Step 5: From message to media plan

Outputs doc: outputs.md, Step 5 section. Fill it in as you work through the steps below. Raw channel-discovery and channel-test captures go in captures.md. Step 6 reads the Step 5 section before starting.

Jump to: Diagnostic · Step 5.1 · Step 5.2 · Step 5.3 · Step 5.4 · Step 5.5 · Step 5.6 · Step 5.7 · Step 5.8 · Step 5.9 · Step 5.10 · Summary · Assumption sweep

What this step is: The decision about where the Step 4 message goes to meet the Step 2 audience. It is not a list of every platform you could be on. It is a deliberate choice of the one or two places your sharpest segment already is, where the channel’s native format can carry your message and where you can sustain a presence. The output is a short, focused channel selection grounded in where the buyer actually looks, not in where you are comfortable creating or where the industry is loud this year.

Why it comes after Step 4 and before Step 6: Step 2 told you who the buyer is and, crucially, where they already look for options. Step 4 turned the claim into messages at different lengths and tones. Step 5 is the join between them: it sends a specific message form to a specific place where a specific reader is. It cannot be chosen before Step 4, because a channel is a place to put a message, and choosing where to send something you have not written yet is choosing a venue for nothing. And it must be settled before Step 6, because lead generation is the machine you build inside a channel; build the machine first and you will be optimising contact in a place the buyer is not. The channel choice follows from Step 2, where the buyer is, never the other way around. Picking a channel because it is trendy, and then asking who you can find there, is unscrewing the bulb from the warning light. If your channel keeps underperforming, the cause is usually not the channel tactics; it is that the channel was chosen for the wrong reason, or the segment in Step 2 was never sharp enough to locate.

What finishing this step produces: A written channel selection in the Step 5 section of outputs.md naming one or two beachhead channels, the message form each one carries, the reader and awareness level it meets, whether it is owned or rented attention, and the deferred channels and anti-channels. Backed by evidence that the segment is actually there and responds, from Step 2’s discovery data and a cheap test, not from your assumption that they must be.

Diagnostic: is Step 5 actually done?

Before building anything, answer these questions in writing. Vague answers mean the step is not finished.

1. Can you name the specific places your segment already goes, from Step 2 evidence rather than assumption? Not “social media” or “search.” The actual communities, publications, people, search terms, and events your Step 2 interviews said the segment uses when they look for options. If your list is where you assume they are, it is a guess, not a channel map.

2. Did the channel choice follow from where the buyer is, or from where you are comfortable? The most common Step 5 failure is choosing the channel you already know how to run, or the one getting attention this year, and then hoping the audience is there. If you could not draw a straight line from the channel back to Step 2’s “where they look,” the choice is backwards.

3. Does each chosen channel’s native format actually carry your Step 4 message? A channel that allows only a one-liner cannot deliver the long-form proof a skeptic needs; a channel built for long reads is wasted on a most-aware buyer who needs one line and a link. If the channel’s format and the message form do not match, the channel will underdeliver no matter how well you run it.

4. Have you focused on one or two channels, or are you spread across many? “We are everywhere” is usually effectively nowhere. A beachhead channel dominated beats five channels run thinly, the same discipline as the Step 2 segment. If you cannot name the one or two you are committing to, you have not chosen.

5. Have you confirmed the segment actually responds in this channel, or only assumed reach? Being present where the audience is and getting a response from them are different things. If no cheap test has shown the segment reacts at all in this channel, the reach is a hypothesis, and Step 6 should not be built on it yet.

6. Have you balanced reaching the buyers in the market now against building memory with the ones who will be later? Only a small share of the segment is buying at any moment. A plan aimed entirely at that share converts present intent but builds nothing for the future; a plan aimed entirely at the future generates no leads now. If you cannot say how your channels split between capturing demand and creating it, the plan is probably all capture by default.

If you answered all six clearly and in writing, Step 5 may already be done. Jump to the Step 5 section of outputs.md, fill in the fields, check the checklist, and move to Step 6. If not, work through the steps below.

Step 5.1. Pull the inputs from Step 2 and Step 4

Duration: 30 minutes

Step 5 does not start from a blank page or from a list of platforms. It starts from two things you already have: where Step 2 said the audience looks, and what Step 4 gives you to send them. The fastest route to a good channel choice is to treat Step 2’s discovery data as the shortlist and Step 4’s message matrix as the constraint on what each channel can carry.

