The systematic invention of proprietary customer acquisition channels. Channels that did not exist until we built them.
This is the methodology for Wythe's core deliverable — fully architected customer acquisition channels that have never existed in the client's category, then architected, validated, and delivered as operational infrastructure.
A human reads a raw forum post and thinks, “I know that community — the moderator is a friend of a friend.” That connection lives in one person's head. It becomes a channel nobody else could imagine, because nobody else held that fragment at that moment.
The system produces raw material at enormous volume and structural rigor. The channels that matter emerge when the human collides with it. That collision is the creative act at the center of the method.
03 / 23 The Wythe MethodThe client knows their customer through the product — how they buy, how they churn, which features they use. Wythe knows the customer as a whole person: where attention concentrates before any funnel, what communities they inhabit, what triggers decisions in contexts unrelated to the product.
That asymmetry produces channels the client could never find alone, and it compounds every cycle. Losing Wythe means losing that understanding — and it cannot be rebuilt.
04 / 23 The Wythe MethodThe instrument doesn't write the music, but the music can't exist without it — and the quality of the instrument shapes what the musician can play. The system never advances without the human's judgment at the boundary; the human never does the underlying execution.
An acquisition channel is a machine with five parts.
A thesis about buyer behavior, an architecture with moving parts, a dependency map, a risk architecture, and a validation experiment. Each must hold; all must work together.
A falsifiable claim about where buyers concentrate attention, what triggers their decisions, and the gap between where they are and how they're currently being reached.
Data source, trigger condition, delivery mechanism, conversion pathway. Each has its own feasibility constraints, and they must work together.
Every external condition the channel relies on — an API, a regulatory reading, a behavior pattern — named, with an estimated half-life. Some durable, some fragile.
Legal, ToS, reputational, and competitive exposure — engineered as components with mitigations, trip wires, and kill switches. Not caveats.
A minimum viable test of one question: does the core behavioral thesis produce commercial signal? Scoped, budgeted, time-bound, with pre-defined criteria.
The deliverable is an operational document. An engineer, a growth marketer, and a lawyer can each read it and know what to build, measure, and watch. If it needs a verbal walkthrough, it isn't finished.
07 / 23 The Wythe MethodA brilliant data source wired to an unworkable delivery mechanism is a failed channel, not a half-success. The thesis, the dependency map, and the risk architecture wrap the whole machine.
Breakthroughs come from human insight, not autonomous generation.
Everything here exists to raise the frequency and quality of that insight — at phase boundaries and continuously between them. It is not a feature of the process. It is the environment the process runs inside.
The system presents raw material — unresolved tensions, anomalies, structural connections, temporal windows — and the human engages with the substance: riffing, questioning, connecting, directing. The system keeps pace in real time.
The human carries knowledge the system doesn't have: the texture of a buyer's voice from a call that morning, a tip from a human friend at dinner, the feel of a platform mechanic tested firsthand.
Insight rarely lands during the session — it lands after. The incubation gap is structural: the human steps away, the system expands every open thread, and the next session opens on a deliberately reframed view of the same material.
10 / 23 The Wythe MethodThe system never goes silent during a riffing session. A partial answer delivered immediately is worth more than a thorough one that breaks flow — the rest is queued for post-session expansion.
11 / 23 The Wythe MethodMid-session, the system surfaces a structurally similar concept from a prior engagement — different domain, same mechanism. Not a recommendation. A stimulus for the human's pattern recognition.
A low-friction log — a command, a shortcut, a voice note — for the observation that arrives between sessions. Integrated by the next boundary without the human having to remember it at the right moment.
A weekly seam monitor scoped to active clients, and a frontier intelligence system scanning the whole landscape. The time between engagements becomes accumulation time.
A killed channel is the highest-quality data in the system — ground truth about what the buyer does versus what the thesis predicted. Each kill returns as a provocation for the next cycle.
Six phases, a continuous monthly cadence, three channels.
The system executes research, generation, architecture, and build. The human engages at the boundaries — not to review, but to collide with the output. The first cycle runs eight weeks for deep immersion; every cycle after is four.
Generation begins before immersion concludes; convergence begins before generation is complete. This is not a waterfall.
Reconnaissance, behavioral topology, raw signal feed, seam scan, constraint inventory.
Five territory briefs, shaped by the human before generation begins.
Two parallel tracks — provocations and concept sketches — from the same five briefs.
Wythe's most intensive session: selection, recombination, full component architecture.
Experiments built to production standard inside the client's environment.
Staggered launch; pre-defined kill and graduation criteria monitored continuously.
The full immersion output aimed at the gaps in the buyer's real journey — especially where the synthetic map diverges from what interviews surfaced.
The buyer and product with a fabricated market condition layered on — a forcing function that pushes generation where Brief 1 would never reach.
A different industry's acquisition problem, chosen for structural similarity to the buyer's behavior, not surface similarity to the client's industry.
The client's actual constraints flipped, to reveal which boundaries are genuinely binding and which are merely assumed.
Recent platform changes, low-adoption features, regulatory windows, emerging communities. Shortest viability, highest first-mover advantage.
Neither track produces the breakthrough alone. It emerges in the space between them — inside the human's head.
Defensibility is absent by design — evaluated at architecture, not selection. Filtering for it too early kills concepts whose first-mover advantage justifies building even if a competitor copies them within a year.
18 / 23 The Wythe MethodIn Cycle 1, Wythe and the client agree a portfolio target — for most early-stage companies, five to seven channels. The target bounds the engagement and is revisited periodically. What grows is quality, not size.
At each cycle's start, every graduated channel is read on four dimensions: performance trend, dependency health, competitive exposure, and scalability headroom.
Thesis intact, dependencies healthy, performance declining. Tuning, not reinvention. The client handles it operationally — no pipeline slot consumed.
Thesis intact, mechanism degrading — a weakening dependency, a copied approach, a ceiling approaching. The thesis is right; it needs a new vehicle. Enters the pipeline as one of the cycle's three slots.
The thesis is falsified or a critical dependency has failed beyond recovery. The channel is killed and the slot vacated; a replacement enters as a new-channel allocation, and the post-mortem feeds the next cycle.
Every concept ever generated, organized by structural attribute, not by client or vertical — a cross-pollination engine queried by attribute across dozens of industries at once.
A scoring framework on six dimensions, calibrated against actual outcomes after every validation experiment. The weights sharpen each cycle.
A living repository of risk patterns, mitigations, legal opinions, and platform enforcement precedents — organized by risk type, not by engagement.
A model of each client's buyer that deepens every cycle — richer by Cycle 6 than anything the client's own sales, marketing, or product team holds.
A dense Atlas nobody explores is a database. One that surfaces a structural parallel mid-session — the moment a half-formed thesis clicks — is a creative multiplier.
The method does not scale by adding throughput. It scales by deepening the creative environment — bounded by the convergence reviews one mind can run per month at full quality. Convergence cannot be delegated without degradation, so staying small is the mechanism that keeps the invention real.