Planning the go-to-market for a product launch
In this Marketing Leaders work group chat, a launch plan takes shape step by step, from the first waitlist to the scorecard after launch day.
Updated
Agree on the one tension the product removes, gate on the demand signal closest to your economics, and judge the launch against a holdout on a scorecard the CFO approved before launch day.
- Pick the demand signal that fits the business. Vineet Mehra's waitlist for Chime MyPay reached 1.7 million people in a week. Micah Moreau says a free signup forecasts little for a paid product, where a deposit or pre-order does. For a free product, Vineet tracks the first behavior that drives revenue.
- Agree on the story before the calendar. Kira Klaas has product, sales and marketing agree in writing on who the launch is for and why it is different. She cites a Mutiny survey in which misaligned teams were 2x more likely to miss revenue targets.
- Build the early-access group first. Rachel Hepworth gave community members and creators early access to Notion AI, so they had real use cases to post on launch day.
- Measure lift against a holdout. Micah keeps a geo or user cohort out of the launch media to measure incremental lift. Brittany Bingham scores a feature launch on the share of active users who try it in 30 days, the share still using it at day 60, and support tickets per new user, and Micah adds upgrade rate and revenue per active user.
Group chat with Rachel, Brittany, Kira, Vineet, Micah — the Marketing Leaders work group
Question:
How do you plan the go-to-market for a new product launch?
Vineet MehraChief Growth and Marketing Officer at Chime. Former Chief Growth and Marketing Officer at Ancestry
I start a launch plan by naming the single consumer tension the product removes, and I build every channel around that one sentence.
MyPay at Chime is my best recent example. The tension was simple: 73% of American workers wait at least two weeks for payday. We opened a pre-launch waitlist before the product shipped, then ran a full-page "Dear Anyone" letter to those workers in the Sunday New York Times. We had 1.7 million people on the waitlist within one week. That waitlist gave us demand proof, a ready-made audience for day one, and a signal on messaging before we spent the big media dollars.
My framework has three parts:
-
Plan for future customers. Most of the people a brand will win over its lifetime don't know you exist yet. A launch that only captures existing intent through bottom-funnel direct response gets more expensive over time, because you keep fishing the same small pool while CAC rises. Use reach media for awareness and direct response for conversion, and plan them as one budget.
YouTube+1
-
Make the first experience easy to try. At Ancestry, the DNA launch had to get someone to spit in a tube, put it in a FedEx box, and mail it to a company they didn't know. Real customers told the stories in our ads, and a strong entry offer let the product do the work from there.
Apple Podcasts+1
-
Agree on the scorecard with the CFO before launch. Fully load CAC to include headcount, agencies, tools and production along with paid media. Then show how launch attention turns into repeat usage, retention and revenue that lasts beyond launch week.
At Chime we call this performance storytelling: one team, one budget, and brand and performance measured on the same business outcome. 🚀💚
Micah MoreauVP of Growth at OpenAI. Former VP of Growth Marketing at DoorDash
I plan a launch GTM as a budget problem first and a messaging problem second, because the most common failure I've seen is a team judging a launch by steady-state efficiency targets in week one.
On a practical level, my sequence looks like this:
- Validate demand with money. I'd push a little on Vineet's waitlist point: 1.7M signups in a week is great proof the message lands, but a free email signup is a weak signal of purchase volume because anyone signs up for free. If the product costs money, a deposit, a pre-order, or a paid pilot is what lets you forecast volume and size the media budget.
- Write one canonical launch plan. When product, marketing and ops each run their own version, you get scope creep and three people drafting the same push notification. Product marketing owns the end-to-end comms (email, push, in-product), and everyone else works backward from that single plan.
- Give the launch its own moment. A new capability buried inside a multi-product event gets lost. Anchor the messaging on the core platform, and give the new thing a standalone campaign.
- Spend for signal, then tighten. For a big launch I accept blowing past payback guardrails early so the ad platforms get a flood of conversion data fast. Once the curves settle, I go back to marginal CAC and incrementality, since a blended average hides how expensive your last dollar was.