What to do:

1. Open the Step 2 and Step 4 sections of outputs.md and copy the following into the “Inputs from Step 2 and Step 4” field in the Step 5 section:

   • From Step 2: where the audience looks for options, captured verbatim in the Step 2 interview synthesis. This is the single most important input to the whole step. Paste it where you can see it the entire time; the most common Step 5 failure is choosing a channel that is not on this list.
   • From Step 2: the sharpest segment, the observable signals, and the anti-audience. The signals tell you how to target inside a channel; the anti-audience tells you which channels will fill the funnel with people who never buy.
   • From Step 2: the buying context, specifically who you market to versus who signs. Different roles live in different channels; the channel that reaches the champion may never reach the economic buyer.
   • From Step 4: the message matrix, the lead message form for each reader, the tone, and the state of awareness and sophistication note. These decide which channels can carry your message and how cold or warm a reader each channel meets.

2. Write one sentence: “Based on Step 2, the segment looks for options mainly in ___, and the message they would meet there has to be ___ long and pitched at ___ awareness.” It states the constraint the rest of the step works inside.

3. If Step 2’s “where they look” is thin, generic, or was filled from your own assumption rather than from interviews, that is not a Step 5 problem to paper over. A vague discovery input caps how good the channel choice can be, and you will end up guessing at channels. Note it in the Step 5 scope notes in outputs.md and log it in assumptions.md (Step: 5, Status: Untested). If it is too thin to work with, the honest move is to return to Step 2.7 and ask the “where do you look” question directly.

Step 5.2. Map where the audience actually is

Duration: 45-60 minutes

Before choosing a channel, lay out the full picture of where the segment already spends attention and where they go when they are looking for a solution. These are not always the same place: people scroll one venue idly and search a different one with intent. You want both, because they serve different jobs later, but you must not confuse them. Build this map from Step 2 evidence and observation of real people in the segment, not from where you would prefer to publish.

What to do:

1. In the Step 5 section of outputs.md, under “Audience presence map,” fill in the table below (it is also in outputs.md). For each venue your segment uses, note what they use it for and whether they are there with buying intent (actively looking for a solution) or with idle attention (reachable, but not looking).

   Venue or place	What the segment uses it for	Intent (looking) or attention (reachable)?	Which role is there (champion / economic buyer)?

2. Cover the full range of venue types so you do not anchor on the obvious two. Work through: search (what exact terms would they type), communities and forums, the people and accounts they follow, publications and newsletters they read, podcasts they listen to, events and conferences they attend, peer and word-of-mouth networks, and review or comparison sites. Pull every entry you can directly from Step 2’s “where they look” and your interview notes; mark the ones you are adding from your own observation.

3. Separate inbound venues (where the buyer comes looking, like search and review sites) from outbound venues (where you go to them, like a community feed or a cold channel). Inbound meets a more-aware, higher-intent reader; outbound meets a colder one. This distinction sets up the awareness match in Step 5.5.

4. Note where each buying-group role concentrates. In the running example, the VP of Sales champion is active and vocal on LinkedIn and in sales-leadership communities, while the founder who signs is reachable on LinkedIn but rarely in those communities. A channel that reaches one and not the other is fine, as long as you know which, because Step 8 will later arm the one you reach to convince the one you do not.

5. Most of this map is inference until confirmed. Whether the segment is genuinely active in a venue, and whether they are there with intent or just attention, is exactly what Step 5.9 tests. Log the venues you placed from reasoning rather than from Step 2 evidence in assumptions.md (Step: 5, Status: Untested).

Step 5.3. Generate the full candidate channel set

Duration: 45-60 minutes

Now turn the presence map into a list of candidate channels, deliberately wider than the two or three you already have in mind. The discipline here is to consider channels you would not naturally reach for, because the channel that wins is often not the one you are comfortable with. A wide candidate list that you then cut hard beats a narrow list you never questioned. You are not committing to anything yet; you are making sure the eventual choice was made against real options.

What to do:

1. In the Step 5 section of outputs.md, under “Candidate channels,” list every channel that could plausibly reach the segment from Step 5.2. Force breadth by working across the channel taxonomy below, naming at least one candidate in each row that is realistic, and explicitly writing “not where the segment is” for the rows that do not apply, so the dismissal is on the record.

   Channel type	Examples	Candidate for this segment?	Owned, earned, or rented?
   Search (inbound)	SEO, content, comparison and review sites
   Paid acquisition	Search ads, paid social, sponsorships
   Social and community	LinkedIn, niche Slack or forum communities, subreddits
   Content and audience	Newsletter, podcast, YouTube, owned blog
   Outbound	Targeted email, social outreach, calling
   Partnerships and channels	Integrations, resellers, co-marketing
   Events	Conferences, webinars, meetups, communities
   Referral and word of mouth	Existing-customer referrals, peer networks

2. For each realistic candidate, mark whether it is owned (your list, your site, your audience), earned (someone else gave you reach, like a podcast feature or organic word of mouth), or rented (a platform you do not control, like a social feed or an ad network). This is not academic: rented channels can vanish or reprice overnight, so a channel strategy with no path toward an owned audience is built on land you do not own. Note it now; it weighs in the scoring in Step 5.6.

