What If You Could Actually Explain the Score?
Picture yourself in an expansion committee meeting. You're recommending a location. Someone asks why this site scored an 84.
And you can answer.
Not with "the model says so." Not with a vague reference to the algorithm. You pull up the breakdown: foot traffic scored 91 because the adjacent grocery anchor draws 14,000 weekly visits. Demographics scored 78 because the median household income matches your top-performing stores within 4%. Competition scored 72, which is solid but flagged because two direct competitors sit within a seven-minute drive. Visibility came in at 68 because the site is set back from the road with limited signage options.
That's a conversation a committee can work with. That's a site score framework built on lenses, not a black box.
I wrote recently about why opaque site scores are a liability in committee. This article is the other side of that argument: what transparent scoring actually looks like when you break it into components.
At GrowthFactor, we score every site across five lenses. Each one measures a distinct dimension of location quality. Each one produces its own score with a plain-language justification. And each one can be weighted differently depending on how your business actually works.
Here's the framework.
Why a Single Number Falls Short
A composite site score is useful for triage. If you're evaluating 200 locations for a franchise rollout, you need a way to sort. A number does that.
But a number without context creates two problems.
First, it hides tradeoffs. A site that scores 82 might have incredible foot traffic and terrible visibility. Another 82 might have the opposite profile. The total looks identical. The risk profiles are completely different.
Second, it blocks the conversation. Expansion committees exist to pressure-test recommendations. When the score is a single number from an opaque model, the committee has two options: trust it or reject it. Neither is a good use of anyone's time.
ICSC research found that more data hasn't made retail site selection decisions easier. More data without explanation has made them harder. The solution isn't less data. It's data organized into dimensions you can evaluate independently.
That's what a lens-based framework does. Instead of one score, you get five. Each one tells you something specific. Together, they tell you whether a site works, and why.
The 5 Lenses, Explained
Every site analyzed on GrowthFactor receives a score across these five lenses. Each lens has its own 0-100 rating, its own data sources, and its own justification. The lenses combine into a composite score, but the individual lens scores are where the real insight lives.
Lens 1: Foot Traffic
What it measures: How many people move through the area around a site, and where those visits come from.
Foot traffic is the most intuitive lens. Retail depends on people showing up. This lens quantifies that by looking at pedestrian activity patterns, visit counts at nearby points of interest, and the general flow of movement around a location.
The data comes primarily from mobile device signals aggregated by providers like Unacast. These signals show visit frequency, dwell time, and cross-visitation patterns at nearby businesses. A site next to a grocery anchor drawing 14,000 weekly visits has a different foot traffic profile than one in an office park that empties at 5pm.
A high foot traffic score means the area generates consistent, relevant visits. A low score doesn't automatically disqualify a site, but it shifts the burden: if people aren't already coming to this area, your location needs to be a destination on its own. That works for some formats (furniture stores, medical clinics) and fails for others (coffee shops, convenience retail).
When this lens matters most: QSR, convenience retail, impulse-purchase categories. Less critical for destination businesses with loyal customer bases that will drive 15 minutes regardless.
Lens 2: Demographics Fit
What it measures: How well the people living in the trade area match the customers who actually spend money at your stores.
This is where site scoring gets brand-specific. A gym chain targeting 25-to-34-year-olds with household incomes above $75,000 needs a different demographic profile than a dollar store serving price-conscious families. The demographics lens scores the match between the trade area population and your proven customer profile.
The data comes from ESRI demographic datasets: population by age bracket, household income distribution, education levels, household composition, and consumer spending indices. We overlay these with your existing store performance data. The stores where you're actually making money tell us who your real customer is, which often differs from who marketing thinks the customer is.
Customization matters here. A site in a college town might score high on demographics for a fast-casual brand targeting young professionals, while the same location scores poorly for a home improvement retailer. Same data, different weights, different score. This is why scoring frameworks that apply the same model to every brand miss the point.
Lens 3: Market Potential
What it measures: Whether the market around a site is growing, stable, or contracting.
Foot traffic and demographics tell you what the area looks like today. Market potential tells you where it's headed. This lens looks at forward indicators: population growth projections, new residential and commercial development permits, employment trends, and economic momentum signals.
