When to use
- You are planning a 6 to 12 month demand program and need a channel mix, not a list of one-off sends.
- A product launch has to land as a business moment, not just a release note.
- Pipeline is coming from one channel and you need to diversify before it saturates.
- A new distribution surface opened (a platform in its open phase) and you have a short window to ride it.
- The CFO is asking what marketing returned and the dashboard answers in MQLs.
How to use
The Campaign Operating System
01 Start with the business question, not the channel list.
Before you pick a channel or draw a calendar, write the decision this campaign has to move. A campaign without a business question is a content calendar. It is activity, not a plan.
State the one or two outcomes the campaign must produce in the next two quarters: pipeline in a named segment, adoption of a launched capability, preference in a category where you are unknown. Then check every workstream against those outcomes. If a webinar or a paid push does not connect to one of them, cut it. Diagnose before executing, refuse the playbook ask
Separate three things that teams blur together: a release ships a feature, a launch builds a business moment around it, and go-to-market is the commercial engine across all revenue teams. A campaign sits inside go-to-market. Most "campaigns" are really launches dressed up, which is why they spike and fade. Smart teams move to fewer, larger launches and always-on GTM.
Release ≠ Launch ≠ GTM
📦 Release: Development ships. Feature exists.
📣 Launch: PMM builds a business moment to drive adoption, revenue, and market perception based on PM objectives.
· Maddy Hirshan, LinkedIn, 2026-04-10
02 Split the funnel: demand creation for the 95%, demand capture for the 5%.
Most B2B campaigns over-index on demand capture, reaching the small slice of buyers actively searching, and starve demand creation, building preference with the majority who are not yet in market. The two motions need different channels, different KPIs, and different measurement. Stop optimizing for the 5% in-market; build preference with the 95% who aren't
The reason teams default to capture is that attribution software rewards what it can track. Trackable channels look efficient on a last-click dashboard. The channels where preference actually forms (LinkedIn, podcasts, communities) show zero in that view, so they get cut first. By the time a buyer raises a hand, the brand decision is mostly made. Capture only harvests demand that creation built.
"Passetto attributed 97% of their revenue to dark social using customer-reported attribution, while their software-based attribution showed 0% from those channels."
· Chris Walker, State of Demand Gen, 2026-03-03 · “Stop optimizing for the…”
Decide the split before you allocate budget. Tag every workstream as creation or capture. If everything is capture, you are buying pipeline that someone else's brand work created, and you will keep paying for it.
03 Tier your leads as a three-asset portfolio, not one blended funnel.
Leads are not interchangeable. They arrive in three asset classes, each with different physics. Treating them as one number hides which engine is broken. Treat leads as a three-asset portfolio, not a single funnel
| Lead type | Source | Physics | Owns the quarter? |
|---|---|---|---|
| Seeds | Referral, organic, word of mouth | Compound 2 to 5x higher conversion, but slow. Cannot be bought. | No. Build over months. |
| Nets | Inbound marketing, content, paid | Scale with budget. Hit a market-wide demand ceiling. | Partly. |
| Spears | Outbound, SDR-driven | Lowest conversion, but output is proportional to input. The only predictable lever. | Yes. The lever you pull on a quarterly horizon. |
A mature program runs at least two of the three. Single-source dependency is an existential risk.
"No mature B2B company should rely on fewer than two types. Single-source dependency creates existential pipeline risk."
· Aaron Ross, Seeds, Nets, and Spears, predictablerevenue.com · “Treat leads as a…”
Map your demand-creation work (step 2) onto Seeds and Nets, your demand-capture work onto Nets and Spears. The portfolio frame is how you decide where the next dollar goes: if pipeline is all Spears, you have no compounding asset; if it is all Seeds, you cannot forecast the quarter.
04 Audit channels by maturity, then earn before you rent.
Map every channel you run against a maturity stage: Up Next, Testing, Scaling, Core, Big Bet, Not Working. This is Emily Kramer's MKT1 channel model. At any point at least one channel should be a Big Bet that could change the growth trajectory, not just defend it.
Then sort the mix into rented and earned. Algorithm channels (paid search, paid social, SEO) are rented attention. The cost trends up forever and the rent goes to the platform. Earned channels (sharing loops, referral mechanics, user-generated content, an audience you own) accrue to you. With AI eating the search interface, rented costs rise faster. Build earned channels, every dollar in algorithm channels makes Google richer, not you Distribution and earned channels are the new moat
"When you're doing organic search or paid search, you're making Google richer. With AI eating the search UI, your cost of acquisition is only going to go up. You're constantly going to be praying to algorithm gods."