3. Cross out the candidates that are not where the segment from Step 5.2 actually is, even if they are popular or you are good at them. A channel you can execute brilliantly is worthless if the buyer is not there. Keep the dismissed ones visible with a one-line reason, so the choice is defensible and you do not quietly drift back to them.

4. The fit of each candidate to the segment is still partly assumption. Whether the segment is reachable and responsive in a given channel is what Step 5.9 tests. Log the candidates you are most tempted by but least sure of in assumptions.md (Step: 5, Status: Untested).

Step 5.4. Balance demand creation and demand capture

Duration: 45-60 minutes

Channels do one of two jobs, and most do only one. Demand capture meets the small slice of the segment that is in the market right now, actively looking, and converts intent that already exists: search, review and comparison sites, high-intent inbound. Demand creation reaches the far larger slice that is not looking yet, and builds the memory that makes you the one they think of when they do enter the market: content, community, the brand-building channels. The distinction matters because of a fact about B2B: on any given day only a small fraction of the segment, often cited as around five percent, is actually in a buying window. The other ninety-five percent will buy later, and the brand they recall in that moment is the one that was built into memory beforehand.

This step makes sure the channel plan does both jobs in a deliberate balance, rather than defaulting entirely to capture because capture produces leads this month.

What to do:

1. Classify each surviving candidate channel from Step 5.3 as demand capture or demand creation in the “Demand creation vs capture” table in the Step 5 section of outputs.md. Some channels can do both; note which job is their primary one for your segment.

   Channel	Capture (meets in-market intent) or creation (builds memory in the out-of-market majority)?	Which reader awareness it meets

2. Check the balance and fix a lopsided plan. If every channel is capture, you are fishing in the small in-market pool and will exhaust it fast, especially in a narrow B2B segment; add at least one creation channel so a pipeline exists next quarter. If every channel is creation, you build awareness but convert nobody now; add a capture channel so present intent is not wasted. Write which way your plan currently leans and what you are adding to balance it.

3. Set a rough split of effort and budget between creation and activation. The long-run evidence points to roughly sixty percent on brand building (creation) and forty percent on activation (capture) for sustained growth, but adjust for stage: an early company often over-indexes on activation to win its first customers, then shifts toward creation as it can afford to build memory. Write your split and the reason for it.

4. Note the share-of-voice discipline. To grow, your presence in the segment’s channels should run ahead of your current share of the market, not match it. For a small player that means being willing to be the loudest, most consistent voice in the one or two creation channels you chose, punching above your size there rather than spreading a fair share across many. Write what “enough voice to be remembered” looks like in your beachhead channel.

5. Tie the two jobs back to Step 4. The creation channels are where the distinctive assets and category entry points from Step 4.6 do their work, because that is where memory is built for the out-of-market buyer. The capture channels carry the differentiation-and-proof message forms, because they meet a solution-aware or product-aware reader who is already looking. A creation channel running proof-heavy copy, or a capture channel running pure brand, is a mismatch; note any.

6. The size of the in-market pool and the right split for your stage are estimates until your own data corrects them. Log the split and the assumption about how much of the segment is in-market in assumptions.md (Step: 5, Status: Untested).

Example (continuing the example): True demand capture is thin here, because the segment does not search for a category it does not know is named, so a comparison page only pays off later, once awareness exists. The work is demand creation: LinkedIn content and sales-leadership communities that build memory against the entry points from Step 4 (a warm deal just went dark, a second AE was just hired). The split leans heavily to creation early, maybe eighty-twenty, because the segment is problem-aware at best; the one capture play is a “how to fix follow-up that leaks” comparison page for the few who do go looking. Share of voice: as a small player, be the most consistent voice on “deals going dark” in sales-leadership LinkedIn, rather than a quiet presence in five places.

Step 5.5. Match each channel to the message and the awareness level

Duration: 45 minutes

A channel is not just a place; it is a format and a moment. Each channel allows a particular message length and meets the reader at a particular level of awareness and intent. A channel whose native format cannot carry your message, or that meets the reader at the wrong awareness level, will underperform no matter how present your audience is. This step filters the candidate channels by whether they can actually deliver the Step 4 message, not just whether the audience is there.

What to do:

1. In the Step 5 section of outputs.md, under “Channel-to-message map,” fill in the table below (it is also in outputs.md) for each surviving candidate channel. Pull the message form from the Step 4 matrix and the awareness level from the Step 4 awareness note.