A site in a trade area with 8% projected population growth over five years has a different risk profile than one where the population is flat. Both might score well on foot traffic right now, but five years into a 10-year lease, the growth market is compounding while the flat market is eroding.
This lens catches emerging markets before they show up in foot traffic data. A new 500-unit residential development three miles from your site won't generate foot traffic for 18 months. But it will change the demographics and demand profile of the trade area permanently. Market potential is the lens that makes site selection forward-looking instead of backward-looking.
Lens 4: Competition Analysis
What it measures: The competitive density around a site, and what that density signals about demand and saturation.
Competition is the most misunderstood lens. Most people assume more competitors equals worse. That's too simple.
A cluster of direct competitors near your site can mean two things. In a high-traffic corridor, it signals strong demand: people already come here for what you sell, and there's room for one more. On a declining strip where three competitors are all struggling, it signals saturation: the market can't support the existing supply, let alone a new entrant.
This lens evaluates competitor proximity, visit counts at competing locations, and the ratio of supply to estimated demand. For multi-unit operators, it also includes cannibalization analysis: would this new location pull customers from your existing stores, or would it capture net-new demand?
The cannibalization piece matters more than most teams realize. We've seen operators open locations that looked strong in isolation but pulled 30% of revenue from a store eight minutes away. The individual site scored well. The portfolio impact was negative. Competition analysis at the lens level catches this before you sign the lease.
Lens 5: Visibility
What it measures: How easy the site is to see, reach, and access from the road network.
Visibility is the most physical of the five lenses. It accounts for road frontage, signage opportunities, traffic counts on adjacent roads, ingress and egress patterns, and the general accessibility of the site.
The data comes partly from StreetLight vehicle traffic analytics, which provide daily volume counts and road type classification. But visibility also depends on physical characteristics that data alone can't fully capture: is the site set back 200 feet from the road? Is there a turn lane? Can drivers see the signage from both directions at 45 miles per hour?
This is where the analyst layer adds value beyond the score. A visibility score of 65 tells you the site has moderate exposure. The analyst report tells you whether that's because of a fixable signage issue (landlord willing to add a pylon sign) or a structural problem (the building faces away from traffic and sits behind a berm). One is a negotiation point. The other is a dealbreaker.
How the Lenses Combine Into a Score
Each site receives a composite score on a 0-100 scale with four grade bands:
| Grade | Score Range | What It Means |
|---|---|---|
| Great | 80-100 | Strong across most or all lenses. Worth serious evaluation. |
| Good | 70-79 | Solid overall with one or two areas that need scrutiny. |
| OK | 60-69 | Mixed signals. Proceed only if weak lenses are fixable. |
| Bad | Below 60 | Fundamental issues. Rarely worth pursuing unless data is missing. |
The composite isn't a simple average. Lenses are weighted based on what actually drives performance for your brand. A QSR chain might weight foot traffic and visibility at 25% each, demographics at 20%, competition at 20%, and market potential at 10%. A fitness chain expanding into suburban growth corridors might weight market potential and demographics higher and visibility lower (their members drive to them regardless of signage).
These weights aren't fixed. When GrowthFactor onboards a new customer, the analyst team works through the weighting with you. We test different configurations against your existing portfolio. The weights that best predict where you're already succeeding become the weights for evaluating new sites.
Every lens score comes with a justification. Not a number and a grade. A sentence explaining what drove the score. "Growing suburban population offers strong match for value retailers" is more useful than "Demographics: 78." The justification is what you bring to committee.
What Makes This Different From a Black Box
Three things separate a lens-based framework from the opaque models that dominate this industry.
You see each component, not just the total. When a site scores 74, you know exactly which lenses are strong and which are weak. You can make an informed decision about whether the weak areas are acceptable, fixable, or disqualifying. A black box gives you 74 and lets you guess.
You can ask why. Every lens has a justification. Every justification traces to specific data. If you disagree with the competition score, you can look at the competitor list and visit data and challenge it. With a proprietary model, there's no mechanism to question the output.
You can change the model. This is the part competitors don't offer. At GrowthFactor, the analyst team builds scoring models collaboratively. We walk through every variable and weighting across multiple meetings. If you believe drive-through access matters more than general visibility for your brand, we test that hypothesis and adjust. If the data proves you wrong (it sometimes does), we show you the evidence. Either way, you understand the model because you helped build it.
