Playbook
The Content to Pipeline Loop
Content becomes pipeline when a specific person reads a specific piece and then talks to you. The loop has four parts. Pick the question a buyer asks just before they buy, answer it properly, put it in front of named people by hand, then follow every reply into a conversation.
Last reviewed 27 August 2026
- Outcome
- A weekly publishing loop measured in conversations started rather than sessions, with a distribution step that actually runs
- Timeframe
- One piece a week, reviewed quarterly
The steps
Run it in this order.
01
Build the question list from calls, not from a keyword tool
Go through your last twenty sales calls and support threads and write down every question a buyer asked in their own words. Keyword tools tell you what has volume, which at your stage is mostly questions asked by people who will never buy. Your call recordings tell you what the people who might buy are actually stuck on.
02
Rank each question by how close it sits to a decision
Score every question from one to five on how near it is to someone spending money. How do I compare X and Y is a five. What is B2B marketing is a one. Write the fives first, even though they have less search volume, because a hundred readers at decision stage are worth more than ten thousand at curiosity stage.
03
Write one piece that answers one question completely
Direct answer in the first two sentences, before any context or preamble. Then the specifics, the caveats, and the cases where your product is the wrong choice. A piece that fully closes one question gets forwarded internally by the person evaluating you, and internal forwarding is the mechanism that actually moves a deal.
04
Budget your hours in thirds and protect the last third
Split the time for each piece roughly evenly between research, writing, and distribution. Publishing is about a third of the job, not the whole job. Most content programmes spend everything on the middle third, ship into silence, and conclude that content does not work for them. Put the distribution hours in the calendar before you start writing.
05
Send it by hand to twenty named people
Open the piece, list twenty specific people it is genuinely relevant to, and message each of them individually saying why you thought of them. Customers, prospects who raised that exact question, people who posted about it publicly. This is an hour of unglamorous work per piece and it is where the conversations come from.
06
Post it where the question was originally asked
Find the forum threads, community questions, and comment sections where people asked this, and answer them properly in place, with the link as supporting material rather than as the answer itself. Follow each community rule on self-promotion and disclose your affiliation. A link dropped without an answer attached gets removed and earns you nothing.
07
Reply the same day and ask one question back
Every reply, comment, and forward gets a response within the day, and every response ends with a question about their situation. This is the step that converts a reader into a conversation. Without it you have an audience rather than a pipeline, and the two feel similar in a dashboard while behaving completely differently in a forecast.
08
Measure conversations started, not sessions
Keep one sheet with a row per conversation and a column for the piece that preceded it, filled in by asking people directly how they found you. Traffic tells you a page was loaded. Only the conversation count tells you whether the loop is working, and it is usually a much smaller and much more useful number.
09
Refresh what produces conversations, retire what does not
Each quarter, sort your pieces by conversations produced. Update and redistribute the top few, because a piece that worked once usually works again with fresh examples. Delete or consolidate the ones that produced nothing in six months. A smaller set of pieces that each produce conversations beats a large archive that produces sessions.
Traffic and pipeline are different things
A content programme can be working perfectly by every traffic metric and producing no revenue at all. This is not a measurement problem. It is usually an accurate reading of a programme aimed at the wrong readers.
Pipeline comes from a specific person, with a budget and a problem, reading something that resolves the thing blocking their decision, and then talking to you. Everything in the loop below exists to make that sequence more likely. Volume is not one of the ingredients.
The practical consequence is that you should be willing to publish something with a hundred readers if they are the right hundred, and that you should be suspicious of any piece that performs well and produces no replies.
Where the questions come from
From your call recordings, not from a keyword tool.
Keyword research tells you what a large number of people are searching. At an early stage, most of that volume is people who will never buy anything from anyone, and writing for it produces an audience with no commercial relationship to you. Your last twenty sales calls contain a different list, in the buyer's own words, weighted towards the questions that appear right before someone spends money.
Score them by proximity to a decision.
| Stage | Example shape | Volume | Worth writing |
|---|---|---|---|
| Decision | How does X compare to Y for a team of five | Low | First |
| Evaluation | What does implementing X actually involve | Low to medium | Second |
| Problem aware | Why does our current process keep breaking | Medium | Third |
| Curious | What is category X | High | Last, or never |
The inverse relationship between volume and value at this stage is the whole reason keyword-led content programmes disappoint.
The step everybody skips
Publishing is roughly a third of the work. Research is a third and distribution is a third, and the third that gets cut is always the last one, because by the time a piece is written it feels finished.
It is not finished. A published page that nobody has been pointed to is a page waiting on search indexing, which takes months and may never deliver. The distribution hour is what makes the piece do something this week.
Concretely, that hour is twenty individual messages to named people who have a live reason to care, plus a real contribution to two or three places where the question already gets asked, plus one owned-channel post that works as an argument on its own rather than as a link with a caption. None of it is automated and none of it can be, because the value is precisely that a person chose to send it to another person.
Following the reply
The loop closes on the response, not the publication.
Every reply gets an answer the same day, and every answer ends with a question about their situation. This is not a sales technique, it is the only way a reader tells you enough for you to know whether there is anything here. Sometimes the answer is that there is not, and saying so quickly is what makes the next conversation with that person possible.
Keep one sheet. A row per conversation, a column for what preceded it, filled in by asking. Over a quarter, that sheet tells you which three pieces are worth updating and which fifteen are worth deleting, and it will disagree with your analytics almost every time.
What to expect
The hand-distributed half of this can produce conversations in the first week, because you are going to people directly. The compounding half, where search and AI assistants surface the pieces to people you never contacted, takes months and cannot be hurried.
Run both. The fast half is what makes it survivable to wait for the slow half, and most content programmes die in the gap between them because nobody planned for the gap to exist.
Playbook questions
How much traffic do I need for this to work?
What does the distribution step actually look like in practice?
Should I gate anything behind a form?
How long before content produces pipeline?
How do I know which piece caused a conversation?
Or have us run it.
This playbook is yours to use. If you would rather not run it yourself, that is the entire point of Zway.
30 minutes. You leave with the plan either way.