Marketing Agent
You do the work. Your agent reads every marketing signal you have and finds the few campaigns actually worth running, with the cost-benefit worked out.
Suggested campaign
3rd this month · waiting on your call
🏘️
Maple Grove waterproofing
Neighbourhood campaign · built from the Ellis Ave job, closed Tue
Scheduled · Tue 6 pm
Tell-me mode
You're in control. This goes out Tuesday at 6 pm. Review, edit, or hold it anytime before then, and you can undo after it sends. Nothing is locked in.
Before💧
After🛡️
Drywise Basement · Maple Grove
Your street just took on a lot of water
We just finished a full waterproofing and sump system two doors down on Ellis Ave — dry through the whole storm. If your basement smells damp or you've seen pooling, it's worth a 15-minute look before the next one. We're in Maple Grove this week.
1,200 homes across Maple Grove · Facebook + email + mailer · sends Tue 6 pm
See it before it sends:
Add an incentiveOptional — stand out when neighbours are comparing quotes
AI
Agent: No discount needed — your reviews on this block are strong and demand is high. A free inspection converts browsers without touching price.
Held. Nothing sent. Your agent logs that you passed and factors it into what it suggests next.
Projected cost vs. return
Estimated cost
~$1,600
Facebook ~$450 · postcards ~$1,150 · email $0
vs
Projected revenue
~$12,000
4–5 jobs · avg ~$2,700
≈7.5×on spend
How it gets there
1,200Homes
~1.4%Click-through
~9Conversations
4–5Jobs
~$12kRevenue
Replies land in your AI Receptionist — answered and booked 24/7. Modeled against your April campaign (~6× on spend). That's a ~0.4% booking rate, normal for local outreach — the win is the low cost per job.
How it built thisEvery marketing signal you have, combined into one call
Combined:
ReviewsJob historyShowcase photosOnline presenceGoogle profileSeasonalityWeatherNeighbourhoodPast campaigns
🏅What you're known for
You list nine services. Your record says basement waterproofing — so that's the lead.
Waterproofing jobs (12 mo)61%
5★ reviews naming "basement"34
Fastest-paid serviceWaterproofing
🌧️Why now
Three timing signals lined up this week.
7-day rain vs normal3.4×
Spring thaw~2 wks out
City flood subsidyActive
LIVE · Toronto
Checking conditions…—
Pulling current weather…
📍Why these homes
Maple Grove is older housing on a low grade. The radius covers the block that shares the risk.
Homes in radius (500 m)1,200
Pre-1980 builds (higher risk)~900
Basements you've done nearby3
Your last campaign
1.1%Click-through
6Conversations
3Jobs booked
$7,400Revenue
WorkedCost $1,150 → $7,400 booked · ~6× on spend
Folded into this one
↑
Photo beat text-only 3:1 → this one leads with the photo.
↑
Tuesday evening beat weekends → scheduled Tue 6 pm.
↓
10%-off didn't lift conversion → no discount this time.
Jobs held back
7 this monthClosed jobs the agent chose not to marketFinding the campaigns worth running means knowing which jobs aren't worth one. Of the 10 jobs you closed this month, 3 became campaigns and 7 didn't. Here's each one the agent set aside, and why.
⚠
Elm St — 5 homes that floodedSame storm as the campaign · water damage reported
Storm damage
Marketing to families currently bailing out their basements is chasing an ambulance, not offering help. All 5 held — flagged for a personal, no-pitch check-in from you instead.
↩
Danforth — warranty call-backReturn visit on last month's work
Warranty fix
A call-back on your own work isn't a win to showcase, and neighbours don't need to know one needed a second visit. No marketing generated.
🤝
Ms. Okafor — invoice in disputeStrong job · bill contested
Bill unsettled
Featuring a client's home while their bill is contested would sour a fixable situation. Held until the invoice settles, then reconsidered.
🔕
The Bergers — opted out of marketingKentworth Ave · privacy preference on file
Opted out
This client turned off marketing outreach in their profile. The agent respects the setting without being asked — consent isn't something it re-litigates each time.
🏢
Scarborough — commercial unitProperty-management contract, not a home
Not residential
A neighbourhood campaign to nearby homeowners is the wrong play for a commercial job. Different buyer, different pitch — held out of the residential campaign.
🔁
East York — repeat client, just servicedThird visit this quarter
Contacted recently
You served this street twice already this quarter. Another outbound this soon reads as noise, not service. The agent spaces contact instead of maximizing it.
📷
Weston — no before/after on fileJob closed without photos
Nothing to show
Your last campaign proved the photo does the work. This job closed without one, so there's nothing strong to lead with. Flagged you to add photos next time rather than send something weak.
How your agent gets sharper
The more it runs, the better it targetsWhat it's learned about your business
🏅
Your money-maker is waterproofing61% of jobs, your best reviews, fastest paid
📸
Before/after photos sell3× the click-through of text-only
🕕
Tuesday evenings land bestBeat weekend mornings 2:1 on opens
🏷️
Discounts don't move youNo lift on conversion, so it stopped suggesting them
📍
Older, low-lying streets convertWhere your waterproofing wins cluster
💧
Rain is your best triggerStorm-week campaigns book 2× a calm week
Campaigns learned from
What never changes
You keep the send button. It gets smarter about what to suggest, never about acting without you.
C
Simulated data, except the live Toronto weather reading in "Why now" (Open-Meteo). Interface built on Jobber's public Atlantis design tokens.
Concept prototype — Corey Brown
Exploring Jobber's Marketing Agent · not affiliated with Jobber
Exploring Jobber's Marketing Agent · not affiliated with Jobber