Every record rides the same belt. One person keeps it moving.
Pull, skip trace, route, score, assign. Five handoffs. The data manager owns four of them.
Five stations, one belt. The data manager runs stations 1-4. Callers take the hand-off at station 5.
Marketing starts
SiftMap precision pulls need no build-out. Pull today, dial today.
Numbers cut
Each Trestle pass drops half the dials, keeps about 80% of correct numbers.
Attempts in 72 hours
The initial dial plus 3 follow-ups on every first-to-market record.
Records per month
The floor a single precision strategy needs for consistent deal flow.
Elite lists, doors per deal (lower is better)
The rule that makes it work: escalate from the cheapest touch to the most expensive. Cold call is the first live touch, never SMS. Track every action at the property level. Revenue from each cheap touch funds the next one.
Stage 1: Pull SiftMap Precision, Market the Same Day
Three precision pulls run in parallel, not stacked on top of each other. The preset does the daily work.
AI / Investor Score
The most efficient lever: 22.8 to 30.6 doors per deal in the Knox+Blount backtest. Score 70 runs 37.9 and score 50 runs 66.9, so drop down only when you want coverage.
Acute distressor lists
Each Elite list runs as its own parallel pull: judgments, pre-foreclosure final judgement, negative equity, estate sales, pre-probate. Doors per deal from real sold data.
Native obituary
Auto-adds weekly, nationwide, nearly uncontested. A living flag: if probate files later, the flag flips and a last-obituary-date field keeps the record findable. Pair with deep prospecting on the heir.
Do not AND-stack Standard Distress on top of the AI score. Stacking makes efficiency worse, not better: Distress AND Investor 80+ runs 48.4 doors per deal versus 30.6 for the score alone. Run the score and each distressor list as separate pulls, then work them in parallel.
Build the filter preset
County plus zip plus list type plus equity plus asset class. One preset per pull, in SiftMap.
Toggle Auto Add New Records
The preset becomes a daily job. Every newly qualifying record lands on the records page, already skip traced on the Expert plan.
Filter to new records only
Use the records in my account / exclude my account filters so only NEW records hitting a list surface. You reach the owner before the list gets marketed.
Tag the preset
The tag set here routes the record in Stage 3. Precision tag means niche pipeline, no tag means bulk.
The exclude-my-account filter is the lowest-saturation position you can hold. The AI score flags a deal before it becomes an obvious marketed list, so you are first even on predicted data.
Access: native distressors and obituary ride SiftMap Pro, included in the $499/mo Expert plan with unlimited skip trace, or a $297/mo add-on on the $149/mo Professional plan. The Investor AI Score runs on the $1,250/mo AI plan.
Broad Score 50+ and Driving for Dollars are coverage: 70 to 85% of the market at 100 to 180 blended doors per deal. Reach, not the lead.
Stages 2-4: Skip, Route, Score
The middle of the belt is pure data manager work. Clean it, trace it, tag it, score it.
Standardize first. Every raw pull becomes a DataSift-compatible CSV:
- Clean address fields and proper headers
- Residential only, duplicates removed
- Messy PDFs or photographed screens: have Claude extract the structured data, never type records by hand
Skip tracing is a guess at which number belongs to a person. It aggregates roughly 6,000 sources, returns around 15 candidate numbers, and surfaces the top 5 by cross-source frequency.
Order of passes: DataSift built-in first ($0.10 to $0.15 per record), Skip Genie second, Forewarn if you hold a license. The Expert plan auto skip-traces records as they come in.
The tag on the Stage 1 preset decides the pipeline. Not manual uploads. On Expert there is no reason to add lists by hand anymore.
- Precision tag (niche list, acute distressor, obituary, courthouse data): niche sequential, click-to-dial every number, lists under about 1,000 records
- No tag or a bulk tag: bulk sequential, multi-line power dial, lists of 1,000 plus
- Current direction: pull top-neighborhood pockets and run niche click-to-dial, treating bulk-quality data like first-to-market
Build the suppression preset: recently sold owners drop out of marketing. Suppress, never delete. Go back about 3 years to cover county-to-Sift data delays.
Trestle scores every number 0 to 100 on activity, not correctness, reading carrier data on texts sent and numbers dialed. About 1.5 cents per number through the Claude phone-validator skill.
Skip tracing guesses which number belongs to the owner. Trestle confirms which numbers are active. Different jobs.
- Dial the dial-first (81-100), dial-second (61-80), and dial-third tags
- Skip dial-four and dial-five entirely
- Each pass removes about 50% of numbers while keeping roughly 80% of the correct ones
- Connect rate lifts from 2-3% to 10-12%: about 1 in 10 dials becomes a real conversation
The Trestle pass is what lets you treat bulk data like first-to-market quality. Half the dials, 80% of the correct numbers. That arbitrage wins every time.