That collaborative process is what we mean by "Glass Box." Not just showing you the five lens scores, but building the model with you so you never have to take a number on faith.
One example: a frozen dessert brand hypothesized that stores selling more pint varieties would perform better. It seemed logical, so they wanted pint mix weighted heavily in the forecasting model. We built the custom model, ran the numbers, and proved the correlation wasn't there. Pint mix wasn't a significant performance driver. Without that test, they'd have chased the wrong variable across dozens of new locations.
Legacy platforms don't work this way. The typical process: a vendor builds a model over six to nine months, delivers it with minimal explanation, and updates it rarely. Their customers go to committee holding a forecast they can't explain. That's not analysis. It's an expensive guess with better production values.
How to Read a 5-Lens Report
If you're working with a lens-based scoring framework (or evaluating whether to adopt one), here's how to get the most out of it.
Start with the composite for triage. When you're sorting 100+ potential sites, the total score is your filter. Below 60 goes to the bottom of the stack. Above 80 goes to the top. The composite earns its keep here.
Then open the lens breakdown for anything serious. Once you're down to 10-15 candidates, stop looking at the total. Look at which lenses are strong and which are dragging. Two sites can both score 78, but one might have broad consistency (all lenses in the 70s) while the other has extreme variance (95 on foot traffic, 55 on competition). Consistency is generally lower risk.
Ask whether weak lenses are fixable. A low visibility score caused by missing signage is a negotiation point with the landlord. A low demographics score caused by the wrong income distribution in the trade area is structural. You can't change who lives there. Fixable weaknesses are costs. Structural weaknesses are risks.
Compare sites lens by lens, not just on total. If you're choosing between two sites and both score 81, compare the lens profiles. Which tradeoffs does your team prefer? A site with higher foot traffic but lower market potential is a near-term play. The reverse is a growth bet. Neither is wrong, but they serve different strategic purposes.
Cross-reference with an analyst deep dive before committee. The five lenses are the quantitative layer. For anything going to the expansion committee with capital attached, add the qualitative layer: a site visit assessment, a conversation with the broker, a look at the lease terms. The lenses tell you whether a site is worth investigating. The analyst tells you whether it's worth signing.
FAQ
What is a site score in retail real estate?
A site score is a composite rating that quantifies how well a potential location fits a retailer's expansion criteria. Scores typically range from 0-100 and factor in demographics, foot traffic, competition, visibility, and market growth. The value of a site score depends on whether you can see how it was calculated. An opaque score from a black-box model is harder to defend in committee than one broken into transparent components.
How many factors go into a site selection score?
It depends on the framework. Some models use hundreds of variables. GrowthFactor organizes site evaluation into five primary lenses (foot traffic, demographics fit, market potential, competition analysis, and visibility), each drawing from multiple data sources. Within each lens, dozens of individual data points contribute to the score. The five-lens structure makes the output interpretable without sacrificing analytical depth.
Can you customize scoring weights for different retail formats?
Yes, and you should. A QSR chain cares more about foot traffic and visibility than a furniture retailer does. A fitness brand expanding into suburbs may weight market potential and demographics higher. At GrowthFactor, scoring weights are set collaboratively with each customer based on what actually predicts performance in their existing portfolio. Fixed-weight models that apply the same formula to every brand miss format-specific drivers.
What's the difference between a glass box and a black box site score?
A black box site score gives you a number without showing how it was calculated. A glass box site score shows every component: which lenses contributed, how they were weighted, and why each lens scored the way it did. The practical difference shows up in committee. A glass box score lets you answer "how did you get that number?" with specifics. A black box score forces you to say "the model says so," which rarely satisfies a room full of people about to approve a million-dollar lease.
Making the Score Work for Your Team
A site score should be the beginning of a conversation, not the end of one. The five-lens framework exists to give your team a shared language for evaluating locations: specific enough to be useful, transparent enough to be questioned, and flexible enough to match how your business actually works.
If you're currently working with a single opaque score, or building site recommendations from spreadsheets and gut feel, the gap isn't more data. It's structured data you can defend.
The full argument for why opaque scores fail is worth reading alongside this piece. Together, they lay out both the problem and the framework.
See how GrowthFactor's 5-lens scoring works for your portfolio.