---
title: "How to Get a Refund for Bad B2B Contact Data [2026 Guide]"
description: "How to get a refund for bad B2B contact data: measure hard bounces, meet the notice window, and compare against 0.55% across 37,040 mailed contacts."
canonical: "https://leadsonar.io/blog/how-to-get-a-refund-for-bad-b2b-contact-data"
last-updated: "2026-11-06"
---

> How to get a refund for bad B2B contact data: measure hard bounces, meet the notice window, and compare against 0.55% across 37,040 mailed contacts.

LeadSonar [LeadSonar](https://leadsonar.io/)

Blog

# How to Get a Refund for Bad B2B Contact Data [2026 Guide]

November 6, 2026By [Sabo Nagy](https://leadsonar.io/authors/sabo-nagy)

[Sabo Nagy](https://leadsonar.io/authors/sabo-nagy)

Founder & CEO

How to get a refund for bad B2B contact data comes down to a dated paper trail. Find the remedy clause in your contract and measure hard bounces from a small sample with SMTP logs. Then send a written claim with a count and a deadline before the renewal window closes. I'm Sabo Nagy, founder and CEO of LeadSonar, a lead intelligence platform that cascades 18-20 data providers and bills for the contacts the waterfall finds. We built the billing that way so a buyer has less to claim back.

Refund outcomes depend on your contract and your jurisdiction, and I write as a data vendor's founder, so take a live dispute to counsel.

As of Q4 2026, LeadSonar's reference bar for contact quality is 204 hard bounces across 37,040 mailed US outbound contacts, a 0.55% hard-bounce rate on a scale where under 5% is healthy.

## What counts as bad data in a refund claim

As of Q4 2026, LeadSonar's mailed reference sample of 37,040 US outbound contacts produced 204 hard bounces, led by Google's NoSuchUser and DisabledUser reply codes.

A refund claim needs defects you can count, because "the list is bad" gives a vendor nothing to approve. We sort defects by how hard they are to dispute. The ranking in this table is our own read from running enrichment at volume, and your contract's wording can move any row up or down.

| Defect | Evidence to keep | Our read on claim strength |
|---|---|---|
| Mailbox does not exist | SMTP log showing 550 5.1.1, with send date | Strong if mailed soon after delivery |
| Mailbox inactive | SMTP log showing 550 5.2.1, with send date | Strong if mailed soon after delivery |
| Blank field you were charged for | Delivered file with the empty column counted | Strong under a per-field charge |
| Domain has no mail server | MX lookup results, dated | Strong and quick to verify |
| Duplicate records | Dedupe count against the delivered file | Medium, depends on how the order counts records |
| Consumer-domain address on a work list | Count of gmail, yahoo, hotmail, outlook and aol addresses | Medium, depends on what the order specified |
| Wrong person or stale title | Sample of manual checks against current company pages | Weak alone, because a job change after delivery looks like decay |
| Soft bounce | Keep it out of the claim | A temporary failure proves little about the record |

Google's [Gmail SMTP errors and codes](https://knowledge.workspace.google.com/admin/support/troubleshooting/gmail-smtp-errors-and-codes?hl=en) reference gives you neutral wording for the first two rows. It says 550 5.1.1 means the account you tried to reach does not exist, and 550 5.2.1 means the account is inactive. We saw the same two codes dominate the 204 hard bounces in our 37,040-contact US outbound sample.

Use our samples as reference points for two more rows. Of those 37,040 contacts, only 6 lacked a valid MX record, so 99.98% had one. In a separate 103,000-contact B2B file, 0% of addresses sat on free consumer domains. A purchased work list with a large consumer-domain share differs measurably from those samples, and you can state that difference as a count.

Timing is the weak point in several of these rows. Expect the vendor to answer that the contact was correct on the day of delivery, so keep the send date close to the delivery date. Our guide to [B2B contact data quality](https://leadsonar.io/blog/b2b-contact-data-accuracy/) covers how to judge a file's accuracy before you dispute it.

## How to get a refund for bad B2B contact data in seven steps

As of Q4 2026, LeadSonar treats a 10% hard-bounce rate as the stop line for a campaign, a threshold our 37,040-contact US outbound sample cleared at 0.55%.

The steps run in this order because the contract sets which evidence counts, and the renewal date sets how long you have to collect it.

### Step 1: Find the remedy and the notice clause

Pull four documents: the order form, the master services agreement, any statement of work and the last renewal email. Search them for "accuracy", "credit", "refund", "renew" and "notice". Write down the data-quality promise and the remedy beside it, then add the last date you can give non-renewal notice.

