Article
How to Calculate Email Delivery Rate: Formula, Denominator, and Worked Example
By Unlimited Verifier Team ·

Summary
Email delivery rate is (attempted sends minus final bounces) divided by attempted sends, times 100. For example, 48,600 accepted out of 50,000 attempts equals 97.2%. Count only attempted sends after suppression, and wait for soft bounces to finish retrying before calculating.Email delivery rate is the percentage of attempted sends that a receiving mail server accepted, calculated as (attempted sends − final bounces) ÷ attempted sends × 100, so 48,600 accepted out of 50,000 attempts equals 97.2%. The formula is simple. Most disagreements about delivery rate come from two choices: what counts as "attempted," and what counts as a "bounce."
Quick answer
- Formula: (Attempted sends − final bounces) ÷ attempted sends × 100.
- Denominator: Use attempted sends after suppression, not list size. Report suppressions separately.
- Numerator: Count final failures only. Wait for soft bounces to finish retrying.
- Target: Aim for 98% or higher (our recommendation). Investigate anything below 95%.
- Split it: Calculate the rate per receiving domain. A healthy overall average can hide one provider that is failing.
What does email delivery rate measure, and what does it miss?
Delivery rate measures whether the receiving server accepted your message. It does not measure whether a person saw it. Treat it as a ceiling on inbox placement, not a measure of placement.
When your server connects, the receiving server answers with an SMTP status. RFC 5321 defines the basic reply codes. RFC 3463 defines enhanced status codes, where the first digit does most of the work:
- 2.x.x: accepted. Counts as delivered.
- 4.x.x: temporary failure. The sender retries later.
- 5.x.x: permanent failure. Counts as a hard bounce.
An accepted message can still land in spam, so a 99% delivery rate and a 99% inbox rate are different claims. Opens are no substitute for either. Since Apple Mail Privacy Protection, machine-triggered image loads inflate open counts, as covered in why open rates are unreliable after Apple Mail Privacy Protection. For the complaint and placement signals that delivery rate cannot show, see deliverability metrics every sender should track.
Bottom line: Delivery rate tells you what the receiving server accepted, not what the recipient saw.
What is the formula for email delivery rate?
Use this formula:
Delivery rate (%) = (Attempted − Final failures) ÷ Attempted × 100
- Attempted: messages your platform handed to a receiving server. Exclude addresses suppressed before send, because they were never attempted.
- Final failures: hard bounces (5.x.x) plus soft bounces (4.x.x) that failed after all retries ended.
- Pending deferrals: messages still being retried. Wait until they resolve before you calculate. Do not count them either way.
Delivery rate and bounce rate add up to 100 when both use the same attempted base. If your final bounce rate is 2.8%, your delivery rate is 97.2%. For bounce rate on its own, see how to calculate email bounce rate.
Bottom line: Delivery rate is 100 minus your final bounce rate, measured on attempted sends after retries close.
Which denominator should you use?
Use attempted sends, not list size. Suppressed contacts were never mailed, so counting them as failures makes your performance look worse than it is.
Worked example 1: one campaign, two denominators
A campaign targets 52,000 contacts. The platform suppresses 2,000 of them because of earlier hard bounces and unsubscribes, which leaves 50,000 attempts. The send produces 1,100 hard bounces and 300 soft bounces that failed after retries.
- Delivered: 50,000 − 1,100 − 300 = 48,600
- Against attempted sends: 48,600 ÷ 50,000 = 97.2%
- Against audience size: 48,600 ÷ 52,000 = 93.5%
The 3.7-point gap comes entirely from the 2,000 suppressions. Report 97.2% as the delivery rate, and report the suppressions separately as a list-hygiene figure.
Bottom line: Keep suppressed contacts in their own column so they do not distort the delivery rate.
Which bounces should you count as failures?
Count every final failure in the numerator, but classify it by cause. Two different hard-bounce types look identical in a single "bounced" total and call for different fixes.
- Address failures (5.1.x): The address is wrong or the mailbox does not exist. For example,
5.1.1means the destination mailbox does not exist. The fix is list hygiene: suppress the address and verify the list. - Policy or reputation blocks (5.7.x): The message was refused for policy or security reasons, not because the address is bad. Many large providers use 5.7.x codes for reputation-based rejections. Cleaning the list will not fix these. Look at complaints, authentication and sending patterns instead.
- Soft failures (4.x.x): Count these only when they have failed for good after retries. A deferral that later succeeds is a delivery.
