CRE Underwriting: Foot Traffic and Property Data Signals That Actually Matter

Foot-traffic panels and property-data files appear in every CRE underwriting deck. The stress cycles they are meant to anticipate are tracked in the Federal Reserve financial stability report and the Senior Loan Officer Opinion Survey, and usefulness is narrower than sales slides imply. Retail-center underwriters benefit from visit comparables; multifamily underwriters need different signals; industrial Class-B distribution centers benefit little from foot-traffic alone. Strong shops ask which signal moves the credit decision before paying for dashboards. Pair global mobility and POI data with real estate data 201 and CRE due diligence playbook.

Lenders repeating vendor marketing without asset-class fit waste committee time. Signal selection should follow asset taxonomy: what question each metric answers, what refresh cadence credit memos require, and what ground-truth validates panel reads.

Retail and Mixed-Use: Where Foot Traffic Moves NOI

For strip centers, malls, and street retail, mobility panels inside POI polygons benchmark tenant visit trends against comps, detect anchor decline early, and stress-test co-tenancy clauses. Compare same-store visit indices year-over-year with documented panel composition. Traffic without spend context misses monetization: pair with card-spend panels where category spend validates visit quality.

Multifamily: Absorption and Trade-Area Context

Multifamily underwriting uses origin-destination and daytime population patterns more than store-visit counts, layered on public baselines such as HUD multifamily program data. Where do renters work, shop, and commute? Origin-destination for site selection frames the question. Foot traffic to ground-floor retail is supplemental, not primary, for residential credit.

Property Data: Legal and Physical Truth

Lending-side context comes from CFPB HMDA data and CREDA research, with residential-adjacent sales and inventory context from NAR research and statistics. Assessed values, building characteristics, ownership entities, tax status, zoning, and lien priority come from property records: the chassis every memo needs regardless of asset class. Foot traffic cannot fix bad title, undisclosed liens, or wrong square footage. See property data for CRE due diligence and location intelligence for CRE.

Industrial and Office: Limits of Visit Signal

Industrial yards and office towers generate sparse consumer foot-traffic signal relative to retail. Underwriting leans on lease terms, tenant credit, utilization proxies, and market vacancy tracked in NAIOP office and industrial research, not mall-style visit indices. Using retail-calibrated mobility products on industrial assets produces noise dressed as analytics.

CRE Signal Diligence Checklist

Before licensing:

1. Asset-class fit: which signals apply to this property type?
2. Polygon or parcel definition: footprint matches collateral?
3. Panel DUAs inside footprint, not national marketing totals?
4. Refresh cadence: aligns with covenant monitoring frequency?
5. Ground-truth sample: owner-reported traffic or tenant sales where available?
6. Governance: aggregation floors, sensitive-location exclusions, and Fair Credit Reporting Act scope reps where residential-adjacent lending data is in play?

Credit committees should see signal lineage in memos (vendor, geometry version, date range, and known limitations) so challenged loans replay the same evidence.

Visit Signal Lags the Lease, and Underwriting Runs on the Lease

Underwriting decisions turn on contracted cash flow, and visit data describes activity. Those move on different clocks. A tenant can be trading poorly for months while paying rent on time under a lease with years left to run, and a strong visit trend says nothing about a renewal option that will be exercised or declined on economics the sensor cannot observe.

Use visit signal for the questions it can answer, which are about relative performance, trade-area strength, and whether a location is gaining or losing ground against its comparison set. Keep it out of the questions it cannot, which are about contracted income and covenant strength. The useful framing is that visit data changes your confidence in a renewal or a re-letting assumption; it does not replace the lease abstract, and an underwriting model that lets it do so has quietly changed what it is measuring.

The Comparison Set Decides the Answer

Almost every conclusion drawn from CRE visit data is a comparison, which makes the comparison set the most consequential and least examined choice in the analysis. A property benchmarked against a national average will look one way and against its immediate trade area another, and both are defensible until someone selects whichever supports the conclusion already reached.

Define the comparison set before looking at the numbers and write down why those properties belong in it. Then hold it fixed across periods, because a comp set that quietly changes composition produces trends that belong to the benchmark rather than the asset. Where a genuine change is needed, report the old and new basis together for at least one period so the discontinuity is visible.

Choosing the Comp Set Before You See the Numbers

A visit index is a ratio, and the denominator does more work than the numerator. Once the subject asset's polygon is clean, the remaining judgment is which properties it gets measured against, and in practice that judgment usually happens after somebody has already seen a first read. Picking comps once the answer is visible is how a foot-traffic exhibit turns into support for a conclusion the sponsor reached before the data arrived, and a credit committee cannot detect it from the chart.

Write the selection rules first, in the memo, before pulling any panel output. Rules should be properties of the asset rather than properties of the result, which means format and gross leasable area band, anchor category, trade-area density, drive-time profile, category-level tenancy mix, and market. State how many properties qualify under those rules and use all of them, or state the tiebreaker that trims the list. The real test of a rule set is that a second analyst reading only the memo can rebuild the same comp set without asking you, which is also the standard a challenged loan gets replayed against later.

Then document the breaks. A renovation, an anchor replacement, a road or transit project, a competing center opening inside the trade area, and a change in the vendor's polygon version or panel composition each interrupt the series, and any one of them can move an index further than the tenant performance you are trying to observe. Mark the date, say which side of it a given read comes from, and never span a break silently. Period alignment deserves the same discipline, because a partial period compared against a full one, or a year where a holiday shifts across a reporting boundary, produces a difference that has nothing to do with the asset. Where the subject has no clean history at all, from new construction or recent re-tenanting, say so plainly and lean on the comp set and the property record instead of manufacturing a trend line. The CRE due diligence playbook covers the surrounding workflow.