One channel is no longer a strategy
The playbook that said pick a channel and go deep was written for a search engine that judged you on your own pages. Three things changed at once, and the minimum viable channel mix went from one to three.
For about fifteen years the standard advice was to pick one channel and go deep. Master search, or master paid, or master outbound. Diversifying early was the mark of a team that could not commit.
That advice was correct, and several things underneath it moved.
Three, not one
The current floor is three channels: one social channel at the top of the funnel that genuinely matches where your buyers are, search if — and only if — you are actually excited by it, and an email list from day one.
The reason for the third is the interesting one, and it is not the obvious reason. Search engines increasingly weight whether you exist anywhere else at all: whether you are a recognised entity, whether people talk about you off-search, whether any of your traffic arrives from somewhere other than a query.
Which inverts the old trade-off. Under the previous model, effort spent on social was effort not spent on search. Under the current one, existing off-search makes search easier, because it is part of what search is measuring. The channels stopped competing for the same hours and started feeding each other.
Discovery platforms and relationship platforms
The most useful way to sort your options is not paid against organic. It is by who controls the connection.
Discovery platforms have an algorithm that connects strangers to content: Instagram, TikTok, X, LinkedIn, YouTube. They can hand you an audience you did not have. They can also stop, without notice, without explanation, and without any change on your part.
Relationship platforms are the ones where the connection is yours: email, podcasts, SMS, private communities. Slow to build. Nobody can switch them off.
And the operating rule that follows, which is the single most useful sentence in this area:
The game is using discovery platforms to reduce your dependency on discovery platforms.
Jay Clouse
Discovery is not the destination; it is the on-ramp. Every piece of work on a discovery platform should have an explicit job of moving someone onto a relationship platform. If a month of posting produced reach and no addresses, it produced nothing durable — you rented attention and gave it back.
The warning that goes with it dates the old playbook precisely:
Followers have never meant less. You're basically auditioning for even your own audience with everything you publish.
Jay Clouse
Follower counts stopped being a durable asset when the feeds stopped being chronological. Your list is an asset. Your following is a lease.
What is happening to cold outreach
Cold email and cold LinkedIn are saturating, and the next phase is worse: the volume is about to rise sharply as sending gets automated end to end. The recommended alternative in that world is physical — handwritten letters, gift boxes, video mailers — and it comes with a condition that most people repeating the advice leave out.
The physical touch has to be paired with a trust layer waiting for the search the recipient will immediately run. Someone who receives an unexpected package looks you up within the minute. If what they find is a thin site, no third-party mentions, and nothing that says other people take you seriously, the package has converted mild interest into a decision not to reply.
Two honest caveats. Physical mail at any scale is expensive per touch, so it only works on a short, high-value, well-chosen list. And the evidence for it in software is thinner than the enthusiasm — it is largely second-hand, which makes it a reasonable experiment on twenty named accounts and a poor basis for a plan.
The broader point stands regardless of whether you post anybody a box: as the cost of sending a message falls to zero, the scarce thing is not the message. It is having something a suspicious recipient can check.
The reporting fix that costs nothing
Zero-click search and answers written by language models have broken last-touch attribution in a specific direction. Someone reads about you in a newsletter, hears your name on a podcast, then searches your brand and arrives. Search takes the credit. The newsletter and the podcast look like they did nothing, and next quarter you cut them.
The fix is one optional free-text field at signup — how did you hear about us? — and it moves a typical business from roughly 55% attributed to 75–80%. It costs one input and it changes which channels survive the budget conversation.
It is also the only instrument that can see the channels that never get a click.
Budget one channel a year
The last constraint is the one that makes the other advice tractable. Every channel skill takes about a year to learn properly — one operator's sequence was content and search, then paid search, then display, then copywriting, then audience, one per year, while holding down a full-time job.
Search is slower still: five years to anything genuinely interesting. If you go there anyway, take the cheap structural wins early — put the blog in a subfolder rather than a subdomain (worth a 20–30% lift on its own), make each URL slug exactly the keyword, and rank indirectly for "best {category} software" by getting reviews onto the directories that already rank for it.
So "three channels" is not "start three this quarter". It is: one channel learned properly per year, and an email list from day one, which is the one that does not take a year.
Five things to do
- Count your channels. One is fragile, whatever it is currently producing.
- Start the list today if you have not. It is the relationship platform with the lowest cost and the longest half-life.
- Give every discovery-platform activity an explicit job: which relationship platform does it move someone to?
- Add the attribution question. One field. This week.
- Pick the one channel you will learn this year, and stop grading the others as though they were also being run properly.
Keep reading
- Treat the first call as a research interview you also happen to be selling on You will lose most first meetings. The ones you lose are worth something only if you left with an answer you did not have before — which requires deciding, before you dial, what you are trying to learn.
- How long to give an idea before you are allowed to conclude anything A week is not enough and six months is too long, and the number matters less than the thing it protects you from: quietly retreating into the work that feels productive while the question you were testing goes unanswered.
- Your price decides which sales motion you are allowed to run Whether you can afford demo calls, an account-based motion, or only a self-serve funnel is not a strategy question. It is arithmetic, and the thresholds are specific enough to check yourself against this afternoon.