Bottom line: A failure should be classified by its cause, not only by its first digit, or your cleanup will target the wrong problem.
Which time window should you use?
Use a single campaign for diagnosis, a rolling 30-day window for trends, and avoid drawing conclusions from very small sends. Small denominators move quickly. One bounce on 500 attempts changes the rate by 0.2 percentage points. The same bounce on 50,000 attempts changes it by 0.002 points.
Calculate only after the retry window has closed. Your platform's documentation states the exact window, and it varies by provider. As a working rule, compare figures only when they cover the same window.
Bottom line: Use campaign-level rates to find problems and rolling rates to judge trends, and never read a trend from one small send.
How do you calculate delivery rate step by step?
- Pull the send log. Export per-recipient results for one campaign or one time window.
- Count attempts. Use messages handed to a receiving server, after suppression.
- Classify each failure as address (5.1.x), policy (5.7.x), soft-final (4.x.x after retries), or deferred (still retrying).
- Wait for retries to finish. Do not calculate while any message is still pending.
- Apply the formula to final counts only.
- Split by receiving domain (Gmail, Outlook, Yahoo and others) and by list segment.
- Record the result with its date range, denominator and platform, so you can compare it with later sends.
Bottom line: Calculate on final counts after retries, and always record which denominator you used.
Why does an overall delivery rate hide problems?
An overall rate is a weighted average of unequal parts. A large, healthy provider can mask a small, failing one.
Worked example 2: the same campaign, split by receiving domain
Using Worked example 1 (50,000 attempts, 1,400 final failures):
| Receiving domain | Attempts | Final failures | Delivery rate |
|---|---|---|---|
| Gmail | 30,000 | 360 | 98.8% |
| Outlook | 12,000 | 240 | 98.0% |
| Yahoo | 5,000 | 400 | 92.0% |
| Other | 3,000 | 400 | 86.7% |
| Total | 50,000 | 1,400 | 97.2% |
Yahoo and "Other" together account for 800 of the 1,400 failures (57%), while receiving only 16% of attempts (8,000 of 50,000). The overall 97.2% looks fine, but the split shows a segment doing most of the damage. Check where that segment's addresses came from before you send to it again. Older or purchased contacts are a common cause, though the split alone does not prove the source.
Bottom line: Always split by provider, because the segment that is failing is rarely the one that moves the average.
How do you estimate delivery rate before you send?
You can forecast deliverability by verifying the list before the send. Verification finds the addresses most likely to produce hard bounces, and those are the bounces you could have predicted. For catching bad addresses at the point of signup, see real-time email validation.
Worked example 3: a 40,000-contact list
- Valid: 33,600 (84%)
- Risky: 4,400 (11%), addresses the verifier could not confirm, such as those on catch-all domains
- Invalid: 2,000 (5%)
The Invalid group sets a hard floor on bounces. Even if every other address delivers, 2,000 bounces cap the rate at (40,000 − 2,000) ÷ 40,000 = 95.0% for a full send. Remove Invalid addresses before you send and the cap disappears. A Valid result is not a delivery guarantee, and Risky addresses remain unknown. Send to the Valid group first and measure its first-send rate before you widen the audience.
The cost and limits are straightforward. Unlimited Verifier's Pro plan is $29 per month, or $25 per month billed annually, and it handles lists of up to 25,000 emails each. The 40,000-row file splits into 25,000 and 15,000. The Free plan caps each list at 150 emails, so this file needs Pro. Standard verification is unlimited on every plan. The 99.5% accuracy figure for deep verification is Unlimited Verifier's own stated number, not an independent benchmark. To spot-check a single address, use the free email checker. For a full file, upload and verify a list. For background on what verification costs across vendors, see email verification pricing.
Bottom line: Verification does not measure inbox placement, but it removes the bounces you can predict before you send.
Which method or tool should you use to measure delivery rate?
