The warming pipeline · step by step

Here's what we actually do when you hand us a cold list.

Every step, no magic words. Upload to ranked call sheet is usually under five minutes — this is the order it happens in and what we read at each stage.

  1. 01

    You upload

    ~30 sec

    Drop a CSV, XLSX, TSV, or just paste names into a textarea. No need to clean it first, no column-mapping gymnastics. We parse whatever you drop in and show you a 5-row preview before we touch anything.

    Example

    A roofing contractor pastes 420 rows copied out of a storm-damage aggregator. Half the columns are garbage. The parser finds name, address, ZIP, and phone; the rest gets ignored.

  2. 02

    You tell us about your business

    ~30 sec

    Industry, target customer, location, price range. More context = sharper warming. It's a 30-second form, not a 20-field interrogation — and we remember it next time.

    Example

    Industry: Residential solar. Target: dual-income, owned-home 5+ years, no recent mortgage event. Service area: Phoenix metro + outlying ZIPs. Avg deal: $22k installed.

  3. 03

    We load your industry's correlation playbook

    ~5 sec

    Every vertical has a different signal stack. We pull the rule set tuned for yours — tuned (one of 12 flagships), learning (outside the flagships, sharpening weekly), or fresh (brand new, bootstrapped from your context).

    Example

    Auto sales looks at lease-end density + vehicle age + resale ZIPs. Solar looks at roof pitch + utility tier + HOA friction. Personal injury looks at ER visit patterns + adjuster ZIP density. Different industries, different signals.

    Input
    "Residential solar, Phoenix AZ metro"
    Output
    8 correlation rules loaded (roof-pitch, utility-tier, HOA-friction, ...)
  4. 04

    We build your ideal-customer profile

    ~20 sec

    From your context + what actually converts in your vertical, we formalize who you're hunting. Not 'adults 25-54' — specific. Specific enough that we can match against it row by row.

    Example

    "Dual-income households in ZIPs with >$85k median, owned home 5+ years, no recent refinance, high utility tier, detached single-family." That's the filter — not a persona deck.

  5. 05

    We read the community around each lead

    runs continuously

    We're not just looking at the name on your list. We're looking at the neighborhood around the address — median income, owned vs rented density, household composition, buying-behavior baseline for that ZIP.

    Example

    Row 184 lives at a specific address in 85032. That ZIP is 67% owner-occupied, median income $78k, utility tier 3. Row 184 inherits those community signals as context for every rule that runs.

  6. 06

    We layer local + industry + seasonal trends

    runs continuously

    Current market conditions in that geography, crossed with your vertical, crossed with time of year. Buying intent isn't static — the same lead is warmer in October than in April, depending on what you sell.

    Example

    Toronto auto sales in Q4 (lease-end wave, year-end incentives) looks nothing like Toronto auto sales in Q2. Phoenix solar in summer (bill shock) looks nothing like Phoenix solar in winter. We factor it in.

    Input
    ZIP 85032
    Output
    Median income $78k / 67% owned / utility tier 3 / 12% YoY solar-install growth
  7. 07

    We check real-time triggers

    refreshed daily

    News events, market moves, community-level catalysts that shift buying intent this week. We watch the signals that matter for your vertical and tag any lead that sits in their blast radius.

    Example

    Major local employer announces layoffs? Mortgage rates drop 40 bps? Big hailstorm rolls through a metro? We tag affected leads and push them up the ranking.

  8. 08

    We correlate your list against all of it

    ~2 min per 1k leads

    Each row gets checked against the full signal stack. Every correlation rule that fires gets logged with its confidence weight, so you can see exactly why a lead scored the way it did.

    Example

    Rule 'dual_income + recent_lease_end + owned_home' fires on row 184 with 0.78 confidence. That's logged against the lead — not hidden inside a black-box score.

  9. 09

    We rank and map

    ~instant

    Every lead gets a temperature — Hot, Warm, or Cold — and a human-readable 'Why' hint. Sort defaults to Hot-first, so the call sheet writes itself.

    Example

    Hot (call today): 38 leads. Warm (email this week): 112 leads. Cold (quarterly check-in): 97 leads. Each row shows the top 2 reasons it landed where it did.

    Input
    247 uploaded leads
    Output
    38 Hot (call today) / 112 Warm (email this week) / 97 Cold (quarterly)
  10. 10

    We learn from what happens next

    runs continuously

    Every outcome you log — Won, Lost, No Contact, Stale — tunes the weights for your vertical. Your results feed back into the playbook so the next list is sharper than the last.

    Example

    You mark 12 Hot leads as Won and 3 as Lost. The rules that fired on the Wons get their weights bumped; the rules that only fired on the Losts get tightened. Over time the engine gets smarter about YOUR market.

Try it with 50 leads, free.

No credit card. Watch the pipeline run on your own list before you commit to a plan.

Start free trialSee the 12 tuned verticals