A regional director opens the pipeline and finds three hundred candidate addresses across a dozen markets. Every one of them looks plausible on a map. The question a growing team actually needs answered is not “are these good sites?” but “which five are worth a plane ticket and a committee slot this quarter?” Screening at scale is the discipline of getting from the three hundred to the five without quietly lowering your standards along the way.
The instinct at most enterprises is to add analysts. That works until it does not: headcount grows linearly, the pipeline grows faster, and the review backlog becomes the real constraint on how quickly the brand can expand. The teams that pull ahead treat screening as a funnel with explicit, agreed-upon gates, so the expensive work only ever touches sites that have already earned it.
Screening hundreds of sites at scale means converting your criteria into hard filters, applying them to the whole pool at once, and forecasting only the finalists. Filter first to protect throughput; forecast second to protect rigor. The output is a short, defensible list of sites worth visiting, produced in days rather than months.
The Bottleneck Is Visits, Not Sites
Candidate sites are effectively free to generate. Brokers send them, data platforms surface them, and any market map produces more than a team can ever act on. What is scarce is everything that comes after: a site visit, a full underwrite, a slot on the real estate committee agenda, and the attention of the senior people who sign off. Those are the resources that gate expansion, and they are the ones an enterprise cannot simply buy more of overnight.
That reframes the whole problem. The goal of screening is not to evaluate every site thoroughly. It is to be ruthlessly efficient about which sites reach the thorough stage. A good screen spends almost nothing per site at the top of the funnel and reserves the deep, human, expensive work for the handful that survive. Get that ordering right and a small team can cover far more markets than its headcount would suggest, which is exactly the promise of modern site selection software.
Filter, Then Forecast
The two jobs in screening are different and should not be mixed. Filtering is cheap, fast, and binary: does this site clear a threshold or not? A composite site score, a customer-match floor, a cannibalization cap, and a trade-area population minimum will each knock out large chunks of a pool in milliseconds, and they can run across three hundred sites as easily as three. Forecasting is the opposite: it is expensive, nuanced, and continuous, projecting expected revenue for a specific address against analog stores. You cannot afford to forecast three hundred sites, and you should not have to.
The sequence matters. Filter first to get from hundreds to a shortlist, then forecast new-store sales only on the survivors. Reversing the order is how teams burn weeks and still miss deadlines. It also keeps the filters honest: because they are cheap, you can tune thresholds live and watch the shortlist grow or shrink, rather than arguing about hypotheticals in a meeting.
The model below shows the shape of it. Start with a pool of three hundred synthetic sites and tighten each filter in turn. Watch how a stricter score or a lower cannibalization cap collapses the funnel, and how the final shortlist responds to every move.
Screen 300 sites down
Set your filter thresholds
Keep the Committee Honest
The point of explicit thresholds is not just speed; it is discipline. When the gates are written down and applied to the whole pool the same way, the committee spends its time on the sites that matter instead of relitigating why a favored corner keeps getting championed. Everyone agreed to the score floor and the cannibalization cap before anyone saw a specific address, so a site that fails a gate fails on the rule, not on a personality.
That is what separates a screen from a shortcut. A screen that quietly waives its own thresholds for a pet site is just gut feel with extra steps. The value of a consistent, data-driven process, the kind that pairs a calibrated score with retail location intelligence, is that it makes the standard the same for the tenth market as it was for the first. Rigor that survives volume is the whole game.
From Shortlist to Signed
A shortlist is a beginning, not a verdict. The three to seven sites that clear the screen are the ones that deserve a real forecast, a site visit, and a full underwrite. This is where the human judgment that a model cannot replace goes to work: reading the specific parcel, the traffic pattern at the actual intersection, the landlord, and the deal terms. The screen did not make the decision; it made sure the decision was made about the right five sites.
Done well, the whole path compresses. What used to be weeks of manual review per market becomes a screen that runs in an afternoon and a forecast pass that finishes in days. The committee sees a tight, comparable list with the reasoning attached, approves faster, and the brand moves on the best sites before a competitor does. Speed and rigor stop being a trade-off and start being the same workflow.
Spend almost nothing per site until a site has earned attention, and spend generously once it has. Filters are how you afford to look at everything; forecasts are how you decide. If your process forecasts before it filters, or waives its filters for favorites, it will not survive the jump from three markets to thirty.
Common Questions
- How do I screen hundreds of potential sites quickly?
- Start by turning your criteria into hard filters, then apply them in sequence to the whole pool at once rather than reviewing sites one by one. A site score, a customer-match threshold, a cannibalization cap, and a trade-area population floor will typically cut a pool of hundreds to a shortlist of a handful. The slow, expensive work (visits, underwriting, committee) then only touches sites that already cleared the bar.
- What criteria should I filter sites by?
- The durable filters are a composite site score, how well the surrounding population matches your best customers, how much a new unit would cannibalize existing stores, and whether the trade area has enough population to support your format. Layer in access, co-tenancy, and competition once the pool is smaller. The goal is a small set of thresholds everyone on the committee agrees on before anyone falls in love with a specific corner.
- How many sites should make a shortlist?
- For most enterprise pipelines, a shortlist of roughly three to seven sites per market is the sweet spot. That is few enough to visit and underwrite properly, but wide enough that you are not betting the market on a single address. If your filters leave you with dozens, they are too loose; if they leave you with zero, loosen the least important threshold first.
- How does site scoring work?
- A site score combines many weighted signals (demographics, psychographics, foot traffic, competition, co-tenancy, accessibility, and your own store performance) into a single number that ranks candidates on a comparable scale. Locate calibrates that score against your existing stores so it reflects what actually drives revenue for your brand, not a generic average. The score narrows the field fast; a revenue forecast then does the deeper read on the finalists.