PART 01

A Freddie Mac veteran in his first sales role, hunting $50M of multifamily originations.

Drew spent 8 years at Freddie Mac before joining KeyBank. He knows agency execution cold — Fannie DUS, Freddie Optigo, the whole stack — but he's never been the one on the phone closing deals. He has the technical credibility. What he doesn't have is a repeatable outbound engine, and he's the first to admit it.

Honest take: Drew is coachable, has real domain expertise, and has explicit budget authority on data tools ("cost is not an issue"). What's missing is a system. That's a much easier problem than missing knowledge or missing budget.

The numbers

Drew at a glance

$50M

Annual origination target

$5–10M

Typical loan size

~5–10

Deals needed to hit goal

8 yrs

At Freddie Mac before KeyBank

8.4/10

Discovery-call coaching score

First

Time in a sales seat

PART 02

What KeyBank actually sells (and where Drew's edge is).

Drew originates permanent multifamily mortgages — but the real differentiator isn't the agency loan itself. It's the balance-sheet bridge → agency takeout combo: a 2–3 year value-add bridge loan that converts into a permanent Fannie DUS or Freddie Optigo mortgage. Vanilla agency lenders can't do that. Vanilla balance-sheet lenders don't have the agency expertise. KeyBank does both, and Drew is fluent in both.

The product

Three things to lead with on every outreach

EDGE 01 Bridge-to-agency in one shop

A sponsor buying an apartment building to value-add doesn't have to stitch together two lenders. KeyBank does the bridge today, then refis it onto its own agency execution at stabilization — same paperwork team, same underwriter, same closing certainty. This is the wedge against pure agency shops.

EDGE 02 Agency execution credibility

8 years at Freddie Mac means Drew knows what underwriting will and won't accept before submission. For sponsors, that translates into certainty-to-close — fewer late-stage surprises, faster timelines, more aggressive structures cleared the first time. This is the wedge against pure balance-sheet lenders.

EDGE 03 KeyBank scale + servicing quality

Drew is not a one-person shop. KeyBank's commercial real-estate book gives sponsors a long-term lender-of-record relationship — refi, expand portfolio, syndicate, all under one umbrella. Important for repeat-deal sponsors, less relevant for one-and-dones.

Who buys this

The four trigger events that create a real conversation

Drew's ICP is anyone who owns or is buying an apartment building. That's too broad to outbound — millions of LLCs match. The targeting that makes outbound work is filtering by which trigger event is happening right now:

Trigger	Why it's high-intent	Window	Priority
Loan maturity in next 6–18 mo	Sponsor MUST refi or sell. Forced-decision moment.	6–18 mo out	P0 — easiest wedge
Bridge loan from 1–2 yrs ago	Agency-takeout window opens. KeyBank's combo is the natural fit.	12–24 mo after origination	P0 — best fit for Drew
Value-add cycle complete	NOI stabilized → ready to swap construction debt for permanent agency.	Renovation visible in permits / listings	P1
New acquisition	Need acquisition financing now.	Immediately, but contested	P2 — most contested

The conclusion: we don't need to find every apartment owner. We need to find the ones with a maturity date in the next 6–18 months, because that's where Drew's KeyBank pitch lands hardest and the timing is forced.

PART 03

What's actually broken — and what isn't.

Drew's product is fine. His credibility is fine. His budget is there. The breakage is upstream of all of that: he has no repeatable way to find sponsors with a forced refi event coming up, no funnel metrics to know what's working, and no daily activity discipline. Fix those three things and the rest is execution.

From the discovery call

The five gaps he named, ranked by leverage

GAP 01 No structured outbound plan or daily activity targets

Today Drew reaches out opportunistically — when something crosses his desk. There's no list of "20 calls / 30 emails / 5 LinkedIn DMs by Friday." That's the difference between random originations and a $50M run rate. Highest leverage to fix.

High impact · Low effort

GAP 02 Manual lead research and trigger-event identification

Drew can identify a refi opportunity once he sees it — but seeing it is the bottleneck. There's no system that walks the maturity wall on the agency books and surfaces "12 sponsors in your geo with loans coming due in Q3 2026." That's exactly what Phase 02 below builds.

