How to build and clean a list of domain ideas before you spend money
A good domain session starts ugly. You have fragments in Notes app, half rows in Sheets, words you liked at 1 a.m., and three prefixes you keep repeating because they feel safe (get, try, use).
The goal of cleaning is not to make the list pretty. The goal is to make the list scannable so your tools can answer availability questions without you constantly fixing typos.
This is the workflow I use before I paste anything into the bulk checker.
Quick answer
Brainstorm wide, normalize once, delete aggressively on spelling and speakability, then convert surviving labels into full domains with deliberate TLDs.
Only then do you bulk scan. If you only have a single strong label and many extensions, use check domain extensions instead of inventing fifty full domains by hand.
Stage 1: capture without filtering
Give yourself permission to write bad ideas for ten minutes. Quantity first.
Useful prompts:
- what verb describes what the product does?
- what noun describes the outcome?
- what metaphor fits your industry without sounding like a parody?
- what two-word combinations still sound good when you shout them across a room?
Write strings, not domains yet. Examples: clearledger, flowkit, northpilot, brightlocaltools.
Stage 2: pick naming patterns, not individual genius
Most workable lists mix a few patterns:
- compound nouns:
ledgerkit,taskforge - verb plus noun:
buildflow,shiptrack - real word plus modifier:
honestinvoice,simpleledger - invented brandable:
zorbly,mantivex(spellability matters more here)
If your list is ninety percent invented brandables, you will spend a lot of time teaching people how to spell your company. If it is ninety percent generic descriptors, you will spend a lot of time fighting SEO sameness. A sane list mixes categories on purpose.
Stage 3: hard deletes before DNS
Some failures are cheaper to catch with your eyes than with any tool.
Delete or quarantine anything that:
- includes hyphens unless you have a strong reason (radio tests fail often)
- includes digits unless you are building a numbered product line people expect
- is easily confused with a major existing brand in your space
- is a joke that will not survive your first investor email
- relies on a TLD to complete the phrase (
example.ly) in a way that reads cute today and tired next year
Keep a separate “maybe” bucket if you cannot decide. Do not let maybes clog the first scan.
Stage 4: convert labels to full domains deliberately
This is where people accidentally create garbage input.
If you care about .com, add .com on purpose. If you also want to compare .io and .app, duplicate rows with those extensions. Do not assume a tool will guess your extension preferences.
Example block for one product concept:
clearledger.comgetclearledger.comclearledger.ioclearledger.app
Notice the pattern: same family, controlled variation. Random jumps (clearledger.com, randomotherthing.io) make comparison harder later.
Stage 5: normalize syntax like a compiler would
Paste your list into a text editor and run boring fixes:
- lowercase everything for consistency
- strip URLs down to hostnames
- remove trailing dots unless you truly mean absolute names in a zone file context (you usually do not)
- collapse duplicate lines
- sort alphabetically so duplicates jump out visually
If you work in Sheets, use a formula column that flags rows without a dot in the string. Those rows are labels, not domains, until you append a TLD.
Stage 6: scan, then sort by outcome and gut
After a bulk scan, I keep three local tags in a sheet:
- dead: DNS shows real use or obvious parking
- maybe: quiet DNS but something feels socially or legally risky
- confirm: quiet DNS and I would happily say the name on a call
Only the confirm rows deserve registrar time. Maybes deserve a walk around the block.
Stage 7: sanity checks that are not DNS
DNS cannot tell you if a name is a good idea. It can only shrink your search space.
Before you pay, do at least:
- say the domain out loud with and without the extension
- type it on a phone keyboard without autocorrect saving you
- search the bare word in your target language markets if you ship internationally
- check obvious social handles if those matter for your launch
Common cleaning mistakes
Mistake: mixing labels and full domains in one bulk paste.
The checker expects full domains. Labels belong in a sweep or in a preprocessing step.
Mistake: keeping twenty near duplicates.
If getflow, tryflow, and useflow all pass DNS, you still have to pick one brand. Narrow early.
Mistake: ignoring pluralization.
taskforge.com and taskforges.com are different products in customer minds. Decide which family you want before you register.
Checklist: is this list ready to scan?
- Every line is a full hostname with a TLD you chose on purpose
- No duplicate rows unless you are intentionally testing resolver variance
- No email addresses, paths, or query strings
- You can read the whole list aloud in under two minutes
FAQ
How large should a first scan be?
Large enough to compare fairly, small enough that you will actually delete losers. For me that is often fifty to one hundred lines per pass.
Should I include premium TLDs I cannot afford?
If budget is a constraint, exclude them. Hope is expensive.
What if every good .com is taken?
Either change the label family or decide intentionally on a non-.com strategy with eyes open about confusion and email.
Next step
Clean one messy list today and run it through the bulk checker.
Keep the spreadsheet habit for one week. You will see patterns in your own naming biases faster than you expect.
BenOpt checks live DNS on the Cloudflare edge and uses RDAP when it can. Treat results as strong signals. Confirm availability and renewal price at a registrar before buying.
Check your domain ideas
Paste one domain per line. We check each one against live DNS and show you which ones look available.
Try the bulk checker