· Elena Verna, Lenny's Podcast, 2026-04-28 · “Build earned channels, every…”
If a genuinely new distribution surface is in its open phase, that is the Big Bet. Every platform moves through a four-step cycle: category competition, moat identification, open platform with free distribution, then lock-down with monetization. Early movers in the open phase get free escape velocity. Late entrants pay rented prices against incumbents who already own the audience. The cycles are getting shorter, so the window is smaller than it feels. Every distribution platform follows a four-step cycle; cycles are getting shorter
"If you don't do it, your competitors are going to go to the new platform and your customer expectations change. There is no opting out of the game."
· Brian Balfour, Lenny's Podcast, 2026-04-28 · “Every distribution platform follows…”
One caution on the Big Bet. Build the audience before you launch loud on it, but do not signal a product that does not exist yet. The failure mode runs both ways: shipping with no audience, or making noise with no product. Build quietly vs. distribution-as-moat
05 Build the calendar around fewer, larger launches tied to the roadmap.
Map a 6 to 12 month calendar organized by campaign type, audience, and product theme, aligned to the product roadmap. The aligning is the work. A calendar disconnected from what engineering ships becomes a list of invented moments.
Move toward fewer, larger launches. Package multiple releases into one narrative on a quarterly or biannual rhythm rather than a constant treadmill of small announcements. Bundling produces stronger signal and more momentum than a stream of individually forgettable sends. Smart teams move to fewer, larger launches and always-on GTM.
Underneath the launch moments, run always-on go-to-market: lifecycle communication, value-based education, routine enablement. The launches are the spikes. The always-on layer is what compounds adoption between them. A campaign calendar that is all spikes trains buyers to wait for the next announcement instead of adopting what is in front of them.
06 Write a messaging brief per campaign before any asset starts.
No asset begins until its brief exists. Each brief carries the persona, the pain or situation, the core message, the positioning frame, the proof, and the call to action. The brief is the contract between PMM and content. Without it, content produces polished assets that say the wrong thing well.
Order the work the way buyers actually decide. For the demand-creation campaigns, the message leads with the cost of the status quo, not a feature comparison. When you survey buyers who did not buy, 40 to 60% made no decision at all. They were not choosing a competitor. They were not convinced the problem was worth solving. 40–60% of B2B buyers say "no decision", your real competitor is the status quo
"When you survey buyers who didn't buy, 40–60% say they made no purchase decision at all."
· April Dunford, Lenny's Podcast, 2026-04-28 · “40–60% of B2B buyers…”
This sets the call to action. A demand-creation campaign earns the next step by making inaction feel expensive, not by demanding a demo. Make the CTA a product-specific verb the buyer can picture themselves doing, not a generic "Learn more."
07 Source content from the frontline, not from keyword tools.
The highest-converting campaign content comes from the live language in sales calls, customer-success threads, and support tickets, not from keyword-volume tools that lag reality by 6 to 12 months. Source content briefs from Sales, Success, and Support, not keyword tools Frontline customer contact is the PMM substrate
"SEO tools are at best 6 to 12 months behind reality."
· Brendan Hufford, 3S Strategy, 2026-04-25 · “Source content briefs from…”
Cluster the three streams weekly. Rank candidate briefs by frequency times deal size. Use the frontline source for bottom-of-funnel and comparison content where the demand is already verified by pipeline. Use category-level research for the top-of-funnel demand-creation work the sales call cannot yet see.
Hold the content to a higher bar than "professional." Ask the diagnostic: whose content are you actually looking forward to reading, and why? If the honest answer is not "ours," you are producing content no one outside the company anticipates, however polished it is. Distribution is more important than creation, but only when the thing is worth distributing. Distribution is more important than creation, the diagnostic question is "whose content are you actually looking forward to reading and why?"
08 Run webinars and events for customers first, prospects second.
Run webinars for customers first. Repurpose for prospects roughly four weeks later. Customer feedback validates the messaging before it meets a cold audience. The four-week gap is the test.
