The Anti-Playbook, Part 2: The 90-Day Roadmap, Inverted
This is Part 2 of The Anti-Playbook, a series on how I’d actually build demand gen at an early-stage B2B SaaS company in 2026. Part 1 covered the setup and why the standard playbook fails. If you missed it, start here.
Most 90-day plans for demand gen leaders look like a pyramid: research at the base, content in the middle, demand generation at the top. The logic is that you can’t generate demand without content, and you can’t build content without research. Each phase depends on the one below it.
It’s a tidy framework. It’s also wrong.
The problem with the pyramid is that it takes 60 days to reach the part where pipeline actually gets generated. That’s 60 days of your salary. 60 days of CEO patience. 60 days of the sales team watching marketing produce decks instead of meetings.
At an early-stage B2B company, you don’t have 60 day.
So we flip it.
The Inversion
The inverted roadmap starts pipeline generation in week one. Research and content are build around the conversations that outbound produces, not in advance of them.
This sounds reckless. It isn’t. It’s a different theory of how learning actually happens.
The pyramid is built on a fear: that if you start outbound without research, you’ll be wrong about the ICP, the messaging, and the offer. So you spend 30 days reducing that risk.
The inversion accepts that you’ll be wrong — and treats being wrong fast as the goal. Two weeks of being wrong with real buyers teaches you more than 30 days of being theoretically correct in a conference room.
The pyramid produces a polished plan that may or may not work. The inversion produces a working motion that may or may not be polished.
At an early-stage company, working beats polished every time.
Here’s what the three phases actually look like.
Days 1-14: Generate Conversations Before You Have Infrastructure
The goal of the first two weeks is one thing: put real messages in front of real buyers and watch what comes back.
This is not the listening tour. There is no listening tour. The listening tour is the consultant’s defense mechanism — it’s what you do when you don’t know what to do, and it produces the illusion of progress without producing pipeline. Skip it.
What you’re doing instead:
Pick one narrow ICP slice and one offer. Not the full ICP. One slice. If the company sells to “B2B SaaS Marketing teams,” you pick “Series B SaaS companies, 50-200 employees, with a marketing leader hired in the last 6 months.” Narrow enough that one message can resonate with all of them. The offer is whatever sales is already closing — don’t invent a new one.
Build a list of 200 named accounts. Enrich with the data points that actually matter for personalization — recent funding, recent hires, tech stack, content they’ve published. Don’t buy a list of 10,000 contacts. The cost of bad data is higher than the cost of a small list.
Write the first sequence by hand. Three messages, sent from real inboxes, by real people. Marketing leader sends some. SDR or AE sends some. Twenty messages a day, manually, while the scaled infrastructure gets built in parallel.
By day 14, you’ve sent around 140 messages, gotten 5-15 replies, booked 2-5 meetings, and started building a documented list of objection patterns and resonant messaging themes.
You’ve also done something most demand gen hires never do in their first two weeks: you’ve put pipeline on the board.
Days 15-45: Layer in the Scaled Motion
Manual outbound proves the concept. It does not scale. Sending 20 messages a day will not produce the pipeline a Series B company needs to grow.
This is where the AI-powered outbound stack comes in — and where most frameworks wave their hands and move on. The mechanics matter and Part 3 of this series covers them in detail. For now, the shape of this phase looks like:
The infrastructure built in parallel during Days 1-14 — secondary domains warmed, deliverability monitoring in place, sequencing platform configured — comes online. Volume scales from 20 sends a day to 200-400. The messaging that worked in the manual phase becomes the spine of the AI-generated variations.
At the same time, the first content moves get made — not based on a content calendar built from a keyword research deck, but based on what you’ve actually heard from prospects. The single best resource you could send to the next 200 people becomes a real piece of content.
By day 45, the outbound motion is generating consistent pipeline, the sales team trusts that marketing is producing real opportunities, and you have data to defend every decision you’ve made.
This is the moment the role becomes defensible.
Days 46-90: Compound and Layer
Only now — after pipeline is flowing and the systems are in place — do you start adding the second and third motions.
The content engine that was hand-fed in Phase 2 becomes a real publishing motion. The marketing leader and the sales team start posting in their own voices, with the credibility that comes from actually talking to the market every day. This is content with a feedback loop, not content built in a vacuum.
The inbound capture infrastructure gets built. Now that demand is being created externally, the website needs to convert when a prospect lands on it. Forms, demo flows, pricing transparency, lifecycle automation. This is the work the consultant playbook would have you do in weeks 1-4. It’s correctly sequences for week 6.
Paid layers in as an amplifier, not a generator. Once you now which messages and offers convert organically, paid becomes a way to scale what’s working — not a way to test into the unknown. Most early-stage companies that start with paid in week one waste money. You’re not testing. You’re scaling proven offers.
A second outbound motion opens. With one ICP slice working, you expand to the adjacent slice. Repeat the manual-then-scaled progression. The framework is repeatable.
By day 90, the company has a demand engine that produces predictable pipeline, a content engine fed by real market conversations, and a sales team that actively partners with marketing instead of merely tolerating it.
The CEO has stopped asking what marketing is doing.
Why This Sequencing Works
Every step of the inversion is built on the same principle: learning velocity beats production volume.
You can spend 30 days building a flawless plan for a market you don’t yet understand. Or you can spend 30 days getting punched in the face by that market and emerging with calibrated instincts. The second path is faster, cheaper, and produces better decisions.
The pyramid optimizes for defensibility — the ability to point to a process when things go wrong. The inversion optimizes for results — the ability to point to pipeline when things go right.
At an early-stage company, you don’t get to choose both. You’re going to be evaluated on one or the other. Pick the one that actually moves the company forward.
What’s Next
The inversion is the strategic argument. The next few posts get into the mechanics:
Part 3: The AI-Powered Outbound Stack — the actual tools, workflows, and operational details that make the scaled phase work
Part 4: Kill the MQL — why the lead scoring model the pyramid was built on is the wrong measurement system for this motion
Part 5: The Channel Bet — why outbound first isn’t just a tactical choice, it’s a bet on a specific theory of how early-stage demand gen works
If the inversion makes sense to you, the rest of the series is the operational manual. If it doesn’t — that’s worth knowing now, because the framework either holds together or it doesn’t.
Tell me where I’m wrong.