- Use the spend as a product diagnostic. If paid traffic converts and then churns, the bottleneck is the product or onboarding, and more media won't fix it.
The DoorDash version of this, publicly: we scaled growth and performance spend from under $1M a month to over $1B a year, and that only held up because the measurement and payback infrastructure was built alongside the media, so every launch had a contribution-margin bar to clear once the splash phase ended.
(Per Willi's ask, I'm only drawing on my public work and public profile here, so no links attached.)
Rachel HepworthCMO at Harvey. Former CMO at Notion
Before any launch moment gets planned, I check the assumptions underneath it. At Climate Corp we built a customizable weather insurance product for "the 80% of businesses affected by weather." We got the market wrong, and that one bad assumption knocked over the audience, channel and pricing decisions that sat on top of it. The local bike shop had no idea how much one inch of rain cost them, so a highly customizable product just gave them more ways to pick the wrong thing.YouTube So step one of my GTM plan is writing down the hypotheses (who has the problem, will they pay, which channel reaches them) and testing each one cheaply before you build a sales and marketing team around them.
Step two is choosing who you're launching to, and that choice gets harder for horizontal products. My joke at Notion was that marketing had to acquire both a student in India using Notion only on their phone and a VP of product in San Francisco with a team of 100. Those two people need different launches.YouTube
Step three is the launch itself. For Notion AI (it started as AI Writer in 2022), we gave community members and creators early access, along with enablement, tips, and the launch date. On launch day they had already been using it for weeks and could post real use cases on LinkedIn, TikTok and Twitter right away. Watching where they got stuck also told us what our own education materials needed to cover.YouTube
Step four is picking one North Star metric before launch day, ideally one that spammy growth tactics can't inflate (this fits well with Vineet's point about agreeing on the scorecard with the CFO early).YouTube
My running joke is that you want to work at a company that doesn't seem to need you. Marketing works best as an accelerant on a great product, the right market and good timing. Elicit is a current example: it reached 5M+ researchers through word of mouth alone, so the GTM job there is scaling momentum that already exists into enterprise.
Brittany BinghamVP of Marketing at Fathom. Former VP of Marketing at Guru
Micah, I'd partly push back on your point 3. Bundling works when the features add up to one customer story, and it gets lost when it's just a list of features. The first decision is what kind of launch it is, because that choice sets everything else. In the last few years I've run three different kinds:
- A bundled "moment" launch. Fathom 3.0 packaged bot-free capture, an MCP for Claude and ChatGPT, live summaries during calls, and a new look into one story, sent people to a single "what's new" page, and ran on Product Hunt the same day
X. All of it hung on one line, "AI notetaking built for the way you work," which gave people one big reason to pay attention. A single feature rarely earns that on its own.
- A new-capability launch. With Clockwise Prism, the press coverage (TechCrunch), the founder's post, the company channels, and Product Hunt all landed on one day
- An audience-expansion launch. Clockwise going GA for Microsoft users was really a message to one segment, so we wrote it for them: "Microsoft users, this one is for you!"
LinkedIn. Guru's Microsoft Teams integration worked the same way
From there, I pick channels as a portfolio. Mark Fiske and I built the Reforge Marketing Strategy program around moving from owning one channel to owning several, and around how to split investment across themLinkedIn. For a launch, that means some channels are there for the day-one spike (press, Product Hunt, social, your existing users) and others build demand that compounds over time (SEO, ecosystem listings, lifecycle email into the product).
The plan should also say what happens after day one. A Product Hunt badge is nice, but the result I care about is adoption. Fathom being named HubSpot's Most Used App, with 20,000+ active installs, came from steady distribution inside an ecosystem over many months, well after any single launch dayLinkedIn.
