idea feedback - leadgen for home service companies using UAVs and computer vision?

[MRiabov]

Hello Fastlane,
Posting in part because this is an idea somebody may execute on, in part because lots of people on FL forum are doing home services.

I was washing windows to pay for rent the other day, and I've seen another salesman selling electricity subscriptions to houses. He was rejected on multiple because "we are already subscribed to Electric 123".
While this was unrelated, I saw something that interested me: door-to-door salespeople are going into regions they basically know nothing of. A "good region" in window cleaning means you clear 8 houses from 15, a "bad region" is about 1 house in 20, or have a significant amount of difficult features.

And those are defined on two things: competition, and owners' wealth.

I was in robotics for some time, and this makes me think: one hundred percent we can collect this intel using drones. Better yet, instead of manually annotating things, a drone operator can simply fly through the street, and a good enough computer vision model would detect:
- Make of their car (or multiple), i.e. price
- How recent/old their house looks
- Quality of their doors/windows. roofs.
- Cleanliness of the house
- Garden: people with good gardens are much more likely to be buying home services.
- Do they have an alarm on their house or not (for alarm companies)
- Existence of solar panels
- Of course, size and cost of comparable property in this region, and auxilary stuff like crime rates.

I.e., if you are a home service business, you may not go blind into regions (and be picky about them); or do more targeted advertising.
Knock on the right doors. Put your doorhangers in the right places.

I've made the numbers, and with a UAV operator+technology costs, data acquisition costs can be down to 5$/20 houses, which can be split to multiple customers... Let's say detailed data of everything about a given house would be sold for 125$/500 houses. Or sell leads directly, just as @MJ DeMarco did years ago. aggregated region data (good/bad street/x mile radius) would be 3x cheaper.
Folks in home services, would this be of any need?

Not rushing into it because I have something else going on I've been developing for over a month, but it's doable and sounds useful. Damn, as a window cleaner, I'd rather never go into bad regions, and always were in good ones.


A trained eye can see dozens of things of interest. One can make the eye automatic!

 

Dislike ads? Become a Fastlane member: Subscribe today and surround yourself with winners and millionaire mentors, not those broke friends who only want to drink beer and play video games. :-)

[Ben_]

O moli vu ati e siem itvevi eqq xovj vji neq gievasi vu tii gus temi / tumf qsodit ug juatit vu heahi vji xiemvj op vji esie O xuamf ci lpudloph op. Op visnt ug wisz apliqv zesft / saccotj O dep neli e raodl fidotoup up vji tquv xjivjis ovt e juati xusvj tloqqoph. (teni opfatvsz, O fuus lpudl gus xopfux dmiepoph)

 

[MRiabov]

O xet vjoploph op qesvodames ecuav cohhis vsefi dunqepoit moli @Kujppz cuz't gus xjun feve dep ci enusvobif uwis meshis fowotoupt/nusi inqmuziit/nusi tiswodit, puv tumuqsipiast.

O moli vu ati e siem itvevi eqq xovj vji neq gievasi vu tii gus temi / tumf qsodit ug juatit vu heahi vji xiemvj op vji esie O xuamf ci lpudloph op. Op visnt ug wisz apliqv zesft / saccotj O dep neli e raodl fidotoup up vji tquv xjivjis ovt e juati xusvj tloqqoph. (teni opfatvsz, O fuus lpudl gus xopfux dmiepoph)

Jnn, vjev't opvisitvoph. Qsivvz ucwouat, pux vjev O vjopl ecuav ov, cav tunijux, O piwis puvodif.