The math of ETA (Entrepreneurship Through Acquisition) is brutal. On average, searchers look at 300–500 companies before acquiring one. That's not 300 deep dives—most are killed in under 5 minutes. The buyers who close aren't superhuman analysts. They've built a system.
Why Most Deals Die on First Contact
Roughly 70% of deal flow from brokers and marketplaces is eliminated before a second email. The reasons are consistent:
- Wrong sector fit — the business is outside the buyer's defined criteria
- SDE math doesn't work — asking price implies a multiple above 4.5x for a mediocre business
- Red-flag keywords — "owner must stay 2 years," "proprietary software," "single customer = 80% revenue"
- Geography mismatch — remote-capable requirement not met
None of these require a CIM. They can be caught in a 60-second scan. The problem is that most buyers aren't doing a 60-second scan—they're either ignoring deals or spending 30 minutes on deals that should have been killed in 60 seconds.
The Three-Layer Screening Stack
Layer 1: Listing Filter (60 seconds)
Before opening a teaser, apply mechanical filters:
- Revenue ≥ $500K (enough scale to pay a fair salary + debt service)
- SDE ≥ $150K (otherwise the acquisition premium doesn't justify the risk)
- Asking multiple ≤ 4.0x SDE for service businesses, ≤ 3.5x for retail/seasonal
- Business age ≥ 5 years (survival bias is real)
Most marketplaces let you filter by revenue and asking price, but not by implied multiple. You have to calculate it. AI deal screening tools can do this automatically from listing text.
Layer 2: Teaser/CIM Scan (5–10 minutes)
Once a listing passes the number filters, pull the teaser and scan for:
- Revenue concentration: Does the top customer account for >25% of revenue? That's a deal-breaker for most bank-financed acquisitions.
- Owner dependency: Is the owner the business? CIMs often bury this. Look for "owner manages all client relationships" or "owner is the primary technician."
- Recurring vs. transactional: A service business with 80% recurring revenue is worth 30–40% more than a comparable transactional business. Price should reflect this.
- Add-backs credibility: Excessive owner add-backs (personal travel, family payroll, depreciation on owner-used vehicles) inflate reported SDE. Model a conservative recast.
Layer 3: The Deal Memo
If a deal survives Layers 1 and 2, it earns a proper memo. This is where most buyers lose time. A deal memo typically takes 2–4 hours to write well. It synthesizes the CIM, broker conversations, and preliminary due diligence into a one-page investment thesis that you can share with lenders, partners, and advisors.
The memo should answer five questions:
- What does this business actually do, and who are its customers?
- Why does it earn the money it earns? (competitive moat, if any)
- What are the 3 biggest risks to the thesis?
- What's the debt-service coverage at the proposed purchase price?
- What's the first 90-day operating plan?
Where AI Fits in Deal Screening
AI doesn't replace judgment—it eliminates drudgery. The highest-leverage applications in ETA deal screening are:
- Automated CIM extraction: Pull key metrics (revenue, SDE, customer concentration, employee count) from PDF CIMs without reading them manually
- Deal memo first drafts: A well-prompted AI can produce an 80% complete deal memo from a CIM in 60 seconds, leaving you to focus on the 20% that requires judgment
- Thesis scoring: Match deals against your documented acquisition criteria automatically—surfacing the best fits, not just the most recent listings
- Comparable analysis: Cross-reference multiples against historical sales data from Bizbuysell and similar sources
The ETA buyers generating 10–20 quality memos a month aren't working 10x harder. They've built a screening pipeline that handles the commodity work automatically.
Building Your Personal Screening System
The framework above scales to any deal volume. The key is consistency: define your filters before you see a deal, not while evaluating it. Bias toward deals that fit your criteria on paper, even if they feel boring. The exciting deals are usually expensive.
DealPacket is built around this workflow—paste a listing or upload a CIM, get a structured deal memo in under a minute, scored against your thesis. It won't close deals for you. It will stop you from wasting hours on deals that were never right.