Define segment plays before the campaign launches, not after the event:
| Segment | Who | Play |
|---|---|---|
| Expansion-Ready | Existing customer, high engagement | CSM expansion motion |
| Adoption-Boost | Existing customer, low engagement | Adoption motion |
| Awareness-Only | Prospect, low intent | Nurture sequence |
| No-Show | Registered, did not attend | Recording plus a one-question survey |
For live events and demos (interactive demos, workshops, live builds), record a dry run at least 48 hours before going live. The reason events fail is rarely the content. It is the unrehearsed run.
09 Design launches to meet the buyer at the right moment, not in the inbox.
A feature-launch email lands at 10:47am into a buyer who has already handled a crisis, 47 unread emails, and a hard 1:1. The launch never stood a chance. The scarce resource is the buyer's attention, and a changelog email spends none of it well. Feature launches fail because buyers have no attention left.
The fix isn't better copy.
It's understanding that attention is the scarcest resource your buyer has.
· James Doman, LinkedIn, 2026-04-10
Design the launch to meet the customer at the right moment, in the right place, with the right message. In-app at the point of relevant use beats a broadcast email. This is why step 5 favors fewer, larger moments: a bundled narrative earns a place in the buyer's attention; a stream of small sends trains them to ignore you.
10 Build the follow-up and nurture as the part that actually closes.
A campaign that ends at the event is half a campaign. Segment the post-event motion into hot, warm, and no-show with distinct sequences, defined before launch. Hot gets a fast human follow-up. Warm gets a nurture track. No-show gets the recording and a single question that re-qualifies intent.
Hand the follow-up to the team that owns it. Demand gen owns the calendar, sales-led growth owns follow-up, PMM owns messaging, content owns assets. Naming the owner per stage is what stops the generic "thanks for attending" send that ignores every segment you just defined.
11 Treat the new earned channel as a 12-month bet, protected from KPIs.
Every loop decays. Allocate roughly 20 to 25% of the team's annual time to introducing a new growth model, and protect it from metric expectations for the first 12 to 18 months. Without a runway, the new loop dies at month six before it can compound. Add a new growth model every 18 months and protect it from KPIs for 12
"Most loops decay over 5–7 years; some, like Dropbox sharing, don't. Allocate 20–25% of the growth team's annual time to introducing new loops with no metric expectation for the first 12–18 months."
· Elena Verna, Lenny's Podcast, 2026-04-28 · “Add a new growth…”
This is the discipline behind the Big Bet in step 4. The campaign calendar carries the channels that fund the quarter. The protected bet is the channel you are growing for the year after. Name the portfolio: currently funding, currently seeding, retired. Defend the seeding channel in reviews without an ROI justification, or it will not survive the first cost-cutting cycle. This rule is a luxury for distressed companies. Fund the new model only when the existing ones are healthy enough to pay for the experiment.
12 Measure with absolute counts and short correlated signals, in the CFO's language.
Track from registration through revenue influence, but pick the right instruments. Optimize absolute counts at each stage, not stage conversion rates. For outcomes months out, find a correlated short signal and validate it, rather than waiting on a long loop. If an experiment cannot reach sample size in 30 days, run pre and post instead of pretending the test is clean. Absolute counts + correlated short signals, not stage rates and long loops
Report the result in the language of finance, not the language of marketing. A dashboard of MQLs, sessions, and engagement scores is unreadable to a CFO. Pipeline translated to revenue contribution, payback period, and burn multiple is readable, and it is what protects the budget when cycles tighten. Marketing must speak the language of finance, not the language of marketing
"Marketing must speak the language of finance, not the language of marketing; the CMO who can translate pipeline metrics into financial outcomes earns strategic credibility, budget protection, and a seat at the revenue table."
· Rowan Tonkin, synergy between finance and marketing, 2026-03-03 · “Marketing must speak the…”
Audit one quarter of "wins" against a one-year holdout and expect a chunk to evaporate. The campaign that looked like a win on a 7-day signal often does not survive the year. Absolute counts + correlated short signals, not stage rates and long loops
Order the work right
The five-step marketing process names the sequence campaigns most often break: invent something worth making, design it for the few who benefit most, tell the story that matches that group's worldview, spread the word, then show up for years. Steps one through three happen before any visible activity. Most teams start at step four, "how do we promote this," which is why the promotion does not land. Five steps in order: invent, design for the few, tell the matching story, spread, show up for years
"In 'This Is Marketing,' Godin codified a five-step marketing process that operationalizes these ideas: (1) invent something worth making with a story worth telling, (2) design it so a few people will particularly benefit, (3) tell a story that matches the built-in narrative of that tiny group (the smallest viable market), (4) spread the word, (5) show up regularly for years to earn permission and enrollment."
