A home improvement company: every ad dollar traced to sold jobs
A home improvement company in the United States spends heavily on Google, Meta, TV, direct mail and lead vendors. Its CRM showed where leads came from but not what they cost, so we built a dashboard that follows each dollar from ad spend to sold jobs and profit.
- Industry
- Home improvement, in-home sales
- What we built
- Marketing dashboard and four voice agents
- Sources
- Salesforce, Google Ads and Meta first, then the rest month by month
- Status
- Built, rolling out in 2027
What the dashboard does
Before, spend sat in the ad platforms and in spreadsheets, and the CRM held leads and sales but no spend. Each week, reports were pulled together through webhooks and pivot tables, with TV and direct mail spend typed in by hand. Leadership could see cost per lead, but not which sources produced sold jobs and profit.
Spend to sold jobs
Each source is followed from spend to leads, appointments, sold jobs and profit, so budget is judged by what it sells, not by what a lead costs.
Every source in one view
Google, Meta, TV, direct mail and the lead vendors sit side by side, by market, product and week, with budget against actual.
Views by role
Executives see totals and trends. The marketing team sees campaign and vendor detail.
Less manual work
The ad platforms connect directly. The only typing left is TV and direct mail spend, a few lines once a week.
Inside the dashboard
Spend, calls and estimated pipeline for each channel on one screen, with the next actions worth taking.
Illustration with sample data.
How we got here
We worked with the owner, the CEO and the head of marketing over several months before anything was built.
- It started with AI voice agents for lead follow-up. They already answered new leads within a minute, so the bigger gap was knowing which spend produced sold jobs
- Every marketing source, and how it connects, was listed before we quoted
- Each number on the screen has an agreed formula, source field and owner
- The highest-spend sources connect first, and the rest are added month by month
- Current numbers are recorded as a baseline before anything changes, so results are measured, not estimated
- Next: a weekly summary written by AI, with recommendations the marketing team can act on
Every call lands in one place
The same hub runs four voice agents, one for each kind of call. Each call is logged with its source, a summary and the next step, with a button to send it to Salesforce after review.
Illustration with sample data.
Security and ownership
The company owns the code, the accounts and the data.
- Named users with multi-factor sign-in and role-based access
- Read-only access to the ad platforms and CRM, so nothing writes back
- Data freshness and source health shown next to the numbers
- Customer names, phone numbers and call recordings stay out of the dashboard unless the company approves them
More work
Tell us how your business runs. We'll tell you what to build.
In 30 minutes we'll walk through your process, point out where software would actually help, and give you a fixed price for the first build.
Book a 30-minute call