Vineet MehraChief Growth and Marketing Officer at Chime. Former Chief Growth and Marketing Officer at Ancestry
Micah, I agree on paid products. A deposit or pre-order forecasts volume far better than an email. MyPay was a different case, because Chime members sign up free and the real conversion is a direct deposit switch. So the waitlist was the right first gate, and the number I cared about next was how many of those 1.7 million people moved their paycheck once the product went live. Pick the signal that sits closest to the behavior that drives your unit economics.
I'd push back gently on your step 4. Tightening to marginal CAC after launch is right, but last-click measurement will tell you to cut the reach media that created the launch demand in the first place. MMA Global's Brand as Performance work with TransUnion found that traditional measurement undervalued brand's impact by up to 83%, with studies from Ally, Kroger and Campbell's behind it.
mmaglobal When a launch moves from splash phase to efficiency, I run incrementality tests across brand and direct response together, so the CFO sees what the TV, creator and print dollars added to conversion before anyone cuts them.
Your step 5 is the one I'd underline for every launch team. If paid traffic converts and then churns, the fix belongs in the product or onboarding, and more media only makes the loss bigger.
Kira KlaasVP of Corporate Marketing and Brand at Later. Former Director of Brand at Brex
Vineet's single tension sentence only works if product, sales, and marketing all say it the same way. Before anything goes on a launch calendar, those three teams have to agree in writing on who the product is for, what it stands for, and why it's different. I use a one-page corporate strategy summary as the single source of truth for GTM teams, and every launch asset gets checked against it
Substack. Mutiny surveyed 500 sales and marketing leaders and found misaligned teams are 2x more likely to miss revenue targets. 70% of those teams believed they were "mostly or completely aligned," so I always confirm alignment out loud before the launch startsSubstack.
I agree with Micah on one canonical plan, but I'd give it two owners with separate jobs:
- PMM owns feature-to-benefit translation, launch coordination, and sales enablement. PMM teams usually pick safe, proven messaging because they're measured on adoption.
- Brand owns getting people to care enough to reach the value prop in the first place, which means taking creative risks to stand out
If you skip the brand job, the launch reaches only people who already understand you.
Positioning has to pick a lane. "All-in-one" is a feature list, and a launch built on it has nothing people can repeat
kiraklaas. At Brex, we tested the story with customers by working directly with UX research, and we ran "taste-making" sessions with stakeholders before any design work started. That got leadership to agree early on decisions that usually blow up late in a launch
Medium.
Last, plan both the big moments and the everyday ones. A billboard and a support rep's reply both shape what people think of the launchApple Podcasts. Rachel's Notion AI early-access program worked because creators had a lived-in story before launch day. For B2B, I'd budget for that from day one and treat specific named creators and operators as launch channelsSubstack.
Rachel HepworthCMO at Harvey. Former CMO at Notion
Brittany's audience-expansion launch is the one where I've seen brand work against you. When Notion moved upmarket, the consumer love gave us huge awareness and a lot of cachet. It also carried a risk: if buyers peg you as a consumer product, they won't trust you to roll out across a large organization or handle sensitive information. So the launch plan for the enterprise audience had to build credibility on purpose (security, admin features, proof that large organizations do serious work in Notion) rather than lean on the brand people already had.YouTube Notion's mix made this sharper, because roughly 80% of customers were individual consumers while more than 80% of revenue came from businesses. A launch can light up the 80% who make noise and barely reach the people who pay.
I agree with Micah on validating with money. In the customer development framing I used in a 2018 talk, discovery confirms the problem is real and validation confirms people will actually buy a solution. Those are separate gates, and Climate Corp passed the first one and failed the second.YouTube For a paid launch I'd treat the waitlist as message testing and the paid pilot as the forecast. For a free product, Vineet's version works: the forecast is the first behavior that drives revenue, like the direct deposit switch.
One practical note on timing: at Notion the AI Writer launch began with the founders telling everyone to rip up their roadmaps and sprint.YouTube A fast launch like that only worked because the community and creator program already existed. If you expect to launch quickly, build the early-access group before you have a product to give them.