Look for an entire-agreement clause as well, and ask counsel how it treats accuracy figures from a website or sales deck. Note the governing law and any arbitration clause while you are there, since both shape the later steps.

### Step 2: Mail a dated sample and keep the SMTP logs

Mail a small sample as soon after delivery as you can, and record the delivery date and the send date. Export the raw bounce log from your sequencer with reply codes and timestamps, because a dashboard screenshot shows a total and hides the reasons.

Stop if hard bounces reach 10%. Our scale for any campaign is under 5% healthy, 5-9% a warning to monitor and fix, and 10% or more stop.

You can test the remaining records with a bulk verifier instead of mailing them, and LeadSonar's Email Validator does this for existing lists, although a vendor may give less weight to verifier output than to real bounces. Our guide on [how to clean a B2B email list](https://leadsonar.io/blog/how-to-clean-a-b2b-email-list/) covers that pass.

### Step 3: Calculate the hard-bounce rate

Calculate the hard bounce rate by dividing hard bounces by contacts mailed. Leave soft bounces out, along with anything your own sending setup caused, such as a blocked domain. A contract threshold, if there is one, outranks our scale.

This worked calculation uses illustrative inputs, and none of them are LeadSonar measurements.

| Input | Value |
|---|---|
| Contacts delivered under the order | 10,000 |
| Contacts mailed in the sample | 500 |
| Hard bounces with 550 5.1.1 or 550 5.2.1 | 60 |
| Hard-bounce rate (60 / 500) | 12% |
| Position on our scale | Stop mailing |

The illustrative 60 of 500 works out at 12%, above the stop line. Our recorded 0.55% across 37,040 contacts, the reference we hold as of this quarter, sits far below it. The contract still sets the remedy, and the vendor now has a number to answer.

### Step 4: Count the other defects separately

Give each defect its own column: blank fields you were charged for, duplicates, consumer-domain addresses, records without an MX record, and records that match the wrong person. Separate counts let a vendor accept one line and argue another, which gives you a better position than a single total it can reject outright.

### Step 5: Send a written claim with a number and a deadline

The FTC's guide to [solving problems with a business](https://consumer.ftc.gov/articles/solving-problems-business-returns-refunds-and-other-resolutions) is written for consumers, and its letter structure transfers well. It tells you to give the basics of the purchase, describe the problem, say what you want, attach copies of documents and keep the originals. It also recommends certified mail with a return receipt.

Here is a sample claim built on the worked calculation above, so every figure in it is illustrative.

Subject: Credit request under our order form for the delivered contact file We received 10,000 contacts under the order form and mailed a sample of 500 within 14 days of delivery. 60 returned hard bounces with reply codes 550 5.1.1 or 550 5.2.1, a 12% hard-bounce rate. The SMTP log is attached, with send dates. We rely on the data quality and credit terms in section 7 of the agreement. We ask for a credit covering the 60 failed records, plus a replacement file or a credit for the unmailed balance. Please reply in writing within 14 days. If we cannot resolve this directly, we will follow the dispute clause in the agreement.

### Step 6: Send the non-renewal notice on its own track

Send non-renewal notice in a separate message, to the address and by the method the contract names, even while the refund talk continues. A refund dispute that runs past the notice date can leave you arguing about one term while a second term begins.

### Step 7: Escalate outside the vendor

Ask for a named owner with authority over credits first. The FTC advises asking for a manager and keeping notes on each conversation and promise. If that fails, the contract's dispute clause, your card issuer and the outside channels in the next sections are the remaining routes, and counsel can tell you which order fits your agreement.

## How to dispute B2B software auto renewal

As of Q4 2026, LeadSonar's list quality reference is 0.55% hard bounces across 37,040 mailed contacts, a figure a buyer can set beside any accuracy claim in a contract before a renewal date arrives.

Public complaint threads that LeadSonar's competitor scans collected during 2026 repeat two themes: annual contracts with long auto-renewal notice windows, and accuracy terms users say were not honored. We attribute these user reports to no particular vendor, and we have not tested them against any contract.

Reading the auto renewal clause in a sales data contract starts with the date arithmetic, and each input comes from your own paperwork.

| Item | Where to find it | What to do with it |
|---|---|---|
| Term end date | Order form or last renewal email | Record it in the claim file |
| Notice period | Renewal clause | Count back from the term end the way the clause counts, in calendar or business days |
| Last day to give notice | Term end date minus the notice period | Set a reminder one full billing cycle earlier |
| Notice method | Notices clause | Send by the named method to the named address |

Before you sign the next annual term, set the discount against that notice burden. LeadSonar's Growth plan is $79 on the monthly basis and $66 per month on the annual basis, about 16% off, which shows the size of discount a buyer can weigh against a twelve-month commitment.