High impact · Medium effort

GAP 03 No funnel baseline metrics

Drew doesn't know his current reply rate, meeting rate, close rate, or cycle length. Without those, you can't tell if a new system is working — you can only tell if revenue went up, which has a 6–12 month lag. We capture these from week 1.

Medium impact · Low effort

GAP 04 ICP not yet quantified

Drew said "anyone with an apartment building." That's a category, not an ICP. We need: asset-size band ($X–Y NOI or unit count), geographies he covers, sponsor-type rules, exclusions. 20 minutes on the phone with him fixes this.

Medium impact · Low effort

GAP 05 Some data tools blocked by KeyBank IT

Drew flagged that subscription access from non-KeyBank environments may be restricted by corporate compliance. We sidestep this entirely by running the data pipeline on Cynthia's infrastructure and only handing Drew the ranked output — no tools for him to install, no IT review needed.

Low impact · Already handled

What's NOT broken

Things we don't need to fix

✓Domain knowledge. 8 years at Freddie Mac. Drew knows the product better than the prospects do.
✓Differentiator. The bridge-to-agency combo is real and underused as a pitch — that's content to lead outreach with, not a problem to solve.
✓Budget. "Cost is not an issue" was said on the call. Tooling spend is not the gating factor.
✓Coachability. Drew was candid about gaps and committed concrete next steps. That's the rarer ingredient.

The unlock

The data falls into three layers.

Once we map it this way, it's clear what's public, what needs an enrichment pass, and where the leverage is.

Layer	What it gives you	Where it comes from
Maturity wall	Every loan, every date, every UPB — but anonymized at borrower level	Public agency disclosures
Owner identity	The LLC or sponsor behind the address	County recorder data + property APIs
Decision-maker	Email, phone, LinkedIn for the principal	B2B contact + LinkedIn enrichment

The agency disclosures (Freddie, Fannie, Ginnie) tell us where the maturity wall is — every loan, every date, every UPB. They don't tell us who owns the property — that's a separate enrichment step. Once we have the LLC, getting to the human inside it is one more pass through B2B contact data and LinkedIn. We handle all three layers end-to-end.

Layer 1 · The maturity wall

Where Drew's leads literally live in the open.

These are public, government-published datasets, updated on a regular cadence. Combined, they cover the vast majority of US multifamily debt — every loan, every maturity date, every unpaid balance.

Public disclosure · Has named borrowers

HUD FHA Multifamily / 223(f) data

The single best public source on this list. Every FHA-insured multifamily loan, with project name, mortgagor name, full property address, original mortgage amount, maturity date, interest rate, and program section. Every row is directly actionable — no enrichment needed to know who to call. Narrower slice than the agencies (FHA-insured only), but every one of those borrowers is exactly the kind of refi target Drew wants to flip onto KeyBank's bridge-to-agency combo at maturity.

Source: hud.gov/datasets Updated: Quarterly Names included: Yes

Public disclosure · Anonymized at borrower level

Freddie Mac Multifamily Loan-Level Disclosure

Optigo SBL + K-Deal + Conventional loan-level files, downloadable as CSV/Excel. Includes original UPB, current UPB, maturity date, origination date, LTV, DSCR, property city/state/ZIP, property type, unit count, prepayment provisions. Does not include borrower name or street address — only ZIP-level location. Used for sizing the maturity wall by geography; we attach the names in Layer 2.

Source: mf.freddiemac.com/investors/data Updated: Monthly (~30-day lag) Names included: No

Public disclosure · Anonymized at borrower level

Fannie Mae DUS Disclose

Same shape as Freddie's. Search by pool/CUSIP and pull loan-level: maturity date, original UPB, LTV, DSCR, MSA, sometimes property name, unit count, year built. Property name appears on many records but borrower entity does not in the standard public file.