Micah MoreauVP of Growth at OpenAI. Former VP of Growth Marketing at DoorDash
Brittany, fair: a bundle with one customer story is a single launch, and my objection is to a capability buried in a grab-bag event with no story. Vineet, agreed on last-click. When I say marginal CAC I mean incremental CAC from holdouts and geo tests across brand and performance together, which is point 3 below. Rachel's early-access playbook needs one more piece: plan day 31 before day 1. Launches that open with generous free usage (creator access, promo limits, free trials) get a big adoption spike, and then a lot of them stall because nobody designed the bridge from "free and delightful" to "paid and habitual."
On a practical level, three things go in the launch plan up front:
- The conversion bridge. If you hand out generous limits to drive trial, decide in advance what the user sees when the promo ends, what the paid plan is anchored to, and how few clicks it takes to upgrade. Free access with no path to paid trains people to expect it for free.
- One motion for every segment. Rachel's student-in-India vs. VP-in-SF problem tempts teams to build separate self-serve and sales launches. I'd rather let anyone sign up as an org of any size and grow into enterprise features on their own, with sales stepping in where the usage data says a team is ready. Rachel's Notion 80/80 split is why the product motion can be shared while the proof differs: security, admin and enterprise references have to be ready on launch day for the buyers who pay, even if the consumers make most of the noise.
- A holdout. Carve out a geo or user cohort that doesn't get the launch media. Week-one dashboards will credit the campaign for demand that was already coming (Rachel's Elicit example, where word of mouth did the work, is exactly the case where attribution overstates marketing). The holdout tells you the incremental lift, which is the number the CFO scorecard Vineet described should be built on.
This is the same discipline we ran at DoorDash: splashy at launch, then every channel had to prove incrementality against a contribution-margin payback bar once the launch spike faded.
Brittany BinghamVP of Marketing at Fathom. Former VP of Marketing at Guru
In PLG, the most important launch audience is usually people who already use the product, so the signal closest to the economics (to use Vineet's test) is whether existing users adopt the new capability and keep using it. For an integration, that means active installs. Fathom reaching 20,000+ active HubSpot installs told us far more than launch-day traffic did, because each of those installs is a team capturing conversations and pushing notes into their CRM every weekLinkedIn.
That changes who belongs on the launch team. At Guru, my team covered brand, product marketing, comms, growth and lifecycle marketing, and scaled support. Because of that, the launch plan ran from the announcement through onboarding emails, in-product prompts, and the help content support would need in the first week. When support and lifecycle are part of the plan from the start, you hear about the "converts then churns" problem Micah described within days. Otherwise it shows up in a retention chart a quarter later.
So my scorecard for a feature launch has three numbers:
- The share of existing active users who try the feature in the first 30 days.
- The share of those users who are still using it at day 60.
- Support tickets per new user, which tells you where onboarding needs fixing.
Vineet MehraChief Growth and Marketing Officer at Chime. Former Chief Growth and Marketing Officer at Ancestry
Kira's point about named creators as launch channels deserves a real budget line, because most launch plans still treat creators as an afterthought. Roughly 4 to 5% of marketing dollars go to the creator economy, while about 80% of consumer attention sits in creator-led environments. For a launch, creators are where discovery and trust form first, so I plan them alongside TV and paid social from the first budget draft.
Community seeding also protects you when a launch touches something members already love. On a Community-Driven Disruption Forum I hosted, Michelle described launching a new version of a shoe to a community built on the original. The team knew customers had asked for changes but still loved and bought the original, so they seeded samples to that community before launch and let them try it first. The new product arrived with advocates instead of a backlash.
Spotify
At Ancestry we took the same idea into the creative itself. Real customers told the stories in our ads and were cast in them, because hiring an actor to tell a customer's story never lands the same way.Apple Podcasts
So in the plan, I give creators and community three jobs: early access weeks before launch day, a share of the paid budget to amplify their best posts, and a feedback loop into onboarding when they get stuck.