Use more than one source. Each one covers a different slice of the picture.
| Method | Where the numbers come from | Granularity | Best for | Limitation |
|---|---|---|---|---|
| Sending platform report | Built-in delivered and bounced counts | Per campaign | Quick checks on each send | Each platform defines "delivered" and its retry window differently |
| Per-recipient log in a spreadsheet | SMTP responses exported from your platform | Any slice you build | Provider and segment splits | Manual work, and it needs a full export |
| Google Postmaster Tools | Google's view of your domain's traffic to Gmail | Domain, Gmail only | Reputation and spam-rate trends | Needs a verified domain and enough Gmail volume before data appears; no per-campaign counts |
| Microsoft SNDS | Microsoft's data on your sending IPs for Outlook.com | Sending IP, Microsoft only | Outlook reputation | You must enrol your IPs; it is IP-based, not domain-based |
| Pre-send list verification | Valid, Risky and Invalid counts per list | Per list | Forecasting and removal | Measures address validity, not inbox placement or engagement |
Verification fits into a broader stack. Teams that automate it often use the email verification API and automation approach. Unlimited Verifier's email verification API accepts single-email checks and batches of up to 10,000 emails per request. The platform has no native ESP, CRM or Zapier integrations, so most teams export a CSV, verify it, and re-import the results. Free-plan results are kept for 24 hours and Pro results for 30 days, so download your exports as you go. For a broader comparison of verifiers, see best email verification tools.
Bottom line: Use the platform report for speed, a spreadsheet for detail, and Postmaster Tools and SNDS for reputation.
What is a good email delivery rate?
Public guidance does not agree on one number, and many sources do not publish a delivery-rate target at all. The closest public anchor is a bounce threshold. Amazon SES's documentation describes a 5% bounce rate as a review trigger. Under this formula, a 5% bounce rate equals a 95% delivery rate. Google's sender guidelines set a different metric, asking bulk senders to keep spam-complaint rates below 0.3%, as set out in its email sender guidelines. Complaints and bounces are related, but they are not the same measurement.
Our recommendation:
| Delivery rate | What to do |
|---|---|
| 98% or higher | Healthy. Keep monitoring each send. |
| 95% to 98% | Find the top three failing domains and segments this week. |
| Below 95% | Pause the segment, verify it, and check its source before resending. |
Benchmarks vary widely with list age and acquisition method. Compare your figure against email marketing benchmarks, then against your sector using email marketing benchmarks by industry 2026. Track the trend rather than any single send.
Bottom line: Target 98% and treat 95% as the point where you investigate.
How do you calculate delivery rate in a spreadsheet?
You need a per-recipient send log and any spreadsheet program. The steps below assume columns: C = receiving domain, D = SMTP enhanced status code, and a sheet named Log.
- Export the log as CSV. Keep the recipient, domain, status and enhanced-code columns. Remove suppressed contacts first, because they were never attempted.
- Set up a Summary sheet. Label cells for Attempted, Hard bounces, Policy blocks, Soft-final and Delivery rate.
- Count attempts. In B2, enter
=COUNTA(Log!A:A)-1to count recipient rows, minus the header. - Count hard bounces. In C2, enter
=COUNTIFS(Log!D:D,"5.*"). In C3, enter=COUNTIFS(Log!D:D,"5.7.*")to count policy blocks separately. - Count soft-final failures. In D2, count 4.x.x rows that your export marks as failed. If it has no final flag, count only 4.x.x rows from sends older than your platform's retry window.
- Calculate the rate. In E2, enter
=(B2-C2-D2)/B2and format it as a percentage. - Split by domain. Add a helper column G with
=IF(OR(LEFT(D2,1)="5",LEFT(D2,1)="4"),1,0), filled down. On a Domains sheet, use=COUNTIF(Log!C:C,A2)for attempts and=SUMIF(Log!C:C,A2,Log!G:G)for failures. Then calculate each rate as=1-(C2/B2). - Verify the next list. Create a free account if you do not have one, then upload and verify a list. Export the results as CSV, Excel or JSON. Lists over 150 emails need Pro, because the Free plan caps each list at 150.
- Suppress Invalid addresses in your sending platform. Keep Risky addresses out of the next send until you have reviewed them, and check any catch-all results with the catch-all email verification guide.
- Log the result with its date range and denominator. Repeat weekly, so a trend shows up before it becomes a problem.
For list upkeep between sends, see email list maintenance best practices.
Bottom line: Four counts and one formula give you a defensible delivery rate for any send.
What are the common mistakes when calculating delivery rate?
- Using list size as the denominator. This counts suppressed contacts as failures.
- Counting soft bounces that are still retrying. Wait for the retry window to close.
- Treating "accepted" as "inbox." A 2.x.x response is not proof of placement.
- Lumping policy blocks in with bad addresses. A 5.7.x block and a 5.1.1 bounce need different fixes.
- Averaging campaign rates. Consider two sends: 10,000 attempts at 99% and 1,000 attempts at 80%. The simple average is (99 + 80) ÷ 2 = 89.5%. Pool the counts instead: 9,900 + 800 = 10,700 delivered, and 10,700 ÷ 11,000 = 97.3%.
