Diablo wind is your new boss (East Bay tree service)

If you run a tree-service crew in the East Bay hills, the late-October Diablo wind events are your peak earnings window — and your worst phone day of the year.

A typical Diablo morning: the wind kicks up sometime around 9 AM. By 11, you're getting calls from every neighborhood north of 580 and east of 24. By 2 PM, your phone has rung 34 times since lunch. You've answered six. Two of those were people on hold for the others. You're standing in a yard in Lafayette with a chainsaw in your hand and 28 voicemails you'll never play because the customers won't have left them.

That's a $30,000+ day's worth of work that walked to whichever crew picked up the most calls.

The Diablo winds aren't getting calmer. PG&E's PSPS shutoffs aren't going away. Insurance pressure on hillside homeowners to clear defensive space is increasing every year. The market is begging for more tree-service capacity in the East Bay hills, and the bottleneck on your business is not crew, equipment, or pricing.

It's your phone.

Concurrent calls are the actual problem

I think this gets glossed over in most "AI receptionist" pitches, so I want to slow down and lay it out.

A normal phone, with a normal human (you, your kid who answers when she's not at school, a part-time receptionist) can handle one call at a time. Maybe two if you're juggling. Three is the upper bound of what one human brain can do without dropping somebody.

A Diablo wind day produces 30, 40, sometimes 60 inbound calls in a four-hour window. Most of them concurrent — the same wind event dropped limbs in Orinda and Lafayette and the Berkeley Hills and Oakland's Montclair district at the same time. Everybody picks up their phone at the same hour.

You cannot answer 30 concurrent calls. Neither can a part-time receptionist. Neither can a regular answering service whose call center has, optimistically, 6 agents available at any given moment to handle dozens of clients.

This is the one specific area where AI is structurally better than every alternative. It handles concurrent calls in unlimited parallel. Three calls in five seconds = three concurrent intake conversations. Eighteen calls in a minute = eighteen intake conversations. The cost of capacity is zero on the AI side.

The hill-fire-region tree-service economy is built on those storm-volume windows. The crew that captures the most of that volume is the crew that wins the season.

The October-November window is the year

This is true of most East Bay tree-service crews, but worth saying out loud for anyone who's still doing this part-time or as a side to a landscaping operation:

Tree-service revenue in the East Bay is not evenly distributed across the year. Roughly 55-65% of annual revenue lands in three specific windows:

1. October-November Diablo wind events. Peak earning window. PSPS-driven defensive-space requirements + active limb-fall + insurance-claim work all stack.
2. Spring storm season (Feb-April). Wet-soil-plus-wind = uprooted hillside trees, soggy-ground falls, tree-on-house insurance work.
3. Late-summer fire-prep (July-September). Insurance-renewal-driven defensible-space work + utility-easement clearing. Lower-emergency, higher-margin scheduled work.

The April-June and December-January windows are mostly steady estimate-and-scheduled-work months, with much lower call volume.

The economics of running a year-round crew depend on capturing as much of those three peak windows as possible. The AI math is roughly:

• Diablo storm event captures: missing 60% of inbound calls vs. capturing 95% changes a $40K storm-event week into a $63K storm-event week.
• Spring storm event captures: same shape, smaller magnitude. ~$10K-$20K incremental per event.
• Fire-prep work: phone-coverage matters less; bigger ticket sizes and longer sales cycles. AI helps but the marginal lift is smaller.

Annualized, the storm-windows alone can move a $400K/year crew to $550K-$650K/year if the phone problem is solved. Same trucks, same crew, same pricing. Just answered calls.

What the East Bay tree call actually needs

Tree-service intake is more complex than most trades I write about. The AI configuration needs to cover:

1. Emergency vs. estimate triage. "There's a tree on my house" or "There's a limb hanging over the power line" is a same-day or next-morning emergency dispatch. "I want to get a quote on removing three trees in my backyard" is a scheduled estimate visit. Different urgency, different pricing, different crew dispatch.
2. Fall risk and access. Steep hillside, access road condition, proximity to the house. The crew needs to know what equipment to bring. The AI captures this on intake.
3. Tree species and rough size. "Eucalyptus, about 60 feet" is a different job than "redwood, 80 feet" or "oak, 35 feet." Pricing, equipment, and crew time all vary.
4. Insurance involvement. "It fell on my house, I have a State Farm claim open" is a different conversation than "I want to take it down before it falls." Different paperwork, different timing.
5. PG&E and utility easement context. Anything within 10 feet of utility lines triggers a different process. Anything that requires PG&E shut-off coordination is a multi-day scheduling problem.
6. PSPS / fire-season urgency. Defensible-space work has different timing pressure during active PSPS notification periods.