Stage 5: Run the Sequential Flow
Cheapest touch to most expensive, tracked at the property level. One override: cold call is always the first live touch.
| Touch, cheapest first | Cost per touch | Rule |
|---|---|---|
| Text | $0.01 | Never the first touch. Follows the call and voicemail. |
| Cold call | $0.03-0.06 | Always the first live touch. Click-to-dial in niche, power dial in bulk. |
| Direct mail | $0.50-0.75 | Never send the same mailer twice in a row. |
| Deep prospecting | $1.50-4.00 | Reserved for records that survived every cheaper touch. |
| Door knocking | $25-50 | The premium touch for saturated, bottom-of-funnel lists. |
Bulk mode (large lists, broad coverage, like a Score 50+ or FTM-at-scale pull): cold call, then direct mail, then deep prospecting as the channel for the unreachable tail.
Precision mode (the Stage 1 lists): aggressive multi-touch. Call, then voicemail, then text, then mail, then email.
Sensitive-list carve-out, non-negotiable. Deceased and obituary lists drop cold text entirely: call up to 4 attempts, voicemail, mail, then deep-prospect the heir or executor, with a monthly rehash after. Never mention the death. Open with "do you have plans with the property," and address mail to the family of the deceased, not the decedent by name.
A full attempt is a manual increment: dial every number on the record, send a text, leave a voicemail. Dialer integrations cannot increment the counter. The team tells Sift the attempt is complete.
Standard first-to-market cadence: the initial dial plus 3 follow-ups. That is 4 attempts inside 72 hours.
Dial back to front. Start at follow-up 3, then follow-up 2, then follow-up 1, then ready-to-call. Start at ready-to-call and you keep pushing people into the step you are about to dial, so you double-dial them the same day.
The three mailers that work
Across 980,000 platform mailers, three types lead. The real-pen handwritten letter with a forever stamp (first-to-market and sensitive lists). The snapback check at 70 to 80% of the Zillow estimate in a window envelope (bulk pattern interrupt). The personalized family postcard (cheap, bulk).
For the leads you most want: a personalized foreclosure-options PDF mailed in a $10 to $12 priority envelope. Value first, then the ask.
| List type | Rehash cadence | Why |
|---|---|---|
| Tax sale | Every 30 days | Hard deadline, fastest forced cycle |
| Probate | Every 45 days | 60 loses deals, 30 annoys people |
| All other first-to-market | Every 90 days | Long nurture, the pipeline builder |
Not-interested campaigns re-call only the correct number. Rehash campaigns re-call the still-callable numbers on the same per-list cadence.
Two independent things set how you work a list. Saturation sets the caller caliber and touch quality. Sales-cycle speed sets the cadence.
Foreclosure is the most saturated niche and sits roughly 45 days from auction: give it your stronger caller and a premium touch. Fresh, low-saturation pulls (obituary, same-day courthouse records, senior, vacant, tired landlord) take the cheap sequence.
Stages 6-7: The Back Half of the Belt
Deep-prospect only what survived every cheaper touch. Then bolt on the courthouse layer for durable consistency.
The queue: records on a precision or courthouse list that survived call, text, voicemail, and mail with no contact. Nobody else can reach these leads, so the biggest spreads live here. One large deep-prospecting deal can make the year.
| Level | Situation | The move |
|---|---|---|
| L1 | Standard skip trace | Marketing grade: the top 5 of about 15 ranked numbers |
| L2 | Title or entity error | Ownership recorded wrong, or a trust or LLC. Find the signing member, skip-trace the actual person |
| L3 | Deceased owner | Pull the native obituary list, confirm ownership, find the heir. Most of the deep-prospecting money is made here |
| L4 | Curative or partition | Buy out heir percentages for control, or buy one share and force the sale legally |
Treat a deceased record as a positive signal to deep-prospect from the start, not a dead end. That is exactly why standard marketing failed on it. Full method: Deep Prospecting.
County filings take roughly 25 to 35 days to reach data brokers and then the platforms. A non-judicial foreclosure runs 3 to 4 weeks from filing to auction, so the deal is gone before any paid source shows it.