- Comparing figures across platforms. Each platform defines "delivered" and retry windows differently. Compare one platform over time.
- Reading a Valid share as a delivery guarantee. Valid addresses can still bounce, defer or land in spam.
Bottom line: Pool counts, wait for retries, separate causes, and never treat acceptance as placement.
How we compared (and limitations)
This guide is based on public documentation: RFC 5321, RFC 3463, Google's email sender guidelines, Amazon SES's published guidance and Google Postmaster Tools. We have not run controlled tests against any platform's delivery reporting. Retry windows, dashboard definitions and sender thresholds vary and change without notice. Worked-example numbers are illustrative. The 99.5% accuracy figure is Unlimited Verifier's own stated number, and plan limits reflect the product's published facts at the time of writing.
Worked example: delivery rate by denominator
| Item | Count |
|---|---|
| List size | 52,000 |
| Suppressed before send | 2,000 |
| Attempted sends | 50,000 |
| Final bounces | 1,400 |
| Delivered | 48,600 |
| Delivery rate (attempted base) | 97.2% |
| Delivery rate (list base, wrong) | 93.5% |
Using list size as the denominator understates performance by about 3.7 points, because suppressed contacts were never mailed.
Frequently asked questions
What is a good email delivery rate?
A delivery rate of 98% or higher is a reasonable target for most senders, based on common industry guidance and our recommendation. Investigate any rate below 95%, because it usually signals list quality problems, authentication issues, or a provider blocking your mail.
Should suppressed contacts be included in the delivery rate?
No. Suppressed contacts were never sent to, so they should be excluded from the denominator. Report suppressions separately so your delivery rate reflects actual sending performance.
Are soft bounces counted as failures?
Only after they fail for good. A soft bounce (4.x.x) is retried by the sender, so it counts as a final failure only when all retries end without acceptance. Until then, treat it as pending.
What is the difference between delivery rate and bounce rate?
They are complements when measured on the same attempted base. Delivery rate equals 100 minus the final bounce rate, so a 2.8% bounce rate means a 97.2% delivery rate.
Does delivery rate mean my emails reached the inbox?
No. Delivery rate only shows that the receiving server accepted the message. The message can still land in spam, so inbox placement needs separate testing and complaint monitoring.
Why is my delivery rate different across email platforms?
Platforms define attempted sends, bounces, and retries differently, and some count suppressions in the denominator. Compare platforms using the same formula and the same time window before drawing conclusions.
How do I calculate delivery rate for one provider like Gmail?
Filter sends by recipient domain, then apply the same formula: (attempted to that domain minus final failures to that domain) divided by attempted to that domain, times 100. A healthy overall average can hide a failing provider.
When should I calculate delivery rate for a campaign?
Wait until all soft bounces have finished retrying, which can take hours to days depending on the receiving server's retry schedule. Calculating too early understates the final rate.
Is a 100% delivery rate possible?
It is rare on real lists, because some addresses are invalid or have closed mailboxes. A very high rate is achievable with good list hygiene, but treat any claim of exactly 100% with skepticism.
Cut Bounces and Protect Your Sender Reputation With Unlimited Email Verification
Every invalid address you send to hurts your sender reputation and your inbox placement. Verifying your list before each send means fewer bounces, healthier deliverability, and a predictable cost. With Unlimited Verifier, standard verification is unlimited on every plan, with no per-email fees.
Unlimited Verifier checks syntax, domain, MX records and the SMTP mailbox. It also detects catch-all domains and flags disposable and role-based addresses. Export a CSV from your email platform or CRM, upload and verify the list, then re-import the clean results. You can also automate checks with the email verification API.
| Free | Pro | |
|---|---|---|
| Price | $0 | $29/month ($25/month billed annually) |
| Lists | 25 a month, up to 150 emails each | Unlimited, up to 25,000 emails each |
| Deep verifications | 750 a month | 25,000 a month |
| API access | No | Yes (100,000 emails a month) |
| Result retention | 24 hours | 30 days |
Why teams choose Unlimited Verifier
- Unlimited standard verification on all plans, including Free, with no per-email fees.
- Catch-all detection so you know which domains accept every address.
- Clear Valid / Risky / Invalid results, exportable as CSV, Excel or JSON.
- An open-source verification engine you can inspect on GitHub.
Unlimited Verifier states 99.5% accuracy for deep verification.
Create a free account and verify your next list before you hit send.
Want to test a single address first? Try the free email checker.