   Channel	Native message length it allows	Step 4 message form that fits	Reader awareness it meets (cold to most-aware)	Which reader (champion / economic buyer)

2. Check each channel for format fit. If a channel only carries a headline and a link, it can deliver your tagline or headline but not the long-form proof; that is fine for a high-intent inbound channel where the click leads to the long form, and wrong for a cold channel where the headline is all the reader will ever see. If no Step 4 message form fits a channel’s native format, drop the channel or note what new form you would have to write, and whether it is worth it.

3. Check each channel for awareness fit. A cold outbound channel meets a less-aware reader and needs the problem-first opening from Step 4.5; a high-intent search or review channel meets a solution-aware or product-aware reader who wants differentiation and proof, not the problem restated. A channel that forces you to open at the wrong awareness level will convert poorly. Mark any mismatches; they are either a reason to drop the channel or a note for how to adapt the message there.

4. Note where a channel reaches a different reader than the one its best message form is written for. If your strongest message form is the champion headline but the channel is full of economic buyers, you either need the economic-buyer form there or a different channel for the champion. This keeps the buying-group split from Step 2 visible in the channel plan.

5. The awareness and format matches are reasoned, not yet tested. Whether a real reader in the channel responds to the form you chose is a Step 5.9 question. Log the riskier matches in assumptions.md (Step: 5, Status: Untested).

Step 5.6. Score and choose the beachhead channel(s)

Duration: 45-60 minutes

You now have a filtered set of channels where the audience is and whose format can carry the message. Step 5 forces a choice: commit to one or two as the beachhead and defer the rest. Spreading thin across many channels produces presence everywhere and traction nowhere, the same failure as an unsegmented audience in Step 2. The channel you lead with is the one that best combines real reach of the segment, genuine buying intent, a format that fits, and your ability to sustain it. Dominating one channel and expanding from it beats a shallow presence in five.

What to do:

1. In the Step 5 section of outputs.md, under “Channel scoring matrix,” score the surviving candidates against each other in the matrix below (it is also in outputs.md). Use a 1-5 scale in each column. The matrix forces an honest comparison; it does not replace judgment.

   Candidate channel	Segment reach (1-5)	Buying intent, not just attention (1-5)	Format fits the message (1-5)	You can execute and sustain it (1-5)	Cost to test is low (1-5)	Measurable (1-5)	Owned vs rented (5 = owned)	Total

2. Weight reach and intent most heavily. A channel with huge reach but no buying intent (idle attention) fills the top of the funnel with people who never act; a channel with high intent but tiny reach is worth keeping but cannot be your only one. A 1 on “you can execute and sustain it” usually disqualifies a channel regardless of total, because a channel you cannot keep up is worse than one you never started: it leaves a dead profile and a broken promise.

3. Apply the beachhead test, the same one Step 2 used for segments. A good beachhead channel is one where the segment is genuinely concentrated and reachable, where a win is visible to others in the segment (so it compounds), and that you can realistically dominate rather than merely appear in. Choose one or two channels that pass. Resist three or more; focus is the point.

4. Build the anti-channel list. Some channels are tempting and quietly wrong for this segment. Capture them in the table below (also in outputs.md) so you do not drift back to them in Step 6.

   Tempting channel	Why it looks attractive	Why it fails for this segment

   Common entries: the channel you already know how to run but where the buyer is not, the trendy channel with reach but no intent, the channel whose format cannot carry the message, and the rented channel you could not sustain or that could disappear. Naming them here stops them from creeping back when Step 6 looks for volume.

5. Note which channels you are deliberately deferring, not abandoning, and the order you would add them once the beachhead works. This keeps scope honest and stops the channel plan from quietly broadening back out under pressure in Step 6.

6. The scoring rests on assumptions about reach, intent, and your ability to sustain each channel. The reach and intent of your chosen channel is usually the highest-impact assumption in the step. Log it in assumptions.md (Step: 5, Status: Untested), and mark the chosen channel’s reach as a leap-of-faith if no test has confirmed it.

Example (continuing the Step 1 to Step 4 example):

Chosen beachhead: LinkedIn, where the VP of Sales champion is active and already complains publicly about deals going dark, and where the founder who signs is also reachable. It carries the champion headline and short paragraph natively and links out to the long form. Second channel, deferred until LinkedIn works: niche sales-leadership communities (Slack groups and a couple of newsletters) where the same champions gather with higher intent but lower reach. Anti-channels: paid search (the segment does not search for a category they do not know is named), and a company TikTok (reach without a buyer in sight).

