A tier-3 ED office may see dozens of projects a year: state-forwarded RFIs, broker inquiries, utility-channel leads, direct company contacts. Most go nowhere for that city. The office declines to respond, submits and gets cut in the first round, or never hears back. Whatever the staff learned about why the project didn't fit ends up in an email thread, a conversation note, or nowhere.
Most projects won't fit most cities. That part is ordinary. What disappears with each one is a specific reading of the gap between what the city has and what the market asked for, and unrecorded, those readings accumulate where nobody can see them.
Take a director who loses three manufacturing prospects over 18 months because the POTW can't issue an industrial pretreatment permit for metal finishing. That is one infrastructure constraint surfacing three times. If the losses live in three separate email chains with three different state project managers, it looks like bad luck.
The 2026 Guild/DCI pulse survey found 61% of responding site selectors naming utility or infrastructure capacity as a current elimination factor. Another 53% named suitable sites. Those are the dimensions where tier-3 cities get cut most often, and they're also the dimensions most likely to repeat across projects, because the underlying constraint doesn't change between RFIs. A structured record of disqualifications makes the repetition visible and converts it into evidence an infrastructure budget holder can evaluate.
Who needs this most
The urgency varies by asset profile, and so does what the record will show.
A legacy industrial city sitting on a base of 1950s-era manufacturing sites will accumulate losses clustered around environmental status. Phase I flags unresolved. Brownfield designations without cleanup plans. Stormwater permits that don't cover the discharge profile a modern process requires. The capital priority those entries point to is remediation and site certification.
An agricultural-corridor city with strong logistics access and a thin labor shed will see losses cluster elsewhere. Projects screen it in on acreage and highway proximity, then screen it out on workforce depth: insufficient annual completions in industrial maintenance, no mechatronics program within 60 miles, a labor-force participation rate that can't support a 300-employee ramp in 18 months. That capital case runs through the community college and the workforce board, which is a different set of institutions on a different funding calendar.
A growing exurban county next to a metro will lose projects on utility capacity its growth trajectory hasn't yet funded. Three-phase power availability in the industrial corridor. Water pressure and fire flow at the spec site. Treatment headroom at a POTW already running at 85% of permitted capacity. Here the record gives the county engineer specific demand profiles to plan against rather than a general sense that capacity is tight.
Each profile produces a different capital argument. The record assembles it before the director has to reconstruct it from memory in a budget hearing.
What to capture per entry
A useful entry has eight components.
Project profile. Sector, approximate employment range, approximate capital investment range, facility type (new construction, existing building, expansion), and any known operational requirements: power draw, water consumption, discharge profile, acreage, rail access, workforce certifications. You won't always have all of it. A useful entry: "food-grade cold storage, 200,000 SF, 150 jobs, 4MW power requirement, rail-served." A weak entry: "distribution project, large."
Source channel. How the project reached you. State-forwarded RFI, utility lead, broker inquiry, direct company contact, regional partnership referral. This matters for weighting, addressed below.
Failed condition. The screening dimension that eliminated the city. Most important field, most often recorded poorly.
A useful entry: "eliminated on wastewater — POTW cannot issue categorical pretreatment permit for metal finishing (40 CFR 433) within project timeline." A weak entry: "infrastructure."
A director reviewing fifteen entries tagged "infrastructure" over two years is reading a word. A director reviewing entries that name the POTW's pretreatment limitation, the substation's available capacity in MW, or the absence of a completed Phase II on the offered parcel is reading constraints that map to specific counterparties and specific budget requests.
Two offices recording the same elimination illustrate the difference. A plastics compounder requires 8MW of three-phase service at the site. The city's industrial park has 3MW available; the utility says a substation upgrade runs 14 months and $2.1 million. Office A writes: "lost on electric." Office B writes: "eliminated on electric capacity — project required 8MW three-phase at site, 3MW currently available, utility estimates 14-month upgrade timeline and $2.1M cost, project timeline was 9 months to occupancy." Office B has documented the size of the gap, the cost to close it, and the calendar mismatch that made the gap fatal on this particular project. When a 6MW requirement shows up eight months later, nobody has to re-investigate.
Qualification threshold. What would have produced a passing answer. If the project required 12MW of available power and the city has 4MW, record both numbers. If it required a Phase II completed within the last 18 months and your most recent Phase II is from 2019, record the requirement and the gap. This field turns the loss into a measurable distance between current condition and market demand.
