already flagged, remediation is possible but not instant. Here is the step-by-step process.

Step 1: Identify which numbers are flagged and where. Do not guess. Check every outbound DID against multiple platforms. Use the Free Caller Registry to check across multiple databases. Use Hiya’s business portal to check T-Mobile/Samsung status. Test-call AT&T phones from your DIDs and note the display. Check Verizon phones. The flagging may be limited to one carrier — in that case, you know which analytics company to target.

Step 2: Register in the Free Caller Registry. Go to freecallerregistry.com and register your business with all associated phone numbers. This establishes a legitimate business identity in the ecosystem. It does not un-flag numbers, but it provides a foundation for remediation requests.

Step 3: Ensure A-level STIR/SHAKEN attestation. If your carrier provides B-level attestation because it is a reseller, switching to a direct carrier like SIPNEX that holds its own SP-KI certificate and signs at A-level is the single highest-impact change you can make for long-term reputation health. A-level attestation signals to every analytics company in the ecosystem that a licensed carrier verified your authority over the number. This does not instantly fix a flagged number, but it shifts the baseline reputation from negative to neutral for every call going forward.

Step 4: Pull flagged numbers from rotation. Do not continue dialing on flagged numbers while you remediate. Every additional call from a flagged number accumulates more negative reputation data. Replace the flagged numbers with fresh DIDs (properly warmed up) and reserve the flagged numbers for remediation processing.

Step 5: Submit remediation requests. File requests with each analytics platform where your numbers are flagged. Hiya, TNS, and First Orion all have business portals with remediation or dispute processes. Include your Free Caller Registry ID, your business details, the specific numbers affected, and evidence that you are a legitimate business (your website, your FCC carrier’s information, your STIR/SHAKEN attestation level). Remediation processing typically takes 5 to 15 business days per platform.

Step 6: Monitor continuously. Reputation is not a one-time fix. Check your numbers weekly — or daily if you are running high-volume campaigns. Track answer rates per DID in your dialer dashboard. A sudden drop on a specific number usually means it was flagged before the analytics platforms update their public-facing status tools. Pull numbers at the first sign of degradation, not after the damage is catastrophic.

Prevention: building a sustainable CID strategy

Remediation is expensive and slow. Prevention is cheaper and faster. Here is how to build a CID strategy that minimizes flagging risk from the start.

Maintain a DID pool with rotation. Do not burn a small number of DIDs at high volume. Spread your outbound volume across a pool large enough that no single number exceeds 50 to 80 calls per day. For a campaign generating 1,000 daily outbound calls, you need a minimum of 15 to 20 DIDs in rotation. For 5,000 calls, 60 to 100 numbers. The math is simple: total daily calls divided by 50 to 80 calls per number equals pool size.

Use local presence DIDs matching your calling geography. Numbers with area codes matching the recipient’s area code get answered more frequently, which improves answer rate metrics, which improves reputation scores. Local presence is both a direct answer-rate improvement and an indirect reputation protection.

Warm up every new number. Start new DIDs at 20 to 30 calls per day. Increase by 20 to 30 percent every two to three days. Reach full campaign volume after 10 to 14 days. This gradual ramp-up establishes a human-like calling pattern that analytics algorithms recognize as legitimate. Jumping from zero to 200 calls on day one triggers robocall detection.

Monitor answer rates per DID, not just per campaign. Your campaign dashboard shows an aggregate answer rate. That aggregate hides individual DIDs that are tanking. If your campaign answer rate is 15 percent but one DID in the pool has a 3 percent answer rate, that DID is flagged and dragging down the rest of your numbers by association (some analytics algorithms consider shared ownership patterns). Identify and replace underperforming DIDs weekly.

Get A-level STIR/SHAKEN from a direct carrier. We keep coming back to this because it is the single most impactful long-term reputation factor. B-level attestation from a reseller means every call starts at a disadvantage. A-level from a direct carrier means every call starts at neutral or positive. Over millions of calls, this baseline difference compounds into dramatically different reputation trajectories. SIPNEX signs at A-level for every DID we provision or verify through porting — no exceptions, no extra charge, no separate registration process. It is built into the trunk.