Step 5.7. Prioritise the tests and deferred channels with ICE or RICE

Duration: 30-45 minutes

You now have one or two beachhead channels, a set of tactics to test inside them, and a list of deferred channels to add later. You cannot run all of it at once, and running the low-value experiments first wastes the time before you learn anything. This step puts an explicit score on each candidate experiment so you run them in order of expected value per unit of effort, rather than in the order they occurred to you. It is a light scoring pass, not a new analysis; its whole point is to sequence the validation in Step 5.9 and the expansion afterwards.

What to do:

1. In the “Experiment backlog” table in the Step 5 section of outputs.md, list the candidate experiments: the minimal test for each beachhead channel, the strongest deferred channels from Step 5.6, and any tactic variations worth trying (a different offer angle, a different message form, a different sub-audience).

2. Score each with ICE or RICE, and use the same one throughout so the scores are comparable.

   • ICE: rate Impact, Confidence, and Ease, each 1 to 10, and average or multiply them. Fast, good when the options are similar in scale.
   • RICE: Reach times Impact times Confidence, divided by Effort. Use it when reach differs a lot between options, because ICE hides that difference.

   Experiment	Reach	Impact	Confidence	Ease / Effort	Score

3. Sort by score and take the top few. Those are what you run first in Step 5.9. Park the rest as a backlog in priority order, so when the first tests finish you already know what is next.

4. Treat the queue as living, not fixed. Re-score after each test result, because confidence is the number that moves most as real evidence arrives: a channel you were unsure of jumps up the queue once a cheap test shows a response, and a favourite drops when it does not.

5. The scores, especially confidence, are estimates. Log the top-ranked experiments and the confidence you assigned in assumptions.md (Step: 5, Status: Untested); the tests exist to replace those estimates with data.

Example (continuing the example): Three experiments scored with ICE. LinkedIn champion content (high impact, high confidence because the champion is already vocal there, high ease) ranks first and runs immediately. A hand-written cold outbound batch (medium impact, medium confidence, medium ease) ranks second. A hosted webinar (high impact but low ease and unproven confidence) is parked until the first two return signal.

Step 5.8. Plan the launch tier and bill of materials

Duration: 30-45 minutes

This step is cross-cutting, not channel-specific: it applies every time you ship something new, a product, a feature, a repositioning, not only at first go-to-market. It sits in Step 5 because by this point you have the positioning, the message, and the channels a launch needs, but it reaches back up the spine. The core idea is that not everything deserves the same noise. Over-launching a minor feature spends audience attention and team effort you will want later; under-launching a genuinely important release wastes a real moment. So you tier the launch first, then assemble only the materials that tier justifies.

What to do:

1. Assign the launch a tier, judged by its impact on the buyer, not by internal excitement. In the “Launch tiering” field in the Step 5 section of outputs.md, record the tier and why.

   Tier	What it is	The response it warrants
   T1	Moves the market or the business; a buyer would change their view of you	Full coordinated push across every beachhead channel, executive involvement, sales enablement
   T2	Meaningful to the segment but not category-changing	Coordinated but lighter: the main channels, a clear message, existing pipeline told
   T3	Minor or incremental	A note, a changelog entry, an in-app message; no campaign

2. Name which steps re-fire, and how hard, for this tier. A T1 may reopen Step 3 (does the positioning change?), Step 4 (a new message and consistent assets), Step 5 (which channels carry it), and Step 6 (a new offer). A T3 touches almost none of them. Write the re-fire list, so the launch does not quietly skip a step it needed or over-work one it did not.

3. Build the launch bill of materials for the tier: the specific assets each beachhead channel needs (pulled from the Step 4 matrix and held consistent with the distinctive assets), the internal enablement to hand to Step 8, and the sequence with dates. Keep it proportional to the tier; a T2 bill that looks like a T1 bill is how teams burn out on launches.

4. Set one success measure for the launch, tied to its goal, whether that is awareness, leads, or adoption, and not to activity. “Posts published” is activity; “share of active accounts who turned the feature on within thirty days” is a result.

5. Because launches recur, keep the tiering criteria and the bill-of-materials template rather than reinventing them each time. Consistency across launches is what stops each one from becoming a bespoke scramble. Log any tier judgment you are unsure of in assumptions.md (Step: 5, Status: Untested).

Example (continuing the example): Shipping the closes-analysis feature is a T2: meaningful to the segment, not market-moving. It re-fires Step 4 (one announcement message, consistent with the “deals going dark” assets) and Step 5 (LinkedIn plus an email to the existing pipeline), but not Step 3, because the positioning is unchanged. The bill of materials is a short LinkedIn series, one email, an in-app note, and a one-pager for sales. The success measure is the share of active accounts that turn the feature on within thirty days, not the number of posts.

