DomainGemsAI
Established Member
- Impact
- 117
After the last thread here got pushback for leaning too hard on theory, I spent this weekend trying to break my own pricing engine against a few recent sales.
Sharing the results in case the pattern is useful — or if anyone wants to poke holes in it.
Names checked:
Rank.ai
honeypot.com
StrandWellness.com
CleverDigitalMarketing.com
AIgoesON.com
Rank.ai I’ve discussed here before — around $10K at expiry auction in 2019, then $200K several years later to an operator, with no execution built between.
I’m including it as a reference point rather than the main story. The earlier version of my engine anchored this much closer to the $10K comp. After adding a separate path for operator-targeted pricing, it lines up better with how the name actually transacted.
honeypot.com sold recently for $150,000.
Real word, multiple use cases — cybersecurity, hospitality, dating, F&B. Current version reads it as a strong single-asset brand and prices it in six figures, which matches the outcome. Earlier version sat too close to generic 8-letter .com comps. That was one of the failure modes I wanted to fix.
StrandWellness.com sold for $717.
Two real words, but wellness is crowded and many similar compounds are interchangeable for the same brand position. The first version over-valued it because it saw structural cleanliness and missed the saturation. Added a saturation check this weekend; now it routes to wholesale clearance, which is where it actually transacted.
CleverDigitalMarketing.com sold for $1,895.
Three-word descriptive phrase. Reads as what the business does, not as a brand. Same pre-patch issue — the engine was treating structural quality as commercial signal. Now it routes to wholesale.
AIgoesON.com — sub-$1K territory.
Sentence fragment, despite the AI keyword. Same story. The engine was rewarding “AI” plus surface-level cleanliness and missing that the construction does not really work as a brand. Patched this weekend.
What I’m taking from the five:
The hard part is not simply valuing a name higher or lower. It is knowing which market the buyer is in.
Some names are investor inventory and clear around wholesale comps.
Some names need the right operator at the right time, and comps can badly understate them.
Most names are in the first group, even if they look clean on the surface.
The engine still has work to do. It now under-values the other extreme — names like AI.com that sold for eight figures still sit in my model’s six-figure band because the ceiling logic was built for portfolio-grade names, not category-defining assets.
That is the next architectural fix.
Curious where others have seen this break in either direction:
Names that looked ordinary on wholesale comps but sold to an operator for much more.
Or names that looked operator-grade structurally but cleared at wholesale.
That is where the calibration actually gets tested.
Sharing the results in case the pattern is useful — or if anyone wants to poke holes in it.
Names checked:
Rank.ai
honeypot.com
StrandWellness.com
CleverDigitalMarketing.com
AIgoesON.com
Rank.ai I’ve discussed here before — around $10K at expiry auction in 2019, then $200K several years later to an operator, with no execution built between.
I’m including it as a reference point rather than the main story. The earlier version of my engine anchored this much closer to the $10K comp. After adding a separate path for operator-targeted pricing, it lines up better with how the name actually transacted.
honeypot.com sold recently for $150,000.
Real word, multiple use cases — cybersecurity, hospitality, dating, F&B. Current version reads it as a strong single-asset brand and prices it in six figures, which matches the outcome. Earlier version sat too close to generic 8-letter .com comps. That was one of the failure modes I wanted to fix.
StrandWellness.com sold for $717.
Two real words, but wellness is crowded and many similar compounds are interchangeable for the same brand position. The first version over-valued it because it saw structural cleanliness and missed the saturation. Added a saturation check this weekend; now it routes to wholesale clearance, which is where it actually transacted.
CleverDigitalMarketing.com sold for $1,895.
Three-word descriptive phrase. Reads as what the business does, not as a brand. Same pre-patch issue — the engine was treating structural quality as commercial signal. Now it routes to wholesale.
AIgoesON.com — sub-$1K territory.
Sentence fragment, despite the AI keyword. Same story. The engine was rewarding “AI” plus surface-level cleanliness and missing that the construction does not really work as a brand. Patched this weekend.
What I’m taking from the five:
The hard part is not simply valuing a name higher or lower. It is knowing which market the buyer is in.
Some names are investor inventory and clear around wholesale comps.
Some names need the right operator at the right time, and comps can badly understate them.
Most names are in the first group, even if they look clean on the surface.
The engine still has work to do. It now under-values the other extreme — names like AI.com that sold for eight figures still sit in my model’s six-figure band because the ceiling logic was built for portfolio-grade names, not category-defining assets.
That is the next architectural fix.
Curious where others have seen this break in either direction:
Names that looked ordinary on wholesale comps but sold to an operator for much more.
Or names that looked operator-grade structurally but cleared at wholesale.
That is where the calibration actually gets tested.