Two pieces of US law come up in these disputes, and each has a limit worth checking with counsel:

- [New York General Obligations Law section 5-903](https://www.nysenate.gov/legislation/laws/GOB/5-903) covers contracts for service, maintenance or repair to or for real or personal property. It makes an automatic renewal provision unenforceable against the person receiving the service unless the provider first gave written notice, served personally or by certified mail, inside a window the statute sets before the termination deadline. The section also excludes short renewal periods, so read its day counts directly, and ask a New York lawyer whether a data subscription fits that wording.

- The [Restore Online Shoppers Confidence Act, 15 U.S.C. 8403](https://www.law.cornell.edu/uscode/text/15/8403), bars charging a consumer through an online negative option unless the seller discloses material terms and gets express informed consent. The seller must also offer simple mechanisms to stop recurring charges. The text speaks of a "consumer", so ask counsel whether it reaches your company's purchase.

## How to get a refund for bad B2B contact data through your card issuer

As of Q4 2026, LeadSonar does not bill the roughly 15% of contacts that miss across its 18-20 provider waterfall in 2026 production data, so a miss leaves nothing to dispute.

A card dispute on a company purchase is a narrower route than consumer guides suggest. [Regulation Z, 12 CFR 1026.3(a)](https://www.law.cornell.edu/cfr/text/12/1026.3) lists extensions of credit primarily for a business, commercial or agricultural purpose, and credit to other than a natural person, among its exempt transactions.

The billing-error rule in the same regulation, 12 CFR 1026.13, is the one consumer guides usually describe. It covers services not delivered as agreed, and it starts its written-notice deadline from the first statement showing the error. Because of the exemption, do not assume those rights apply to a business-purpose card.

Call your issuer and ask which dispute terms apply to the account, and how long you have under the cardholder agreement. Check the vendor contract before filing too, and ask counsel whether a dispute could put you in breach while the contract is still running.

## Where to escalate when the vendor says no

As of Q4 2026, a claim built like LeadSonar's own bounce record, 204 hard bounces out of 37,040 mailed contacts, gives any outside reviewer a count and a denominator to work with.

The FTC's guide lists your state attorney general or state consumer protection office, which may mediate complaints or investigate. It also points to local Better Business Bureau offices, and it notes those offices are supported by local businesses.

ReportFraud.ftc.gov takes reports as well. The FTC says it "doesn't resolve individual complaints", so a report helps law enforcement detect patterns and will not recover your money on its own. All of these channels are built around consumer complaints, and a company buyer may find them a poor fit.

For a contract of real size, the practical next step is counsel and the dispute clause, whether that names arbitration or a venue. Compare the likely cost against the credit you are chasing before you commit to it.

## How to avoid paying for bad data in the first place

As of Q4 2026, LeadSonar charges only for the contacts its waterfall finds, resolving roughly 85% of contacts across 18-20 providers in 2026 production data and leaving the remaining misses unbilled.

Billing structure sets how much you can lose before you notice. Pay-per-result contact data on a monthly plan limits the advance payment to one month, $29 on LeadSonar Starter or $79 on Growth, while a prepaid annual term puts twelve months at risk.

| Billing model | What a miss costs you | Where your remedy lives |
|---|---|---|
| LeadSonar pay-per-result | No charge, because failed lookups are absorbed and never billed | Billing mechanics applied per lookup and per cell |
| Per-attempt credits | The attempt, by definition of the model | Vendor's credit policy |
| Per-seat subscription | Nothing extra per miss, and the seat fee runs regardless | Vendor's terms |
| Annual contract | Already paid or committed in advance | Contract remedy clause and notice window |

If an annual contract is still the right call for your procurement process, negotiate these points before signing:

| Clause to ask for | Why it matters at refund time |
|---|---|
| A defined accuracy metric, such as hard-bounce rate on a mailed sample | Gives both sides one number to check |
| A stated remedy, with credit or refund named | Removes the argument over what happens after a miss |
| A time limit for reporting defects after delivery | Separates a bad record from decay |
| A paid pilot on your own ICP before the annual term | Lets you run steps 2 to 4 with little money at risk |
| A renewal reminder from the vendor, with a shorter notice window | Moves the calendar burden off your team |

Run your own test on any vendor, LeadSonar included. Take 500 contacts in your ICP and mail them within days of delivery, then read the bounce log against our 5/9/10 scale. Dedupe the list before you mail it, so duplicates do not distort the count.