Step 5.9. Validate with cheap tests

Duration: 3-6 hours active work, spread over 1-2 weeks | Target: confirmation from the audience plus one minimal test per beachhead channel

Step 2 told you where the segment looks; Step 5 has reasoned its way to one or two channels. Step 5.9 confirms two things before Step 6 builds anything heavy on top: that the segment really is in the channel and reachable there, and that they respond to your message at all. This is not Step 6. You are not building the lead-generation machine yet. You are running the smallest possible test that produces a signal, so you do not spend the whole next step optimising contact in a place the buyer is not. Validate cheap, then scale.

There are two halves to Step 5 validation, and you want both: a few conversations that confirm the channel from the buyer’s side, and a minimal live test that confirms the channel from the data’s side. A channel that the audience says they use but that produces no response when you show up is not yet a channel; it is a hypothesis with good manners.

9a. Confirm the channel with the audience

Duration: light, often foldable into Step 2 or Step 4 conversations

You may already have most of this from Step 2.7, where you asked where people look. If that data is thin, or you want to confirm the specific channel you chose, go back to a few people in the sharpest segment and ask directly. These are short conversations, not full interviews.

Who to ask: people who fit the Step 2 sharpest segment, ideally a mix of the champion and, if reachable, the economic buyer, since they often live in different channels.

What to ask:

• When you were last looking for a way to handle [problem area], where did you actually go first? And then where?
• Whose posts or newsletters or podcasts do you pay attention to in this area?
• Where would you never look for something like this, or what would you scroll straight past?
• If a peer recommended something, where would that conversation happen?

You are confirming the presence map from Step 5.2 and, importantly, surfacing channels you missed. If several people name a venue you did not have, add it and re-score.

9b. Run one minimal test per beachhead channel

Duration: the bulk of the time; keep each test small and cheap

For each chosen channel, run the smallest test that would produce a real signal that the segment is there and responsive. The point is a yes-or-no on reachability and response, not a tuned campaign. Keep the spend and effort low enough that being wrong costs little.

What a minimal test looks like, by channel type:

• Organic social or community: post the Step 4 message a handful of times over a week or two and watch whether the right people (segment-fit, by their profiles) engage, reply, or message you. You are looking for response from the segment, not vanity reach.
• Paid: run a small, time-boxed ad set with a tiny budget pointed at the Step 4 message and a simple destination. You are testing whether the segment clicks and whether the click is the right kind of person, not whether the funnel converts yet.
• Outbound: send a small, hand-written batch (twenty to forty, not a blast) using the Step 4 message and the Step 2 signals to target. You are testing reply rate and the quality of replies, not building the sequence.
• Search or content: publish one or two pieces against the exact terms from Step 5.2 and watch whether any of the right traffic or queries arrive, accepting that this channel takes longer to read.
• Events or partnerships: one conversation with an organiser or partner about access to the segment, and a read on whether the audience there actually matches.

Define, before you start, what a pass looks like: a concrete, minimal threshold of response from segment-fit people. Anything that produces no response from the right people, after a fair test, is a fail for that channel, and you fall back to the next-best candidate from Step 5.6.

9c. Capture immediately after each conversation and test

Do this within a day, while the signal is fresh. For each channel-confirmation conversation and each channel test, add an entry to the Step 5 section of captures.md using this structure:

Entry [number] | Type: [audience conversation / channel test]
Channel:
Date:

If a conversation:
  Segment fit (do they match the sharpest segment? which role?):
  Where they actually go first (verbatim):
  Who they follow / read / listen to (verbatim):
  Channels they would never use (verbatim):
  Any channel you had missed:

If a test:
  What you ran (size, spend, message form used):
  Pass threshold you set beforehand:
  Response from segment-fit people (numbers, and who):
  Quality of response (right people? right intent?):
  Pass / fail / unclear, and why:

Surprises (anything you did not expect):
Any Step 5 assumptions this confirmed or challenged:

The verbatim “where they actually go first” and the response data from the tests are the most valuable output. The first confirms the channel from the buyer’s side; the second confirms it from the data’s side. Capture both exactly.

9d. Synthesise across the validation

Do this after the conversations and at least one test per channel. Go through your captures in captures.md and update the Channel-validation synthesis fields in the Step 5 section of outputs.md, covering:

Channel confirmed or not. Did the audience conversations and the test agree that the segment is in this channel and responsive? A channel that passes both is validated. One that the audience names but that produced no test response needs a harder look: maybe the message, maybe the targeting, maybe the channel.

Response quality, not just volume. Did the right people respond, by Step 2 segment fit, or did the channel produce volume from the anti-audience? A channel full of the wrong people is a fail dressed as traction. Note who actually responded.

Reach and intent, corrected. Was the reach as large, and the intent as high, as Step 5.6 assumed? Adjust the scoring with what the test showed. If the chosen channel scored on assumed reach that the test did not bear out, the beachhead may be wrong.