Source: mfdusdisclose.fanniemae.com Updated: Monthly Names included: No

Public disclosure · Has named sponsors (large loans only)

SEC EDGAR — CMBS Annex A

Look at 424B5 prospectus supplements, 10-D monthly servicer reports, and ABS-EE filings for new CMBS deals. The Annex A of every prospectus supplement lists the top loans with sponsor name, borrower entity, property address, maturity date, UPB, LTV, DSCR. Useful for loans >$25M (top-20-loan disclosure threshold) — institutional/REIT sponsors at the upper end of Drew's range.

Source: sec.gov/edgar Updated: At deal issuance + monthly servicer reports Names included: Yes (large loans)

Public · Skip

Ginnie Mae Multifamily MBS · NMHC Top 50

Ginnie's multifamily disclosure files are duplicates of HUD's data without the names — just go to HUD directly. NMHC's "Top 50 Apartment Owners" list is publicly available but includes no portfolio detail and is what every other competitor already uses. Table stakes, not edge.

Layer 2 · Owner identity

Turning an anonymized address into a real LLC.

For Freddie/Fannie/Ginnie loans (the bulk of the maturity wall), we have a property address but not a borrower name. Cynthia runs every property address through a county-recorder data layer that returns the owner LLC, last mortgage filing, and recording history. The result joins back to the loan record so every entry on the maturity wall has a real entity behind it.

Cynthia handles this

County recorder data via property API

Owner LLC, last mortgage filing, deed history — covered nationwide via property data APIs that aggregate the ~3,000 US county recorder offices into one normalized feed. Caching is aggressive (owners don't change often), so the same property is only resolved once and reused across runs.

Coverage: Nationwide Latency: Real-time Names included: Yes

Layer 3 · Decision-maker

Going from LLC to a person we can actually email.

Once we have an LLC name, we still need the human inside it — the principal, managing member, GP, or asset manager. Cynthia handles this with a chain of B2B contact and LinkedIn enrichment so every record arrives with name, role, email, phone, and LinkedIn URL ready to outreach.

Capability	What it does
B2B contact resolution	Pulls emails, phones, titles for the people inside the LLC
LinkedIn enrichment	Finds the principal/managing member by company, returns profile URL + role
Sponsor-page scraping	Fallback when an LLC has its own website but isn't in B2B databases
Disclosure parsing	Extracts structured records from agency PDFs and SEC filings
LinkedIn outreach layer	Sends connection requests + DMs through approved accounts
Cold-email infrastructure	Sequences, sending domains, deliverability, reply routing
Address normalization + geocoding	Cleans up disclosure addresses to canonical form for matching
CRM destination	Qualified leads land in Drew's pipeline with the trigger event attached

Every layer of the pipeline runs on Cynthia's infrastructure. Drew sees ranked, named, contactable leads with a trigger event — none of the data plumbing.

End to end

The full pipeline.

Six stages, executed monthly. Stage 1 ingests the raw maturity wall, stage 2 attaches names where the agency files don't, stage 3 ranks by Drew's ICP, stage 4 attaches humans, stage 5 sends the outreach, stage 6 feeds outcomes back into the scorer to make the next month sharper.

                         ┌─────────────────────────┐
                         │  STAGE 1 — INGEST       │
                         └─────────────────────────┘
    Freddie loan-level ──┐
    Fannie DUS Disclose ─┤── Firecrawl + scheduled cron (monthly)
    HUD Multifamily ─────┤── normalize to one schema:
    SEC EDGAR Annex A ───┘     borrower_name (nullable),
                               property_address, msa, units,
                               original_upb, current_upb,
                               origination_date, maturity_date,
                               loan_type, lender, source

                         ┌─────────────────────────┐
                         │  STAGE 2 — RESOLVE NAME │
                         └─────────────────────────┘
    For rows where borrower_name IS NULL:
      Property address ─→ county recorder API ─→ owner_llc
      (Cache aggressively — owners don't change often)

                         ┌─────────────────────────┐
                         │  STAGE 3 — SCORE        │
                         └─────────────────────────┘
    Drew's ICP rules:
      • maturity_date in next 6–18 mo  (highest weight)
      • original_upb in $5M–$50M band   (KeyBank sweet spot)
      • geography ∈ Drew's coverage     (TBD — needs 20-min ICP call)
      • bridge-loan vintage 1–2 yrs ago (agency-takeout window)
      • exclude: KeyBank's own book, recent refis, banned sponsors
    Score = weighted sum → rank.