## What LeadSonar refunds and when

As of Q4 2026, LeadSonar bills CSV enrichment per filled cell and refunds unfilled cells, while its 37,040-contact mailed sample recorded a 0.55% hard-bounce rate.

LeadSonar publishes these five billing mechanics.

| Situation | What happens |
|---|---|
| Failed lookup, all providers miss | Marked not_found and never billed |
| Empty cell in a CSV enrichment | Refunded |
| Verification failure | Refunded instantly |
| Lead search and filtering | Free, with a credit spent only when you reveal a contact |
| Unused credits on Growth and above | Roll over |

LeadSonar runs two independent credit pools. Revealing a contact uses 1 lead credit, and enrichment credits cover AI actions and email, phone and CSV enrichment. EmailShield is a separate paid verification feature inside AI enrichment that you choose to run on an address, and it costs enrichment credits.

Here is what per-filled-cell billing does to a file. This worked calculation applies two recorded rates to a hypothetical 10,000-row CSV: our roughly 85% waterfall hit rate from 2026 production data, and the 58.7% employee-count fill we recorded on the 37,040-contact file.

| Input | Email | Employee count | Total |
|---|---|---|---|
| Cells requested on a 10,000-row CSV | 10,000 | 10,000 | 20,000 |
| Cells filled | 8,500 | 5,870 | 14,370 |
| Cells empty and refunded | 1,500 | 4,130 | 5,630 |

Headcount blanks show up in our other sample too, with 33.6% of records in the 103,000-contact file carrying an unknown headcount as of Q4 2026. A charge per requested field would bill those blanks, while per-filled-cell billing refunds them.

These mechanics apply at the moment LeadSonar processes a record. A contact the waterfall found and charged, who changes jobs months later, sits outside all five rows, so keep your send close to your enrichment date. Apply the Step 1 contract review to our terms too, since any vendor's published billing line should match what you sign.

## What our production runs show about volume and evidence

As of Q4 2026, LeadSonar has run a 293K-record gap-fill and a 220K-record batch through the same pipeline, with 0 pipeline errors recorded across a separate 103,000-row scoring file.

### Case: a 274K-row parity test against an old vendor

A customer already held custom-personalization columns from an existing vendor and ran an A/B parity test against our pipeline across a 274K-row pass. Our pipeline reproduced those columns at par or better quality at a fraction of the cost, and that gave the customer a like-for-like result to judge. We recommend the same shape to any buyer: run the same records through both options and compare the columns before you commit.

### Case: a 293K-record gap-fill on an existing list

One list arrived with empty fields, the position a failed refund claim leaves a buyer in, and we backfilled it in a 293K-record gap-fill through the same pipeline as a 220K-record batch. A file full of blanks is the job where charging per filled cell has the most bearing on the bill, because the cells that stay empty are refunded.

### Case: about 5,000,000 enrichments on a flat package

One enrichment engagement scaled to about 5,000,000 enrichments priced as a flat package instead of per seat. Buyers at that volume should ask how a package treats blanks and misses, since the per-cell logic may be written differently.

In the 37,040-contact export, four columns were empty or constant, among them email_verification_result and department. We excluded them instead of reporting them as if they were measurements, and you should hold a vendor's accuracy report to the same standard. That sample leans toward Software Development at about 27.7% and Advertising Services at about 16.7% of its 37,040 records, so read its figures as representative of that sample and test your own segment.

## Where LeadSonar fits

As of Q4 2026, LeadSonar recorded 0.55% hard bounces across a 37,040-contact mailed sample and charges only for the contacts its waterfall finds.

LeadSonar is focused lead intelligence: find, verify, enrich and score, across 120M+ B2B contacts and 45M+ B2C contacts you can search and reveal. Pair it with any sender you already use. If you want prospecting and sending in one subscription, a bundled suite is the better fit.

It fits buyers who have been through the refund process and want fewer records to dispute next time. Starter is $29/mo and Growth is $79/mo, with credit rollover from Growth upward. To run the 500-contact test on your own ICP before paying anything, start the free trial, with 1,000 credits, 7 days and no credit card, from the [LeadSonar pricing page](https://leadsonar.io/pricing/).

## Frequently asked questions

As of Q4 2026, LeadSonar's first-party bounce reference is 204 hard bounces across 37,040 mailed US outbound contacts, a 0.55% rate.

### How to get a refund for bad B2B contact data?

Send the vendor a written claim with a count of defective records, dated SMTP logs, the contract clause you rely on and a reply deadline. Measure the claim against a fixed scale: under 5% hard bounces is healthy, 5-9% is a warning to monitor and fix, and 10% or more means stop mailing. For reference, LeadSonar recorded a 0.55% hard-bounce rate across 37,040 mailed US outbound contacts. Refunds depend on your contract, so take a live dispute to counsel.