Structural vs. evidentiary. Did the city lack the asset, or have the asset and lack the documentation to prove it? I worked through this distinction in a prior piece on evidence quality. A city with no spec building has a structural gap. A city with a spec building but no current appraisal, no Phase I, and no utility service letter has an evidentiary gap. The evidentiary gap is cheaper and faster to close. A record that doesn't separate the two will overstate the infrastructure investment needed.
A city eliminated because it lacks the asset faces a capital investment problem. A city eliminated because it has the asset but can't document it faces a proof problem. The second is cheaper and faster to close, but only if the record distinguishes the two.
Diagnosis source. Who identified the failed condition: the site selector or company, the state project manager, or your own staff. More on this below.
Confidence level. How certain you are that the recorded condition was actually decisive. This is a separate dimension from diagnosis source. A selector might confirm the elimination and still give feedback vague enough to be low-value — "it was a combination of factors, but your site wasn't quite right" — which is high source quality and low diagnostic confidence. Your staff might identify the failure point with no external confirmation at all, but if the project required Class I rail and your site has none, confidence is high regardless. Record it as high, moderate, or low. When you filter for patterns, a cluster of high-confidence entries on one constraint carries more weight than a cluster of moderates.
Date. When the elimination occurred or was recognized. Without dates, temporal patterns are invisible. The City of Chesapeake's 2021 performance audit found that records migrated from an earlier tracking system carried no dates at all, which made longitudinal analysis impossible.
Weighting entries by source quality
Inquiries don't carry equal signal.
Tier 1: State-screened or direct selector RFI. The project is real, the requirements are specific, and the elimination reflects an actual market test of your assets. Highest diagnostic value.
Tier 2: Utility-channel or regional partnership referral. The project is probably real but may have reached you through broad distribution, and the requirements may be looser. The elimination reason still counts, but it reflects a rougher screen.
Tier 3: Unsolicited broker inquiry or unverified contact. The project may or may not exist. Requirements are usually generic. A broker asking whether you have 500 acres with rail could be carrying a real prospect or fishing.
When you go looking for patterns, count Tier 1 first. Three Tier 1 losses on the same constraint inside 18 months is a signal. Three Tier 3 losses on the same constraint may be one project that reached you through three brokers.
Identifying the same project across channels
A battery materials company scouting the Southeast can generate a state agency RFI, a parallel inquiry through the regional electric utility, and a broker call to the local EDO inside the same month. Three entries, one project. Counted separately, they distort the pattern.
Match on sector, approximate scale, and timing. Two entries with the same sector, similar employment and investment ranges, arriving within 30 days of each other are probably the same project. The 30-day window is a working default drawn from how project timelines tend to generate parallel inquiries, not an empirical threshold; adjust it if your own arrival patterns say otherwise. Consolidate into one entry and preserve the channel information from each source. A project that found you through three channels tells you something about your market visibility even though it counts once.
Missouri Partnership separates projects it sourced or led (69% of 192 recorded losses from FY2016 through FY2023) from partner-led projects (31%), which helps identify duplicates at the state level. At the local level you're more likely to see the same project arrive through a state channel and a utility channel at the same time. Flag the overlap.
Who made the diagnosis
The quality of your diagnosis depends on where it came from.
Selector or company confirmed. Someone on the project side told you why you were cut. Missouri Partnership's loss analysis includes direct company and consultant comments held in its CRM, specific enough to identify site shape, utility upgrade timelines, labor-shed depth, and supply-chain proximity as elimination reasons. Record that feedback verbatim and tag it confirmed.
Staff-diagnosed. Your team read the RFI requirements against your assets and identified the likely failure point, but nobody from the project side confirmed it. Missouri acknowledges that some of its recorded reasons reflect "project managers' understanding of the outcome" rather than direct feedback. Useful, lower confidence. You may have correctly identified power capacity as the constraint, or you may have fixated on power while the real problem was floodplain encroachment on the offered parcel.
Unknown. The project went quiet. Nevada GOED reports that companies commonly stop communicating once they've chosen another location. You know you lost and not why. Record the project profile and source channel anyway; the entry still contributes to sector and scale analysis even when it can't contribute to constraint analysis.