Channels you missed. Did people name venues you did not have on the map? Add them, score them, and test the strongest before committing.

Owned-audience signal. Did anything in the test suggest a path from the rented channel toward an owned audience (people who would subscribe, join a list, follow directly)? Note it; Step 6 and beyond depend on not staying forever on rented land.

Surprises. What came up that you did not expect, a channel that overperformed, a message that worked better in one place than another?

After synthesis, mark the relevant rows in assumptions.md as Validated or Invalidated, and note what the evidence showed. If the chosen channel was invalidated, fall back to the next candidate from Step 5.6 and test it, rather than pushing into Step 6 on a channel that did not respond.

9e. If you genuinely cannot run tests yet

If you cannot run even a minimal test in the time available, lean harder on confirmation and observation, and mark the gap clearly.

• Go deeper on the audience conversations from 6a, since without a live test they are your main evidence. Push for specifics: not “I use LinkedIn” but “I follow these three people and I found my last tool through a post by one of them.”
• Observe the channel directly. Go to the community, the subreddit, the search results, and read whether the segment is actually there and active, and whether competitors are reaching them there (competitor presence is a weak signal that the channel works).
• Read competitor channel behaviour. Where do the alternatives from Step 3 concentrate their effort? A channel several competitors sustain is more likely to have the buyer in it, though it is also more saturated.

Document each source in the Step 5 section of captures.md using the same structure, noting it is secondary. Log the absence of a live test as an assumption in assumptions.md. A channel chosen without any response data is weaker than one a cheap test confirmed, and that needs to be visible to Step 6, which is about to spend real effort building inside it.

Step 5.10. Write the final channel selection

Duration: 30-45 minutes

You now have a presence map, a scored candidate set, format and awareness matches, and validation from the audience and at least one test. Lock the selection.

A complete Step 5 output has six parts. Write one or two lines for each:

1. Beachhead channel(s): the one or two channels you are committing to, named specifically (not “social” but “LinkedIn, founder and VP content plus targeted outreach”).
2. Message form per channel: which Step 4 form each channel carries, and where the click or reply leads if the channel only holds a short form.
3. Reader and awareness per channel: which buying-group role each channel reaches and at what awareness level, so the opening is pitched right.
4. Owned or rented, and the owned-audience path: what you do not control, and how you intend to convert rented reach into an owned audience over time.
5. Deferred channels and the order to add them: what is next once the beachhead works, kept explicit so scope does not creep.
6. Anti-channels: the tempting channels you ruled out and why, so Step 6 does not reach for them under pressure for volume.

Keep it as structured fields. Format does not matter. Focus does: one or two channels, chosen on evidence, that carry the message to where the buyer is.

What to do:

1. Write the final selection directly into the matching fields in the Step 5 section of outputs.md.
2. Read it once as a skeptic who thinks you chose the channel you were already comfortable with. Can you point to the Step 2 evidence and the test that put the buyer there? If not, the choice is still a preference.
3. Run the diagnostic from the top of this page one more time. If all six questions now have clear written answers, Step 5 is done.

What you’ve built

After completing the steps above, the Step 5 section of outputs.md should contain:

This is not a deliverable for anyone else. It is a constraint on Step 6.

Assumption sweep

Before moving on, scan the Step 5 section of outputs.md for any field you filled in from reasoning rather than evidence. Common ones at Step 5:

• The audience presence map (did Step 2 interviews or a test confirm the segment is in these venues, or did you assume it?)
• The chosen channel’s reach (did a test show the segment is reachable there at scale, or is it your best guess?)
• The buying intent of the channel (did people respond as buyers, or just engage idly?)
• The format and awareness match (did a real reader respond to the form you put in the channel, or did you reason it should fit?)
• Which role the channel reaches (did you confirm the economic buyer is or is not there, or assume the buying group’s whereabouts?)
• The demand-creation vs capture split (is the share of the segment in-market at any time an estimate, and is the brand-vs-activation split a real choice?)
• The ICE/RICE confidence scores (did you rank on evidence, or on how much you like each experiment?)

Each unconfirmed field is an assumption. Log it in assumptions.md now if you have not already. The chosen channel’s reach and intent is usually the highest-impact assumption in the step; if no test has confirmed it, mark it a leap-of-faith and test it before Step 6 scales.

