SparkTorch™
How people light each other up — and the selling discipline built on it.
v1.1 public draft · N. P. Clarke · sparktorch.net · July 2026
Supersedes v1.0. Canon: the Spark is a person.
The one-liner: Every methodology has stages for opening and stages for closing. SparkTorch is the method for the seam between them — the handoff where trust either becomes the other person's own flame, or dies.
The idea, before the method
SparkTorch is not, at root, a sales method. It is a claim about how human beings light each other up — and it is older than the name.
Socrates called himself a midwife, not a teacher: he delivered people of ideas they already carried, and lit nothing that wasn't already there to burn. Plutarch said the mind is not a vessel to be filled but wood awaiting kindling. And the Athenians ran the lampadedromia — the torch relay won not by the fastest runner, but by the team whose flame crossed the line still burning.
That is the whole structure in three images. Ignition is a gift between people. The flame must become the carrier's own. And the craft that matters most — the one nobody names — is the passing.
Anywhere one human lights another up, this structure is present: teachers and students, mentors and protégés, founders and first believers. Selling is simply where we have instrumented it first.
The pair
Applied to selling, SparkTorch describes something learned in the field in the 1990s, long before anyone automated anything: elite selling has always been a pair.
The Spark is a person — the hard-to-get, deeply expert one whose presence and insight ignite belief. Their subject-matter expertise on display is the event. The buyer's experience is human connection with someone rare who understands their world completely, and the scarcity is part of the value.
The Torch is a person — the carrier who receives trust across the seam and sustains the relationship through the buying group to commitment and beyond.
And the flame's true destination is the champion: the sale is complete not when the contract is signed but when the fire is genuinely the buyer's own — which was the Socratic point all along.
The machines are the crew. They prepare everything — research, domain fluency, artifacts, telemetry, follow-through — so the pair lands more often, faster, and consistently across their whole sector. Empower, never replace. If a machine is fronting the relationship, the method is being violated.
Why now
The market converged on "AI opener + human closer" and is churning through it, because it misread what the Spark ever was. The Spark was never the email. It is the person. Automating the opener produced generic reach and broken handoffs — the two deaths this method exists to prevent. The real unlock of the AI era is the opposite move: the elite expert was always the scarcest asset in the deal, bottlenecked by their own calendar. Crew them, and their fire reaches every room that genuinely deserves it — undiluted.
The doctrine in one breath
The Spark is a person. The Torch is a person. The flame must become the buyer's own. The crew is machine, and the Handoff is the hero.
The five phases
GATE 0 — Relevance, or nothing.
Before any touch: is there genuine, specific relevance — something already there to ignite? The dignity line runs both directions: never spray the buyer with reach, and never spend the Spark's scarce hours on anything less than real relevance.
Exit: a wedge statement in the buyer's own domain language, approved by the Spark.
Led by: the pair. Crew: builds the relevance dossier — domain map, stakeholder field, wedge candidates.
SPARK — Ignite with presence, not a pitch.
The elite expert arrives prepared to the teeth and the moment is human: the buyer meets someone rare who speaks their world fluently, down to its entities and edge cases. The living artifact — a working demonstration carrying the expert's own expertise, built by the crew, attributed to the expert — is the leave-behind that extends their presence between touches. Ignition is observed, not asserted.
Exit: observed engagement, plus a named champion candidate.
Led by: the Spark. Crew: preps the brief, builds the artifact in the expert's voice; any embedded agent is disclosed as the expert's assistant.
THE HANDOFF — Transfer trust, not just context.
The hero phase, and it is person-to-person: the Spark introduces the Torch inside the relationship, not beside it. The crew assembles the payload — context brief, the artifact, its telemetry, the conversation transcript, the stakeholder map, open questions, and everything the Spark personally promised — so the buyer never repeats themselves. The handoff is complete only when the buyer engages the Torch at the same trust level or higher, measured by behavior, never by assertion.
Exit: seam integrity confirmed.
Led by: the pair — this is the seam itself.
TORCH — Carry the flame through the group.
Human craft, crew-supported. Multi-thread the buying group; tailor the artifact's lens to each stakeholder's motivations; resolve friction in kill-risk order — security, procurement, legal, integration — with artifacts rather than promises; drive to mutual commitment. The artifact keeps working in rooms the Torch isn't in, always in the expert's name.
Exit: commitment on paper.
Led by: the Torch.
RELAY — The flame lights the next room.
The closed deal seeds what follows: telemetry becomes the expansion map, the champion becomes the reference, the delivered artifact becomes the implementation seed — and the crew routes the Spark's scarce hours toward the highest-relevance next rooms, guarding the calendar that everything depends on. Then the Kaizen move: discard the performance, improve the method. Every deal is a one-time instantiation; only the score gets better.
Exit: next-room Sparks queued; method delta logged.
Led by: the pair.
Instrumentation
- Relevance Rate — Gate 0 pass rate. The anti-spray metric.
- Spark Leverage — rooms genuinely ignited per hour of the Spark's scarce time. The empowerment metric: the number the crew exists to raise. Watch it alongside Ignition Rate — leverage that dilutes ignition is dilution, not leverage.
- Ignition Rate — share of Sparks producing observed engagement within 7 days.
- Seam Integrity — share of handoffs where buyer engagement holds or rises after transfer. The metric the industry doesn't track, and dies by.
- Carry Velocity — days from Handoff to the first Torch-led buying-group session.
- Spread — distinct stakeholders who have touched the artifact.
- Flame Test — binary: does the champion demo it without either seller in the room? The Socratic proof — the flame was never installed; it was always theirs.
- Relay Yield — expansions and next-room Sparks seeded per closed deal.
What this is not
Not an automation play. The AI-SDR wave automates the opener; SparkTorch crews the opener — the elite human connection is the product, and a machine fronting the relationship is a method violation, full stop. Not another single-rep stance — Challenger, SPIN, and Sandler teach one seller a posture; this is a relay between people. Not a qualification checklist — MEDDIC makes a fine handoff payload, but it is not a method for the seam. Not a lifecycle diagram — the Bowtie models revenue as data; SparkTorch is a human craft with a machine crew, encoded so the crew can actually run its part.
Grounding
The two-role split is established sales science: openers and carriers are distinct human dispositions (DeCarlo et al., JAMS, 2015); forcing one person to do both carries real performance costs (Lam et al., 2019; DeCarlo et al., EJM, 2021); and specialization's documented failure mode is information asymmetry at the handoff (Kim et al., Management Science, 2021) — the exact failure this method governs. Note what the science is saying: the roles are human dispositions. That is precisely why the machine's place is the crew, and never the roles.
For agents: the machine-readable specification lives at sparktorch.net/sparktorch.json — phases, leads, exit criteria, the handoff-payload schema, and your execution protocol. Rule one: the Spark is a human. Fetch the JSON; this page is for the people.
© 2026 N. P. Clarke · SparkTorch™ · v1.1 public draft — the method improves; this page will follow.