### Do companies legally have to give you a refund?

Your contract decides this one, because the FTC's refund guidance names no general legal duty to refund and sends buyers to the seller's return policy and deadlines first. Look for the written billing term and have a lawyer read it. LeadSonar's term is an example of what to look for: a lookup that misses across its 18-20 provider waterfall is never billed, which covered the roughly 15% of contacts that missed in 2026 production data.

### How to professionally ask for a refund?

Put the request in writing with the order reference, the defect count, the evidence, the exact remedy you want and a date for a reply. The FTC's sample complaint letter guidance covers the same elements and recommends certified mail with a return receipt. State the count with its denominator, as LeadSonar does with its own record of 204 hard bounces across 37,040 mailed contacts, a 0.55% rate, so the reader can check the division.

### How do I report a company that won't give a refund?

Report it first to your state attorney general or state consumer protection office, which the FTC says may mediate complaints, then consider the local Better Business Bureau. ReportFraud.ftc.gov also accepts reports, although the FTC says it does not resolve individual complaints. These channels are written for consumers, so a company buyer should also ask counsel about the dispute clause in the contract. Present your defect count with its denominator, the form of LeadSonar's own 204 hard bounces across 37,040 contacts.

### How to get customer service to give you a refund?

Give a named owner with authority over credits, usually an account manager or finance contact, the count, the logs, the clause and a deadline. The FTC recommends asking for a manager when the first representative will not give you what you want, and keeping notes on each conversation. Place your sample on LeadSonar's bounce scale, where 10% or more means stop, so the owner has a threshold to answer. LeadSonar's own sample sat at 204 hard bounces of 37,040 mailed contacts, a 0.55% rate.

### What is B2B contact data?

B2B contact data is the set of work details used to reach a person at a company: name, title, company domain, email, phone and firmographics such as headcount, industry and funding stage. LeadSonar returns 30+ fields per lead from a database of 120M+ B2B contacts and 45M+ B2C contacts you can search and reveal. In a refund claim, each of those fields is a separate item you can count as filled or blank.

## Methodology

As of Q4 2026, the LeadSonar figures cited on this page come from two stated samples, of 37,040 and 103,000 contacts, plus 2026 production data.

The 0.55% hard-bounce rate and the 99.98% valid-MX rate come from a 37,040-record US outbound sample that we processed, weighted toward US software companies, which also supplied the 58.7% employee-count fill.

The consumer-domain share and the 33.6% unknown headcount come from a 103,000-contact ICP-scored B2B sample, which also recorded zero pipeline errors.

The roughly 85% hit rate and the roughly 15% unbilled misses come from 2026 LeadSonar production data on a full email waterfall. Each figure describes its stated sample, and your own segment may differ.

The bounce scale of under 5%, 5-9% and 10% or more is our operating rule. The complaint themes come from public threads our competitor scans collected during 2026, and we report them as user statements. Tables labelled as worked calculations use illustrative or hypothetical inputs. No competitor price or figure appears here, so no competitor fetch date applies. LeadSonar has not tested any other vendor's data and makes no claim about its accuracy.

Legal references were read on their official or standard published pages on 6 November 2026, starting with the New York Senate's text of General Obligations Law section 5-903 and the FTC's consumer guide.

We read 12 CFR 1026.3 and 1026.13 and 15 U.S.C. 8403 through Cornell's Legal Information Institute on the same date. They can change, and counsel should check them against your contract. The billing basis is monthly list prices unless a sentence says annual, and prices are subject to change.

Used stats: the 37,040 and 103,000 samples and the 2026 production figures listed in this Methodology.

Last updated: November 2026.

Written by Sabo Nagy, Founder & CEO at LeadSonar. Sabo works on LeadSonar's pricing and its pay-per-result billing model, where buyers pay for the contacts the waterfall finds. [Author page](https://leadsonar.io/authors/sabo-nagy) - [X](https://x.com/mrcoldemail)

### Related articles

[B2B Contact Data Accuracy: How to Measure It and Why Most Lists Fail [2026]](https://leadsonar.io/blog/b2b-contact-data-accuracy/)[How to Build an Outbound List From Scratch: A Step-by-Step Workflow](https://leadsonar.io/blog/how-to-build-an-outbound-list-from-scratch/)[How to Clean a B2B Email List (Without Killing Your Deliverability) [2026]](https://leadsonar.io/blog/how-to-clean-a-b2b-email-list/)