                         ┌─────────────────────────┐
                         │  STAGE 4 — ENRICH PEOPLE│
                         └─────────────────────────┘
    For each top-N owner_llc:
      Apollo first (fastest) → managing member / principal
        ↳ fall back to PDL if Apollo misses
        ↳ fall back to LinkedIn search via Gojiberry
      Pull: name, role, email, phone, LinkedIn URL

                         ┌─────────────────────────┐
                         │  STAGE 5 — OUTREACH     │
                         └─────────────────────────┘
    AgentMail (cold email) + Gojiberry (LinkedIn) sequence
      • Opener anchored to ONE signal:
        "Your $14M loan on 2400 Maple matures Sept 2027 —
         wanted to put KeyBank's balance-sheet-to-agency
         option in front of you 12mo out…"
      • One CTA per touch: 15-min call

                         ┌─────────────────────────┐
                         │  STAGE 6 — FEEDBACK     │
                         └─────────────────────────┘
    Log every reply / meeting / closed-won back into the scorer.
    Drew's 5–10 win examples = the initial positive training set.

Honest tradeoffs

What we're NOT going to do (and why).

CHOICE 01 v1 starts with HUD only — fastest path to a real list

HUD multifamily already includes borrower names, so the first 100–200 prospects can be ranked and delivered without waiting on the address-to-LLC layer for Freddie/Fannie. Smaller pool, but every record is directly actionable. Once Drew validates the v1 ranking, we layer in the full Freddie/Fannie maturity wall on top — same scoring model, much bigger pool. This sequencing makes the 48-hour deliverable real instead of aspirational.

Already the plan

CHOICE 02 No legacy CRE platforms (Trepp / RCA / CoStar)

These platforms mostly repackage data that's available through the public agency disclosures plus county recorder feeds. Cynthia goes directly to the source, which means fresher data, faster turnaround, and no dependency on an enterprise vendor that's tuned for analysts, not originators.

CHOICE 03 Drew never touches a tool — only output

All ingestion, name resolution, scoring, and enrichment runs on Cynthia's infrastructure. Drew gets a ranked CSV (and later a CRM feed) with names, contact info, and the trigger event attached. No subscriptions to manage, no IT review, no compliance friction with KeyBank's environment.

The math

What a working pipeline looks like at $50M.

Drew's goal is $50M of originations. Working backwards from typical agency-multifamily conversion rates and the 6–10 deals it takes to hit that:

$50M

Origination target

6–10

Deals required

~120

Real conversations needed

~600

Outreach replies needed

~6,000

Targeted touches needed

~200/mo

High-fit prospects in pipeline

200 high-fit prospects per month, hit on email + LinkedIn, gets us comfortably to the conversation volume that produces 6–10 closed originations annually. The pipeline above is what produces those 200 per month, end to end. Nothing exotic.

Next step

Three things, in order.

01 20-minute ICP call with Drew. Lock the asset-size band, geographies he covers, sponsor exclusions, and his 5–10 win examples. Without these, scoring is noisy. Schedule before any data work.

02 v1 lead pool from HUD only — 48 hours. 100–200 named multifamily sponsors with maturity dates inside Drew's strike zone. Delivered as a CSV. Drew validates the ranking before any outreach goes out.

03 Layer in the full maturity wall + outreach the same week. Once Drew confirms the v1 ranking is good, we extend the pipeline to the full Freddie + Fannie pool (roughly 10× the volume) and start multi-channel sequences against the top 200/mo. Funnel metrics captured from day 1.

Drew already has the credibility, the product, and the budget. The only missing piece is a system that finds the right humans with the right timing. Phase 02 is that system. The 48-hour HUD pull is how we prove the targeting works before scaling to the full maturity wall.

APPENDIX

Receipts & methodology

Source-of-truth for everything in this report — call commitments, transcript notes, and the data-source research that drove the pipeline design.