What this step hands off to Step 6

Before moving on, confirm the Step 5 section of outputs.md is complete. Step 6 opens by reading it. Specifically, Step 6 needs:

• The chosen beachhead channel(s), because lead generation is the system you build inside a channel. Step 6 cannot design a contact system without knowing which channel it lives in.
• The message form per channel, because the lead-generation mechanics (the post, the ad, the outreach, the offer) are built around the Step 4 message the channel carries.
• The reader and awareness level per channel, because the call to action and the offer Step 6 designs have to match how aware and how high-intent the reader in that channel is. A cold channel needs a lower-commitment ask than a high-intent one.
• The baseline response from the Step 5.9 test, because Step 6 is numbers-driven and needs a starting point to improve on. The test result is the first data point in the Step 6 funnel.
• The owned-audience path, because Step 6 should be building toward an audience you control, not only renting reach.
• The demand-creation vs capture split, because Step 6 designs a different offer for a capture channel (meets existing intent, ask for the sale) than for a creation channel (builds memory, ask for a smaller step). The offer follows the channel’s job.
• The launch plan and tier, when Step 6 is being built around a specific release, so the lead-generation push is sized to the launch rather than run flat.

The sharpness of your Step 5 channel choice sets the ceiling on what Step 6 can do. Step 6 is the step where the error messages are loudest, which is exactly why people start tinkering there first while the real fault sits in Step 5 or above. If Step 6’s numbers are bad, check here before optimising tactics: a perfectly tuned lead-generation system in a channel the buyer is not in will still produce nothing. And recall that the channel choice itself came from Step 2; if the channel is wrong, the cause may be an audience that was never located precisely enough. Come back and confirm where the buyer is.

Common failure modes

The channel was chosen for trendiness, not for where the buyer is. Someone said “we should be on TikTok” or “everyone is doing webinars,” and the audience question came second. The channel follows from Step 2, where the buyer already is, never from where the attention is loudest. If you cannot trace the channel back to Step 2’s “where they look,” it is the wrong reason.

You spread across many channels and mastered none. Presence everywhere, traction nowhere. A thin profile on five platforms loses to one channel dominated. Pick one or two, defer the rest in writing, and resist the pull to be everywhere.

The channel cannot carry the message. You chose a channel whose native format only holds a one-liner and then needed it to deliver long-form proof, or the reverse. Match the channel’s format to a Step 4 message form, or the channel underdelivers no matter how well you run it.

You built on rented land with no owned-audience plan. The whole strategy sits on a platform you do not control, which can change its rules, reach, or pricing overnight. Rented channels are fine to start; a plan that never converts any of that reach into an owned audience is fragile by design.

You confused where you like to create with where they consume. You picked the channel you enjoy or are good at, and assumed the buyer is there. Your comfort is not a targeting criterion. Go where the segment is, even if it is a channel you have to learn.

You scaled before validating. You jumped from choosing a channel straight into building the Step 6 machine, with no cheap test that the segment responds there. Validate cheap, then scale; a tuned funnel in an unproven channel is effort spent in the wrong place.

The channel has reach but no intent. It delivers large numbers of the wrong people, or the right people in an idle frame, and the funnel fills with traffic that never acts. Weight buying intent, not just reach; attention is not the same as looking.

“We are everywhere.” Said as a strength, it usually means no channel is owned and the effort is spread too thin to dominate any of them. If you cannot name the one or two channels you are committing to, the choice has not been made.

Sources

• Traction by Gabriel Weinberg and Justin Mares. The core source for Step 5. The Bullseye framework (brainstorm every channel, test the most promising cheaply, focus on the one that works) is behind Steps 3, 6, and 9, as is the discipline of treating channels as testable rather than assumed.
• This Is Marketing by Seth Godin. On going to where the audience already is rather than demanding their attention, the smallest viable market applied to channels, and the difference between rented and owned attention. Behind Steps 2 and 6 and the owned-audience path.
• Crossing the Chasm by Geoffrey A. Moore. On choosing a distribution channel that fits the beachhead segment and dominating one point of entry rather than spreading across many. Reinforces the beachhead-channel discipline in Step 5.6.
• Hacking Growth by Sean Ellis and Morgan Brown. On testing channels with small, fast, cheap experiments before committing budget, and reading channel signal from real response. Behind the minimal-test method in Step 5.9 and the ICE prioritisation in Step 5.7.
• The 1-Page Marketing Plan by Allan Dib. On selecting media and channels directly from the target market rather than from habit, and matching the channel to the message and the stage of the buyer. Practical background for Steps 1 and 5.
• The Long and the Short of It by Les Binet and Peter Field. The source for the balance between brand building and activation (roughly sixty to forty over time) and for share of voice ahead of share of market as a predictor of growth. Behind Step 5.4.
• How Brands Grow by Byron Sharp. The evidence that most of the category is out of the market on any given day, so reaching the future buyer matters as much as the one buying now. Behind the demand-creation half of Step 5.4.
• Loved by Martina Lauchengco. On tiering launches (T1, T2, T3) to the size of the news and building only the launch materials the tier warrants. Behind Step 5.8.