A generic "AI receptionist" can't handle this. The CTI configuration for tree-service shops is trained on each of the above and runs the right branch of the conversation based on what the caller says first. Average call length: 90-120 seconds. Most callers don't realize they're talking to AI until you tell them later.

The "she's asking the right questions" moment

I want to share a specific moment that comes up in our tree-service deployments because it tells you what good intake actually feels like to the caller.

Customer in the Oakland Hills calls. Says "I have a big eucalyptus that's leaning toward my neighbor's roof." The AI's response is roughly:

"Yeah, that sounds urgent. A few quick questions so I can get you on the schedule fast. First — is the lean from a recent windstorm or has it been gradually leaning over months? Second — how close is the neighbor's roof, in feet? Third — is there any visible cracking at the base or any roots that have come up? And fourth — is your address on a hillside road or flat-access? That last one tells our crew what equipment to bring."

That sequence of four questions is what experienced tree-service estimators ask. It feels exactly like calling a knowledgeable shop. It captures the answers our crew actually needs to dispatch correctly. And it lands as text in your phone within ninety seconds of the call ending.

The customer hangs up thinking they got a real, knowledgeable office on the phone. Whether they got a human or an AI — most never know.

What the AI declines

Tree service is one of the trades where being polite about what you don't take actually matters.

• Pure stump grinding without removal: many crews don't do these as standalone jobs. AI politely redirects.
• Out-of-area work: AI knows your service radius. Calls from Brentwood, Antioch, or Concord hills get the right response based on whether you actually go out there.
• Sub-50-foot hedges and ornamental trim: this is landscaper work, not tree-crew work. AI sends those callers elsewhere with a brief explanation.
• Stump-only fire-mitigation work: usually low-margin, owner-installed-disposal work. AI declines politely and saves your team time.
• Solicitor calls from insurance adjusters or wholesale wood buyers: AI handles the intake, doesn't transfer to you, lets you call back when you actually have time.

You'd be amazed how much owner-time gets eaten by calls you'd never have taken anyway. AI handling the no's is one of the underrated wins.

The math on Growth, with storm-day weighting

Our Growth bundle runs $597/mo. Tree-service economics aren't about the average month — they're about the storm windows. Let's run the math both ways.

Average month (non-storm): ~6-10 inbound calls/week, average closed-job value ~$1,200, conversion if answered live ~45%. Recover 5 missed calls/month × 45% × $1,200 = $2,700/mo incremental. Plan pays for itself with one recovered emergency removal.

Storm-week math: 30-60 inbound calls in a 4-hour window. Without AI, you answer maybe 15% live. With AI, you answer 95%+ and capture intake on every one. The delta in a single storm event can be $15K-$40K in recovered bookings, depending on your average ticket and crew capacity.

Annualized across two-to-three storm windows + steady scheduled work: $30K-$80K/year of incremental revenue at a $597/mo plan cost ($7,164/year). Even on the low end, that's a 4x ROI on the line item. On the high end, it's the difference between a one-truck and a two-truck operation.

What changes when we set this up

Setup runs under 48 hours from "yes." First step is a 30-minute discovery call where I learn how your specific crew runs intake — your service area (Berkeley, Oakland, Lamorinda, Tri-Valley, hills only or flats too), your storm-event protocols, your insurance work mix, your PG&E coordination — and configure from there. We test on real calls before handing over.

Then on the next Diablo wind morning, when 34 calls hit your phone in three hours and you're standing in a Lafayette yard with a chainsaw, three things happen at once:

The AI answers all 34. Each customer gets a knowledgeable voice walking through the right intake. Each call lands in your dispatcher's queue with full context — emergency vs. estimate, location, species, lean, access, insurance involvement.

You finish your current job. By the time you wash up, your crew lead has already routed the next eight emergency calls and put the rest on the schedule for next week.

Hear the demo on the homepage, or book a 15-minute walkthrough and I'll show you exactly how we'd configure this for an East Bay tree-service operation.

The Diablo winds are coming. They always are. The only question is who picks up when 30 phones ring at once.