· Seth Godin, This Is Marketing, 2018-11-13 · “Five steps in order:…”
The campaign maps onto the back half. Steps 1 to 3 are positioning and offer work you do upstream. The calendar, channels, content, and nurture are step 4. The always-on layer is step 5. A campaign that skips the upstream work spends step-4 money against a step-1 problem. Market and offer beat funnel optimisation
Check your work
- Every campaign traces to a named business outcome for the next two quarters, not just a date on a calendar.
- The funnel is split into demand creation and demand capture, with different KPIs for each.
- Lead sources are tracked as Seeds, Nets, and Spears, not one blended pipeline number.
- At least two of the three lead types are running. No single-source dependency.
- Channels are tagged rented vs. earned, and at least one channel is a designated Big Bet.
- Every campaign has a messaging brief before any asset starts.
- Demand-creation messages lead with the cost of inaction, not a feature comparison.
- Content briefs cite frontline language (call quotes, ticket counts), not just keyword volume.
- Customer webinars run before prospect webinars. Every live event has a dry run recorded 48 hours prior.
- Follow-up is segmented hot / warm / no-show with distinct sequences and a named owner per stage.
- A new growth model is funded and protected from KPIs for 12 months.
- Metrics include downstream revenue influence and are reported in CFO language, not MQLs.
- CTAs are product-specific verbs, not generic "Learn more."
What goes wrong
- Random acts of marketing. Channel selection by competitor mimicry instead of audience research and motion fit. Map channels by maturity and pick the Big Bet on purpose. Distribution and earned channels are the new moat
- All capture, no creation. Buying the 5% who are already shopping while the 95% form preference somewhere you do not show up. Split the funnel and fund creation. Stop optimizing for the 5% in-market; build preference with the 95% who aren't
- Single-source pipeline. Running on Nets alone, or Spears alone, until the one engine stalls. Run at least two lead types. Treat leads as a three-asset portfolio, not a single funnel
- All rented channels. Every dollar in algorithm channels makes Google richer, not you. Build at least one earned loop. Build earned channels, every dollar in algorithm channels makes Google richer, not you
- Assets before the brief. Production starts, then someone asks what the message is. The brief precedes the asset, always.
- Launch treadmill. A stream of small sends that trains buyers to ignore you. Move to fewer, larger launches with an always-on layer underneath. Smart teams move to fewer, larger launches and always-on GTM.
- Inbox-only launches. A changelog email into a buyer with no attention left. Meet them at the moment of use. Feature launches fail because buyers have no attention left.
- Generic follow-up. "Thanks for attending" to every segment alike. Segment hot / warm / no-show before the event.
- Calendar disconnected from roadmap. Invented moments with nothing shipping behind them. Align the calendar to what product ships.
- Killing the new loop at month six. Judging a 12-month bet on a one-quarter horizon. Protect the seeding channel from KPIs. Add a new growth model every 18 months and protect it from KPIs for 12
- MQL dashboards. Reporting activity the CFO cannot read, then losing the budget in the next tightening cycle. Report in revenue terms. Marketing must speak the language of finance, not the language of marketing
- Promoting before the upstream work. Starting at "how do we spread this" before inventing something worth spreading. Do steps 1 to 3 first. Five steps in order: invent, design for the few, tell the matching story, spread, show up for years
What you get
- Channel audit with maturity staging and a rented vs. earned tag per channel.
- Demand split: creation vs. capture, with separate KPIs and budget allocation.
- Lead portfolio: Seeds / Nets / Spears with distinct conversion math and named owners.
- 6 to 12 month campaign calendar, aligned to roadmap, built around fewer larger launches.
- Messaging brief per campaign: persona, pain, message, positioning, proof, CTA.
- Webinar playbooks (customer and prospect) with the four-segment attendee model.
- Event and demo strategy with a recorded dry run gate.
- Follow-up and nurture system, segmented hot / warm / no-show with stage owners.
- A protected Big Bet channel, funded for 12 months without a metric gate.
- Measurement dashboard with absolute counts, short correlated signals, and a revenue-language attribution model.