Pulling raw from the county makes you the first call. Cheapest per contract once built: $500 to $2,000. The build takes roughly 25 to 35 days with a 3 to 6 month cycle, and that difficulty is the moat. It feeds the same Stages 2 through 6 conveyor.
| Source | How to pull it |
|---|---|
| Online portal | Recorder, treasurer, probate clerk, or code-enforcement site. Set recurring calendar pulls: tax delinquency monthly, code violations weekly |
| In-person courthouse | Photograph the terminal results. Hire a runner on TaskRabbit (post under "Personal Assistant") or Investor Bootz, about $100 per weekly job. One trip covers 2 to 3 list types |
| Public notice | Required newspaper notices hold addresses, case numbers, auction dates. Claude converts the text to a structured CSV. Fully automatable into Slack each morning, tagged by source |
Run 3 to 4 niches only: tax delinquent, foreclosure, probate, tax sale. The First-to-Market County Data skill maps the office, portal, phone number, and difficulty for each niche in any county.
Tax roll: pull the latest roll from the county. Criteria off assessed value: around a $200,000 minimum with a $3,000 delinquency floor, which signals roughly 2 years behind. Ask for "people delinquent in their taxes for more than 2 years" or use-code 0100, never "tax deed sales."
Probate: filings give a case number and executor but no property address. The Probate Property Finder skill searches the assessor by decedent name and returns a Sift-ready CSV. Then run the owner swap to the executor or heir.
Foreclosure: pulling the lis pendens on Day 1 beats waiting on the Day 20-plus newspaper notice. Marketing window: Day 1 through Day 30.
The 200-record month, example stack
Below about 200 new records per month a precision strategy gets lumpy: deals some months, none others. Stack niches to clear it, or blend up to coverage-tier data. Scale precision first, then add geography.
Some operators run the whole cheaper flow only to surface the deep-prospecting set. They break even on the front half and earn all the profit on the back half.
Who Runs the Belt: The Data Manager
One seat owns Stages 1 through 4. It costs $500 to $700 a month and it is your first hire.
This SOP is the repeatable backbone every operator and VA runs against, regardless of market or niche. Hand it over on day one of onboarding. In Phase 1: The Operator, you run this seat yourself until the pulls are producing.
| Rhythm | Done means |
|---|---|
| Daily | Auto-add presets ran and new records are skip traced and tagged. Trestle pass done on new numbers. Dial queues set back to front. Attempt counters current, incremented by the team, not the dialer. |
| Weekly | Obituary auto-adds landed. Code-violation portal pull done. Courthouse runner drop processed with Claude and loaded into the conveyor. |
| Monthly | Tax-delinquency roll refreshed from the county. Suppression preset checked. Every precision strategy counted against the 200-record line. |
Hiring the seat, the interview questions, and the pay structure: the Hiring Guide covers the full process.
What Breaks the Pipeline
Every warning below is in the SOP because it happened. Each one has a one-line fix.
| What breaks it | The fix |
|---|---|
| AND-stacking distress on the AI score | Stacking runs 48.4 doors per deal versus 30.6 for the score alone. Run separate parallel pulls. |
| Leading with SMS or bulk text | Cold call is the first live touch. Text follows the call and the voicemail. |
| Cold text on deceased or obituary lists | Drop it entirely. Call, voicemail, mail, deep-prospect the heir. Never mention the death. |
| Dialing front to back | You double-dial the same people the same day. Start at follow-up 3 and work backward. |
| Suppression preset missing its two settings | Set unrelated parties AND a $1,000 minimum sale price, or inherited and probate properties get wrongly tagged sold. |
| Sending the same mailer twice in a row | Response decays. Alternate the three proven types. |
| Trusting the dialer to count attempts | Integrations cannot increment the attempt counter. The team marks each attempt complete. |
| One niche under 200 records a month | Deal flow goes lumpy. Stack niches or blend up to coverage-tier data. |
| Launching every niche and market at once | Too many data permutations paralyzes the operation. One county, one niche, then add. |
Where to Go From Here
This SOP is the worked example the SOP Creation guide builds from. Go deeper on any stage.
4-Week Deal Flow Workshop
Live Tuesdays, October 6, 13, 20 and 27, 1 to 4 PM ET
4 live Tuesdays with Ty and Tyler plus 4 Thursday Q&As, free, for investors who have already closed a deal. Join any week. Already in? Use this page to prep between sessions.
SOP Creation
How to write SOPs your team follows. This pipeline is its worked example.
Phone Scoring & Spam Management
The full Stage 4 method: activity scores, dial bands, and the connect-rate math.
SiftMap Mastery
The Stage 1 engine: presets, auto-add, precision pulls, and the native distressor lists.
Phase 1: The Operator
Where you run this conveyor yourself before hiring a single seat.
Hiring Guide
How to hire the $500 to $700 data manager seat that owns Stages 1 through 4.