Micah MoreauVP of Growth at OpenAI. Former VP of Growth Marketing at DoorDash
Brittany's scorecard needs a fourth number: a revenue guardrail. Feature launches that hand power users extra capacity (bigger usage limits, reserve pools, bonus credits) can light up adoption and day-60 retention while quietly eating into paid upgrades. If that extra capacity covers what people would have paid for, adoption looks great and revenue per user slides. So I'd track upgrade rate and revenue per active user for the exposed cohort against the holdout, right beside the adoption numbers.
The other thing I'd add on timing: pick the launch date based on when the product is clearly better on the dimension that actually makes people switch. In AI tools that's often speed. Users will move to a new product because it's faster before they move for a longer feature list. A launch calendar driven by a conference date or quarter-end can put you in market while the switching reason is still marginal, and then the media budget pays for trial without conversion.
TL;DR on the framework this group has built so far:
- Name the tension and get product, sales and marketing to agree on it in writing (Vineet, Kira).
- Test the assumptions, and gate on the signal closest to your economics: a deposit for paid, the first revenue behavior for free (Rachel, Vineet).
- Pick the launch type, build the early-access group before you need it, and give creators a real paid budget line (Brittany, Rachel, Vineet, Kira).
- Spend for signal at launch, then hold every channel, brand included, to incremental lift against a holdout.
- Score adoption, retention, support load and revenue per user on one dashboard the CFO signed off on before launch day.
Kira KlaasVP of Corporate Marketing and Brand at Later. Former Director of Brand at Brex
Rachel's upmarket problem is easier to solve if you plan it a few launches ahead. I'd add one question to every launch brief: does this launch's story leave room for the audience you'll need two launches from now?
Brex is the clearest example I know. It launched as "the corporate card for startups." Within its first year, the positioning moved to "the card for growing companies," and then to "the financial operating system for growing companies." Brex wasn't explicitly going upmarket when those changes happened. Even so, the broader story gave the company room when it later moved hard into enterprise SaaS. When Brex offboarded SMBs, the news was briefly confusing, but it didn't feel off-brand because the earlier messaging had prepared the marketSubstack.
For a launch plan, this changes three decisions:
- Positioning: write the launch headline for the segment you're launching to, and check that the company-level descriptor still fits the next segment. Brittany's "Microsoft users, this one is for you!" works because the line targets one segment and leaves the master brand open.
- Proof: Rachel and Micah are right that enterprise proof has to be ready when you launch. I'd go further and start putting enterprise proof (security, admin, named customers) into earlier launches as supporting points, so the enterprise launch confirms what buyers have already seen.
- Sequencing: treat each launch as one chapter of a longer story. A launch that wins its week but narrows the brand makes the next audience-expansion launch more expensive.
This matters most for PLG companies that lean on community love. That love gives you reach, but it can also lock the market's picture of you in place while your revenue is moving somewhere elseSubstack.
Ask a follow-up
+2Marketing Leaders
Ask Marketing Leaders anything...
Frequently Asked Questions
How do you choose an ideal customer profile?
Start from a segment incumbents underserve or from your most retained users, check it in customer conversations, and keep it narrow enough to change your message. Move the core positioning to a new segment once its payback and retention beat the core.
How can you tell if your positioning is working?
Newcomers can explain the benefit and act on it, they name your brand when shown the benefit without a logo, and a matched test group activates and pays back better than the old message.
What has to be ready before a launch?
A named customer and use case, a working path to the promised benefit, a few early users, and measures of source, completion and return. Full campaigns and brand guidelines can follow.
When is a partner program worth building?
When partners already help customers succeed and the team can repeat the sales handoff. Before that, run a small integration pilot and measure activation, retention and support cost.
Which channels bring in technical buyers?
Founder-led outbound when setup is heavy, a self-serve trial when an engineer gets value in one session, and narrow paid search aimed at the trial task, judged on activated accounts.
