Most firms have a buy box. Most buy boxes are a paragraph in a deck, a slide in an investor presentation, or a one-page memo that gets forwarded to brokers when the firm wants to signal what it is looking for. These artifacts describe the buy box. They do not enforce it. The buy box that lives in a memo cannot screen a deal. It only documents what the principals say they want.
The buy box that screens deals is a different artifact. It is a structured set of criteria, machine-readable, that can be applied to an extracted OM in milliseconds and produce a fit score with line-item rationale. It does not replace judgment. It replaces the part of judgment that is mechanical pattern-matching, freeing the firm to apply real judgment to deals that pass the filter.
The conversion from memo buy box to queryable buy box is not a technology project. It is a thinking project. The memo is vague because the firm has not forced itself to be specific. Specificity is the work.
What a Memo Buy Box Hides
A typical memo buy box reads something like: "Multifamily, value-add, sunbelt MSAs, 100 to 300 units, $20 million to $60 million, vintage 1980s and newer, IRR target 16% plus."
This sentence appears specific. It is not. It hides at least a dozen criteria that are not stated and that vary by deal.
Stated Criterion | Hidden Question |
|---|---|
Multifamily | Garden, mid-rise, high-rise, age-restricted, student? |
Value-add | What kind: capex, operations, lease-up, repositioning? |
Sunbelt MSAs | Tier-one only, or tier-two acceptable? Which states? |
100 to 300 units | Strict, or willing to flex 80 to 350? |
$20M to $60M | Total cost or equity? Including capex? |
Vintage 1980s+ | Hard cut, or willing to look at older with capital plan? |
IRR target 16%+ | Levered or unlevered? Hold period? Exit cap assumption? |
When a broker forwards a deal that hits the stated criteria, the firm still says no. The reason is one of the hidden questions, which the broker never knew about. The firm calls this discipline. The broker calls it noise. Both are right.
What a Queryable Buy Box Contains
A buy box that can be applied to a deal automatically has to encode every criterion the firm actually uses, not just the criteria it lists in the memo. The structure breaks into hard filters, soft filters, and weights.
Filter Type | Behavior | Example |
|---|---|---|
Hard filter | Pass or fail; no override | Asset type must be multifamily |
Soft filter | Reduces score but does not exclude | Vintage older than 1985 |
Weight | Determines composite score | Geography fit weighted 25% |
Hard filters answer the question: should the firm even look at this? Soft filters answer: how much does this deal deviate from the ideal? Weights answer: which deviations matter most for this firm?
The components a serious buy box has to encode include the following.
Dimension | Common Variables |
|---|---|
Geography | MSAs, submarkets, exclusion zones, climate risk |
Asset class | Property type, subtype, age, condition |
Size | Units, square footage, deal size, equity check |
Financial | Cap rate, IRR, equity multiple, DSCR, hold period |
Tenancy | Credit, mix, concentration, lease structure |
Business plan | Strategy type, capex range, lease-up risk, time horizon |
Sponsor | Operator type, track record, vertical integration |
Capital | Debt assumption, equity source, JV structure |
Process | Marketed vs. off-market, broker, timeline |
A complete buy box has thirty to fifty fields. The firm does not have to fill all of them. It has to know which ones it has filled and what default applies when it has not.
How Hard Filters and Weights Interact
The buy box has to handle two failure modes. The first is letting through deals that should be auto-declined. The second is auto-declining deals that should have been reviewed.
Hard filters are the protection against the first failure mode. They eliminate categories the firm will not consider regardless of how attractive the rest of the deal looks. A debt fund's buy box has hard filters on lien position. An equity fund's buy box has hard filters on asset type.
Weights are the protection against the second failure mode. They allow a deal that misses on one dimension to compensate by exceeding on another. A deal slightly above the size band but with strong demographics and a credit tenant should not auto-decline. A deal slightly outside the geography but matching every other criterion should surface for review.
The composite score is not a decision. It is a sort key. The principal still decides. The score determines what the principal sees first.
The Exception Pattern
Every firm has a buy box and every firm makes exceptions. The exceptions are where most learning happens. A firm that closes a deal outside its stated buy box has either an opportunistic relationship win, a buy box that needs to expand, or a discipline failure.
A queryable buy box makes exceptions visible. The deal that closed at 78% fit is logged with the components that scored low. The pattern of exceptions is reviewable. After a year, the firm knows which dimensions it bends on, which it does not, and where the stated buy box has drifted from the actual buy box.
Closed Deal Score | Likely Cause |
|---|---|
90% plus | Standard execution within stated criteria |
70 to 89% | Strategic flex on one or two dimensions |
Below 70% | Off-strategy; either opportunity win or discipline failure |
The firm that reviews this pattern quarterly converts the exception data into buy box updates. The firm that does not review it accumulates contradictions between what it says it does and what it actually does, which costs it both broker credibility and internal alignment.
What Brokers See
A buy box that lives in a memo produces a constant stream of mismatched submissions. A buy box that lives in a system can be partially exposed to brokers without revealing the weights. A broker can see the hard filters: asset type, geography, size band, return target. The broker self-screens before submitting, which raises the quality of the funnel and reduces the noise on both sides.
Firms that are protective about their buy box often hide more than they need to. The hard filters are usually inferable from the firm's closed deals. Publishing them costs nothing and improves submissions.
What "Done" Looks Like
A buy box that can screen deals automatically meets the following criteria:
Every dimension the firm actually uses is encoded, not just the dimensions in the memo.
Hard filters are explicit and produce auto-decline with logged rationale.
Soft filters reduce score with documented weights.
Composite scoring is reviewable and tunable based on closed-deal patterns.
Every screening decision references the buy box version that produced it.
Exceptions are tracked and reviewed on a fixed cadence.
If a deal can pass the buy box and the firm still says no for a reason the buy box did not capture, the buy box has a gap. The gap is the work.
Conclusion
A buy box is a tool, not a document. The firms that treat it as a document write it once and forget it. The firms that treat it as a tool maintain it, version it, and use it to screen every deal that comes in. The maintenance is what produces the discipline, and the discipline is what produces consistent decisions across hundreds of deals over multiple market cycles. A queryable buy box does not replace investment judgment. It removes the cases where judgment was just pattern-matching, leaving more time for the cases where judgment is actually required.