Gerrit Huizinga (@gerrith777)

I run AI inside a home-improvement and outdoor-living company (indoor window coverings, awnings, motorized screens, pergolas). Most AI advice ignores businesses with crews, installs, and trucks. So I'm documenting the workflows I'm building for one: - estimating from photos - lead follow-up - install scheduling - fleet checks - review requests - owner dashboards One breakdown a week. Follow if you run a home-improvement business.

199

Gerrit Huizinga · Nov 13, 2026 · 12:02 PM UTC

@gerrith7774h

Which job would you never let AI do alone? Mine: texting a customer without a permission check, and changing ad spend. Both wait for a person's yes. Dollars: not measured. Start Monday: write down your two, and name the person who says yes to each.

Gerrit Huizinga · Nov 13, 2026 · 1:01 AM UTC

@gerrith77715h

Sort what Google or an AI tool could cite about you into four buckets: Own: your website Claim: your Google Business Profile, directories Earn: local news, forum threads Contest: wrong info Dollars: not measured. Start Monday: claim your profile, fix wrong info.

Gerrit Huizinga · Nov 12, 2026 · 4:03 PM UTC

@gerrith777Nov 12

Customers call when a promised update never came. 7 touchpoints: welcome, measure, order + ETA, install confirmed, day-before, demo walkthrough, 7-day check-in. In testing, 40 jobs: 10 overdue. Worth: count "where's my order?" calls. Start Monday: send the order and ETA (#3).

Seven numbered touchpoints per job: welcome and expectations page, measure booked, order placed and ETA shared, install confirmed, day-before reminder, demo walkthrough, 7-day check-in. @gerrith777

ALT Seven numbered touchpoints per job: welcome and expectations page, measure booked, order placed and ETA shared, install confirmed, day-before reminder, demo walkthrough, 7-day check-in. @gerrith777

ALT Morning digest from 40 jobs: overdue 10, six shown with job numbers, the missed step, and days late (12 to 36 days). @gerrith777

Gerrit Huizinga · Nov 12, 2026 · 12:03 PM UTC

@gerrith777Nov 12

Square-foot pricing hides jobs that lose money. In testing, a flat $/sq ft left a 2nd-story awning thin. Start Monday: cost the next $5K+ quote (materials, crew hours, drive, truck), then add a reserve. BLS median, construction laborers: $22.66/hr (May 2025, public estimate).

Gerrit Huizinga · Nov 11, 2026 · 11:30 PM UTC

Most "dead" quotes aren't dead. They're unpaid follow-up. In testing, 150 CRM rows, nothing sent:

more replies

Gerrit Huizinga · Nov 11, 2026 · 11:31 PM UTC

Sorting replies by keyword failed on fresh wording (6/15). So for now, any unclear reply pauses the follow-up and goes to a person.

Gerrit Huizinga · Nov 11, 2026 · 11:31 PM UTC

Discounts come last. Close-out is a real goodbye, then stop. Score it: old quotes that reopen × close rate × gross profit, minus the cost of the hours and the tool. Start with last quarter's open quotes.

Gerrit Huizinga · Nov 11, 2026 · 4:05 PM UTC

One landing page for every ad is a leak I check on a phone: 1. Headline matches the ad 2. Stars in the first screen 3. Booking up top 4. Promised counts match the screens 5. Tap-to-call 6. One page per service Dollars: not measured. Start with your biggest ad's page.

Gerrit Huizinga · Nov 11, 2026 · 12:03 PM UTC

If someone runs your website, ask them for this. Link each blog post to your service pages just below Google's top spots. Max 4 links a post; a person signs off. In testing, a script did all 6 posts in about 5 minutes: 12 links, 7 to those pages. Worth: track calls from them.

Gerrit Huizinga · Nov 10, 2026 · 12:02 PM UTC

@gerrith777Nov 10

Cost per lead is a vanity number for home-service ads. A cheap click that never books a measure still spends your money. I watch cost per booked appointment instead. What do you kill ads on?

Gerrit Huizinga · Nov 9, 2026 · 11:30 PM UTC

An "instant price" that only asks product and size will surprise people later. Sandbox test, 200 fake final quotes against a fake price book:

Checklist card: before, the instant price asks only product and size. After, add three questions: fabric tier, second story, outlet nearby. Sandbox label. @gerrith777

ALT Checklist card: before, the instant price asks only product and size. After, add three questions: fabric tier, second story, outlet nearby. Sandbox label. @gerrith777

Grouped bars, sandbox test of 200 fake final quotes: inside the range 65% with product and size only vs 78.5% with the three questions; sticker shock 25% vs 9.5%. Sandbox label. @gerrith777

ALT Grouped bars, sandbox test of 200 fake final quotes: inside the range 65% with product and size only vs 78.5% with the three questions; sticker shock 25% vs 9.5%. Sandbox label. @gerrith777

Gerrit Huizinga · Nov 9, 2026 · 11:30 PM UTC

No site questions: 65% of finals landed inside the range. Add three: - fabric tier - second story? - outlet nearby? Now 78.5% land inside. Sticker shock fell from 25% to 9.5%. Three questions beat a prettier page.

Gerrit Huizinga · Nov 9, 2026 · 5:15 PM UTC

Your cheapest lead can be your most expensive appointment. Sandbox test, fake Meta export, 12 ads: The worst ad by cost per booked appointment ranked #3 by cost per lead. Rank ads by booked appointments, not leads.

Bar chart: five sandbox ads ranked by cost per lead vs cost per booked appointment. Spring promo is #3 by CPL at $12 but worst by CPBA at $224. Callout: Cheap leads. Most expensive bookings. Footer: Sandbox test, fictional company, made-up data. @gerrith777

ALT Bar chart: five sandbox ads ranked by cost per lead vs cost per booked appointment. Spring promo is #3 by CPL at $12 but worst by CPBA at $224. Callout: Cheap leads. Most expensive bookings. Footer: Sandbox test, fictional company, made-up data. @gerrith777

Gerrit Huizinga · Nov 9, 2026 · 2:49 PM UTC