Brief it once. Every site comes back with its reasons.
A Custom Evaluator is an agent that follows criteria your team writes in plain English. Pick the data it reads, and it rates each site on the scale you set and writes out why.
Write it like you’re training a new analyst
A name, a purpose, the instructions in plain words, and the data it’s allowed to pull, each from as far out as you set. There are no weights and no sliders. The rating is whatever scale you ask for.
- Demographics
- Vehicle traffic
- Foot traffic
- Nearby businesses
- Cannibalization
- Analog stores
- Zoning
- Web search
Morning Commuter Fit
Instructions
We sell espresso and pastries to people on their way to work. Walk-ins are most of our sales, and most of those happen before 10 AM on weekdays.
Score the site from 1 to 5.
Weekday morning foot traffic matters most. Weekend-only traffic doesn't count for much.
Offices, transit stops and gyms within a 5 minute walk help. An independent specialty coffee shop on the same block hurts. A national chain nearby is fine.
If the trade area overlaps one of our existing stores by more than 25%, cap the score at 2.
Name the existing store this site looks most like, and say which data you trusted least.
Ask a different question, get a different answer
Same address, two sets of criteria. The Foot Traffic evaluator gives 700 Mass Ave a 3. Morning Commuter Fit gives it a 4, because it was told what a weekday coffee run needs. Keep one evaluator per format or question, and set one as the default.
A hundred broker sites a week. Every one scored the same way.
Every site you add gets the same evaluator. The weak ones sink with the reason written down, and the week goes to the few at the top.
Change the criteria, rerun, keep both answers
Each run is saved to the location with its time and verdict, so the team can see what changed and why. Study PDF exports carry the history into the committee packet.
Measured against the stores you already run
An evaluator can read your analog stores, so a site is judged against its closest matches in your own portfolio, each ranked by how alike they are.
“With GrowthFactor we’ve been able to expand much faster, make quicker decisions, we don’t have to dig.”
2,000+
Sites Cavender’s has run through it, in markets old and new.
$2M
Roughly what Cavender’s kept by passing on three weak sites the analysis flagged.
Before you write the first one
No. It’s an agent that follows criteria your team writes in plain language. It reads the data sources you picked, rates the site on the scale you set, and writes out why, including which data it trusted least if you ask it to.
Workspace admins write them, and we help you set up the first ones at onboarding. Everyone on the team can run them, from the site panel or by asking the Agent.
The evaluator tells you. Its verdict names the data it trusted least and weighs the rating accordingly.
No. An evaluator judges fit against your criteria. Sales forecasts are built on your own stores in GrowthFactor Labs, and they run in the platform next to your evaluators.
Enterprise and Labs. Book a demo and we’ll build one on a site you’re weighing.
Bring a site and your criteria
Live in a daySOC 2 compliantMIT Sloan Most Disruptive MBA Startup, 2025