Treat diagnosis source as a data-quality flag on every entry. Five confirmed eliminations on the same constraint is actionable. The same count of staff diagnoses is a hypothesis worth testing before it goes into a capital request. A run of unknowns tells you about pipeline volume and sector mix but nothing about constraints.
What the record produces
The value appears when a director sits down with the record quarterly or semi-annually and filters it.
Pull the record. Filter to Tier 1 and Tier 2 entries from the last 12 months. Group by failed condition. Look for the constraint that appears more than once. If three entries in the last year show elimination on electric capacity, check confidence level and diagnosis source on each. Two high-confidence entries with at least one selector-confirmed gives you a documented, recurring infrastructure constraint with demand thresholds attached — the MW requirements sitting in each entry's qualification threshold field. That goes to the utility and the city manager in the same conversation, and it sounds like this:
"We have been eliminated from four Tier 1 projects in the last two years on wastewater treatment capacity, representing a combined 800 jobs and $200 million in proposed capital investment, all requiring industrial pretreatment permits our current POTW cannot issue."
A city manager can evaluate that without knowing what an RFI is.
Prince George County, Virginia offers a precedent for the argument if not the method. County staff built the business case for a 3-MGD pump station and force main to the Board of Supervisors by tying it to industrial projects the county had lost on inadequate water and wastewater capacity. The FY2026 budget carries debt service supporting roughly $30 million in utility capital projects. Staff assembled that case from institutional memory rather than a maintained record. A record would have made the assembly unnecessary and the evidence more precise: project counts, demand profiles, timelines.
At state scale, Virginia's economic development partnership documented at least 65 project eliminations from FY2017 through FY2019 attributed to the absence of developable sites, representing nearly 19,000 potential jobs and more than $5 billion in proposed investment. That analysis fed a $234 million state appropriation for site development across FY2023 and FY2024.
Frederick County, Maryland has taken a middle step. Its 2026 economic opportunity plan tabulates missed projects by sector and assigns each a barrier: electric capacity, water, natural gas, site availability, rail, entitlements. The county attributes 4,466 proposed jobs and approximately $4.69 billion in proposed capital expenditure to those missed opportunities. No capital authorization has followed yet. The analytical foundation is in place.
What the record cannot do
It captures projects you saw. It cannot capture searches where your city was never considered — confidential desktop screens that eliminated you before anyone in your office knew a project existed. What you're measuring is encountered demand, not total demand.
It also doesn't establish that closing a constraint produces future wins. A city that puts $30 million into wastewater capacity after losing four projects on that dimension has addressed a necessary condition. Whether those four would have selected the city absent the constraint, and whether comparable projects will surface again, are questions the record can't answer.
Getting started
A spreadsheet with the eight fields above, maintained consistently, will surface patterns in 12 to 18 months. Entry quality is what makes it work; the platform is incidental.
The non-pursuit record from Issue #4 covers the decision not to respond to a specific RFI and the constraint behind it. A deliberate no-bid is a self-identified disqualification and belongs in the same dataset with the same fields.
Every failed condition points back to an institutional counterparty. Wastewater capacity: the POTW. Power availability: the utility. Environmental status: whoever controls the Phase II. The counterparty map from Issue #2 covers who to approach. The disqualification record tells you which gap to bring to them first.
- Missouri's loss taxonomy: Missouri Partnership's public analysis of 192 lost projects from FY2016 through FY2023 is the strongest publicly accessible model for how a statewide attraction organization records and categorizes elimination reasons, including direct company feedback stored in its CRM.
- North Carolina's loss categories: EDPNC's presentation reporting 60 new-location projects lost to other states in 2024 breaks reasons into building/site availability, search geography, supply-chain proximity, and infrastructure, showing how a state aggregates local losses into statewide constraint categories.
- Frederick County's barrier table: Frederick County, Maryland's 2026 economic opportunity plan tabulates missed projects by sector and assigns specific infrastructure barriers — electric, water, natural gas, rail, site, entitlements — providing a local template for converting anonymized losses into a public planning document before capital authorization.
- Virginia's loss-to-appropriation sequence: VEDP's documentation of at least 65 site-related project eliminations from FY2017 through FY2019 contributed to a subsequent $234 million state appropriation for site development, the clearest public example of a loss record driving a capital response at scale.