Retire numbers before they are flagged. The smartest operators track per-number answer rates over rolling 7-day windows. When a number’s answer rate drops below a threshold (typically 8 to 10 percent, or 50 percent below the campaign average), they pull it from rotation immediately — before analytics companies formally flag it. A number with declining answer rates is a number that is about to be flagged. Proactive retirement preserves the rest of your pool’s reputation.

Frequently asked questions

Why are my phone numbers showing as Scam Likely?

“Scam Likely” is T-Mobile’s label, powered by Hiya’s analytics engine. Your numbers are displaying this label because Hiya’s algorithm has assigned them a low reputation score based on one or more factors: high call volume from a single number (over 100 calls/day triggers scrutiny), low answer rates (indicating recipients are avoiding the number), consumer spam reports (even a few reports can trigger flagging), B-level or lower STIR/SHAKEN attestation, short call durations suggesting robocall patterns, or inherited bad reputation from a previous owner of a recycled number. The label appears for all T-Mobile subscribers and for Samsung phone users on any carrier. To resolve it, register with the Free Caller Registry, check your reputation on Hiya’s business portal, ensure A-level attestation from a direct carrier, and pull flagged numbers from rotation while filing remediation requests.

How do I check my caller ID reputation score?

There is no single reputation score — multiple companies maintain separate databases. Start with the Free Caller Registry (freecallerregistry.com) to check across multiple platforms. Use Hiya’s business portal to check T-Mobile and Samsung status. Use TNS’s caller registration portal for AT&T status. Test-call phones on each major carrier from your outbound DIDs and note what displays. Monitor your per-DID answer rates in your dialer dashboard — a sudden drop is often the first indicator of flagging, appearing before the analytics platforms update their public tools. For ongoing monitoring, establish a weekly check routine across all platforms and track answer rates daily per number.

Can I fix a phone number that has been flagged as spam?

Yes, but it takes time and effort. The process: pull the flagged number from outbound rotation immediately to stop accumulating negative data. Register with the Free Caller Registry. Submit remediation requests through the business portals of the analytics companies where the number is flagged (Hiya for T-Mobile/Samsung, TNS for AT&T, First Orion for branded calling). Include your business registration details and STIR/SHAKEN attestation information. Processing takes 5 to 15 business days per platform, and there is no guarantee of full reputation restoration — some numbers with severe negative history may never fully recover. In many cases, replacing a badly flagged number with a fresh DID (properly warmed up with A-level attestation) is faster and more reliable than remediation.

Does STIR/SHAKEN attestation level affect caller ID reputation?

Yes, significantly. Analytics companies including Hiya and TNS now incorporate STIR/SHAKEN attestation level into their reputation scoring algorithms. A-level attestation (full verification — the carrier confirmed the caller’s authority over the number) is treated as a positive signal that reduces the likelihood of flagging. B-level attestation (partial — the carrier knows the customer but did not verify the specific number) is neutral at best and can contribute to negative scoring when combined with other risk factors. C-level or unsigned calls receive significant negative weight and are much more likely to be flagged. This means that using a reseller trunk with B-level attestation creates a structural reputation disadvantage on every outbound call, regardless of your calling behavior. Switching to a direct carrier with A-level attestation is the most impactful long-term reputation improvement an operator can make.

How many calls per day can I make from one number before it gets flagged?

There is no published threshold because the analytics algorithms consider multiple factors, not just volume. However, industry experience and conversations with analytics providers suggest that exceeding 100 outbound calls per day from a single number attracts algorithmic scrutiny, and exceeding 200 per day makes flagging highly likely regardless of other factors. For safe operation, keep per-number daily volume between 50 and 80 calls. This means a campaign generating 1,000 calls per day needs a minimum pool of 15 to 20 rotating DIDs, and a campaign at 5,000 calls per day needs 60 to 100 numbers. Combine volume limits with gradual warm-up for new numbers, local presence matching, and A-level STIR/SHAKEN attestation for the best reputation protection.

---

SIPNEX is an FCC-licensed carrier that provisions DIDs with A-level STIR/SHAKEN attestation from day one. We help operators build sustainable CID strategies — proper pool sizing, local presence provisioning, warm-up guidance, and the A-level attestation foundation that keeps your numbers clean. Talk to an operator about CID strategy or see our rates.