Open commitments from the April 22 call

Cynthia owes Drew

→48-hour first-lead-pool prototype (committed April 22 — currently overdue, hence Phase 03)
→Audit of available data sources for loan-maturity + multifamily acquisition signals (this document)
→Follow-up email with any data/access questions Drew needs to answer

Drew owes Cynthia

→5–10 example clients with brief bullets on why they chose KeyBank (the positive training set for the scorer)
→Curated list of unreached prospects from his current research, where compliance allows
→Coaching session with Chad on call framework (Thu/Fri this week)

Discovery call · highlights from the transcript

Re-transcribed April 27 (original transcription failed because DEEPGRAM_API_KEY was not configured at the time of the call). 29 min, 185 utterances, 3 speakers (Ricki, Drew, Chad).

Pain points Drew named directly

!"No structured outbound business plan or consistent activity targets" — first sales role, reaches out opportunistically
!"Manual lead research and trigger-event identification" — loan maturity opportunity mapping is possible but labor intensive
!"Data/tool access friction due to corporate environment constraints" — KeyBank IT may restrict subscription access
!"Differentiation difficulty in a commoditized lending market" — needs to lead with execution speed/certainty, not rate
!"Inconsistent system support from internal management/process" — perceives little structured enablement

What Drew said matters

→Consistent pipeline of high-intent refinance/acquisition opportunities
→Higher conversion through hyper-personalized outreach
→Faster trust-building around certainty, execution speed, and aggressive terms
→Building long-term lender-of-record relationships
→Reducing wasted outreach on low-fit prospects
→Establishing repeatable daily workflow and feedback-driven optimization
→Hitting and scaling beyond $50M origination target

Coaching score

8.4 / 10 — strong discovery + alignment call with clear momentum and ownership. Talk ratio: Ricki 60% / Drew 35% / Chad 6% (host-heavy but appropriate for a coaching/strategy session — future calls should give Drew more role-play time).

Data source research — full table

Source	Has names?	Freshness	Role in the pipeline
HUD FHA Multifamily	Yes	Quarterly	v1 anchor — every row directly actionable
Freddie Mac Multifamily	No	Monthly	Maturity wall heatmap; names resolved in Layer 2
Fannie Mae DUS Disclose	No	Monthly	Same shape as Freddie
Ginnie Mae Multifamily	No	Monthly	Skip — duplicate of HUD without names
SEC EDGAR Annex A	Yes	At issuance	Big loans only ($25M+) — top end of Drew's book
County recorder feed	Yes	~Real-time	Layer 2 — turns addresses into LLCs
NMHC Top 50	Yes	Annual	Public but everyone uses it — no edge
Trepp / RCA / CoStar	Yes	Real-time	Repackage of the same upstream data — not used

How this was produced

The Drew Musser discovery call recording (April 22, 2026, session ID 3e00b73a-05f1-45cb-9184-c96c6c682011, 29 min) was re-transcribed via Deepgram Nova-3 with diarization on April 27. The structured analysis (call type, coaching score, action items, qualifying info, pain points) was produced by the meet-analyst agent at analysis.cynthiaconcierge.com's upstream pipeline. Public data source research was conducted via web research against agency disclosure documentation; nothing in this report relies on access to KeyBank-internal systems or Drew's existing tools.

Drew Musser

A Freddie Mac veteran in his first sales role, hunting $50M of multifamily originations.

Drew at a glance

What KeyBank actually sells (and where Drew's edge is).

Three things to lead with on every outreach

The four trigger events that create a real conversation

What's actually broken — and what isn't.

The five gaps he named, ranked by leverage

Things we don't need to fix

The most high-intent leads sit on public agency loan books — but the names are anonymized. Here's how we get to them.

The data falls into three layers.

Where Drew's leads literally live in the open.

Turning an anonymized address into a real LLC.

Going from LLC to a person we can actually email.

The full pipeline.

What we're NOT going to do (and why).

First lead pool: 100–200 named multifamily sponsors with maturity dates inside Drew's strike zone.

What a working pipeline looks like at $50M.

Three things, in order.

Receipts & methodology

Cynthia owes Drew

Drew owes Cynthia

Pain points Drew named directly

What Drew said matters

Coaching score