Define the population
State where records came from, which period they represent, and why the sample belongs to the question.
Measure the list.
Show the method.
RefreshList Lab is the home for first-party research into email list quality. We publish a finding only when the sample, denominator, verification context, and privacy boundaries can be explained beside it.
Research format preview · No fabricated benchmark data · Last reviewed 18 September 2026THE SHORT ANSWER
What is RefreshList Lab?
It is an evidence-first research format for understanding list quality without turning a demonstration into a statistic. The Lab separates observed outcomes from assumptions, keeps unknowns visible, and puts limitations next to every conclusion.
STATUSFORMAT DEFINED
The reporting framework exists. A representative research dataset and published market findings do not yet exist.
EDITORIAL RULENO NUMBER WITHOUT A DENOMINATOR.
Every rate must say what was counted, what was excluded, and when it was measured.
01 / THE RESEARCH CONTRACT
Before a chart, there is a contract.
A reader should be able to reproduce the question a study answered—and identify the question it did not answer.
Declare the denominator
Separate source rows, unique addresses, eligible checks, completed outcomes, and unresolved results.
Protect the data
Minimize identifying content, document authorization, and never treat a customer upload as public research by default.
Show the limitation
Coverage gaps, provider behavior, timing, and missing signals stay beside the finding—not buried in fine print.
02 / HOW A STUDY WOULD RUN
From source rows to a defensible finding.
- FrameWrite the question, population, inclusion rules, and collection window before looking at rates.
- InspectCount source rows, malformed values, duplicates, and eligible unique addresses without charging or inferring outcomes.
- VerifyRecord provider context, completed results, and unknowns. A provider outcome is evidence at a point in time, not a delivery promise.
- ReportPublish the numerator, denominator, date, coverage, exclusions, and limitations together.
EXAMPLE REPORTING FIELDUnique checked addresses
“Unique checked addresses” is not interchangeable with submitted rows. A future report would show both, explain the difference, and state which one supports each rate.
Submitted rowsnot collectedUnique checkednot collectedUnknown outcomeskept separate
03 / EMAIL HEALTH INDEX
A format, not a made-up score.
The planned Email Health Index would organize several defined measures. It would not compress incomplete evidence into a universal “quality percentage.”
Potential measures
- Duplicate rate, using a stated source-row denominator
- Malformed or skipped rate, separate from mailbox results
- Valid, invalid, and unknown outcomes within the checked population
- Supported role, catch-all, or disposable signals with coverage stated
CURRENT STATENo published index value
There is no representative sample, current industry prevalence estimate, trend line, or ranking to present yet. Preview values would be labeled as examples, never measurements.
Read the index format04 / RESEARCH NOTES
Questions worth investigating.
NOTE 01What does list hygiene actually measure?
Separate acquisition context, suppression rules, duplicates, verification, and later delivery feedback.
NOTE 02When does a domain answer stop being a mailbox answer?Catch-all behavior can preserve uncertainty even when a receiving server accepts many names.
NOTE 03How should an unknown result be handled?Keep the provider reason, checking context, and uncertainty visible instead of inventing confidence.
COMMON QUESTIONS
Research questions, answered.
Does RefreshList publish an email quality benchmark today?+
Not yet. The Lab is a research format and editorial standard. There is no published representative industry sample, customer-derived benchmark, or independent RefreshList accuracy study to cite today.
What would a real Lab study disclose?+
A real study would disclose its population, sampling and inclusion rules, collection period, duplicate treatment, denominator for every rate, provider coverage, privacy method, and known limitations alongside the findings.
Would customer lists become public research data?+
No. Customer lists do not become public research data by default. Any future research dataset would require an explicit, documented privacy and authorization process, with identifying content minimized or removed.
How does the Lab treat unknown results?+
Unknown remains separate from valid and invalid. A result that cannot be established is not silently converted into a failure, a pass, or a confidence score.
What is the Email Health Index?+
It is a planned reporting format, not a current market score. It would organize defined measures such as duplicate rate, malformed rate, verification outcomes, and supported signals only when genuine observations and denominators are available.
NO INVENTED INSIGHTS.
When a study is ready, the method and the finding will arrive together.