Platform decision & activation

I was activating 78% of new members on Skool without writing a single DM. Then I moved to Slack.

Sam Hendricks had been running PLG Practitioners — a paid community for product-led growth operators at B2B SaaS companies at $129/month — on Skool for four months when he decided the platform was producing the wrong conversations. His week-one activation rate was 78%. He had not written a single welcome DM. When he moved to Slack, where his ICP already lived professionally, his activation rate fell to 31% in month one. He spent the following three months understanding what Skool had been doing for him and rebuilding it explicitly.

What Skool was doing for free

Sam launched PLG Practitioners in Q1 2026. Forty founding members at $89/month (founding rate), a three-lesson PLG Fundamentals course (measuring PLG signals, building the self-serve funnel, instrumentation), and a community discussion layer around the course. By month two he had 88 paying members. His week-one activation rate — the percentage of new members who posted at least once within 7 days of joining — was 78%. The monthly active rate was 69%.

The gamification was doing the work, and Sam had not thought carefully about that until he started planning the migration. A new member joined Skool, saw their account at level 1 of 10, opened the course module (completing a lesson awarded 50 points), posted a response in the lesson’s discussion thread because each lesson ended with a prompt (“what does your current activation funnel look like at this stage?”), and saw their level counter advance to 2. The leaderboard on the community homepage showed them where they ranked among current participants. Every one of these was a structured first action with a visible reward that Slack does not provide by default.

Sam had run a day-7 review manually each week, checking which new members had not yet posted. He rarely found more than four or five silent members per cohort of 20–25 new joiners. He had not thought carefully about why. He had assumed his community content was good, his founding cohort was warm, and his ICP was naturally inclined to post. All of those things were partially true. None of them was the primary driver of his activation rate.

Why he moved to Slack anyway

The problem was conversation depth. PLG Practitioners existed for one purpose: practitioners sharing real experimental data on self-serve funnel design, activation metric instrumentation, and PQL scoring models. The conversations Skool’s discussion format was producing were course-adjacent — questions about the lesson material, observations about how a concept applied to a specific use case, confirmations that another member was seeing the same pattern. Good conversations. Not the conversations Sam had built the community for.

His ICP had Slack open 40 times a day for their actual work. They were already in paid Slack communities in adjacent spaces — Lenny’s Community, Demand Curve Circle. The conversations they had in those spaces were different from what Skool’s community discussion layer could support: async and real-time, threaded at depth, DM-enabled, with the ability to share a Notion doc or a Mixpanel screenshot and discuss it in the same interface without switching context. Skool’s interface kept them in a learning posture. Slack kept them in a working posture. Sam wanted the community to live in the working posture.

At month five, he ran a survey. 67 of his 92 members responded. Three questions: how many other Slack communities are you in, what would make PLG Practitioners more valuable, do you use Skool for anything outside PLG Practitioners. 89% were in at least two other professional Slack communities. The top two answers to the second question were “more real-time discussion about live experiments” and “a way to ask a quick question when I’m mid-sprint without feeling like I’m opening a course module.” One member wrote: “I check Skool when I’m in a learning mindset. I check Slack when I’m in a working mindset. I want PLG Practitioners in the working mindset.”

Sam decided to migrate. He understood, as he planned it, that he was trading a platform that solved his activation problem by default for one that did not. He did not yet know exactly how large that tradeoff was going to be.

Month one on Slack: the 31% activation rate

The migration was technically clean. He gave all 92 Skool members four weeks’ notice, set up a Launchpass-gated Slack workspace, moved the course modules to a Notion library, and rebuilt his channel architecture. Twelve channels at launch: #general, #intros, #plg-experiments, #self-serve-funnel, #activation-metrics, #pql-scoring, #growth-wins, #wins-and-blockers, #resources, #jobs, #announcements, #random. He had spent two weeks on the channel structure before opening the workspace to anyone.

In month one, 82 of his Skool members joined the Slack workspace. Eleven did not migrate. Sixteen new members joined directly on Slack — word of mouth from the founding cohort plus two newsletter mentions from community members. His month-one cohort was 98 members.

Week-one activation rate: 31%.

Thirty of 98 members had posted within 7 days of joining. On Skool, 72 of 92 had activated in week one at month four. In absolute terms, he had gone from 78% activating to 31% while adding 6 more total members. He had 42 fewer activated members in a larger community.

His first reaction was to attribute this to migration friction. Members who had moved from Skool would be reorienting to a new interface. New members would need to discover where conversations happened. He expected it to recover in month two as the community settled.

Month-two week-one activation rate: 34%.

Not recovered. The gap was structural, not transitional. The 12-channel blank-slate workspace was not producing the first actions that Skool’s gamification had generated automatically.

Understanding the four things Skool had been doing

Sam went back through his Skool member data and mapped exactly what the platform had given new members in their first seven days that Slack was not providing. He found four structural differences.

A first action with a clear reward. On Skool: complete lesson 1, get 50 points, move from level 1 to level 2. The first action was prescribed, the reward was visible, and the progression was immediate. On Slack: here are 12 channels, check out #intros. The first action was open-ended, the reward was undefined, and nothing tracked the member’s progress through any sequence.

A discussion prompt anchored to shared context. On Skool: each lesson ended with a specific question that gave every current member the same contextual frame. When a new member arrived during lesson-3 week, the discussion board had 40 members answering the same lesson-3 question. The new member had an obvious entry point: answer the same question. On Slack: no shared context existed at the moment a new member joined. The #plg-experiments channel had threads from the last few weeks, but a new member had no particular connection to any of them and no clear prompt to contribute.

Visible social proof of participation. On Skool: the leaderboard showed active members by rank. A new member could see immediately that the top-10 were posting weekly and that their own level placed them at the beginning of a visible social structure. On Slack: the sidebar showed channel names and unread badges. A new member had no visibility into who was active, what level of participation was normal, or whether the community was alive or dormant from their vantage point.

An accountability mechanism the member could see. On Skool: the level and points system tracked participation over time. A member who opened Skool and had not completed the week’s lesson could see their level unchanged. Inactivity was visible. On Slack: nothing tracked participation from the member’s perspective. A week of not opening Slack looked identical to a week of opening it daily — the sidebar showed the same channels either way.

None of these were things Sam could replicate inside Slack’s native tooling. One of them — the first-action prompt anchored to the member’s specific goal — he could replace with a direct message. The other three required either adding a third-party gamification layer (which he had decided against on principle — his ICP found points and leaderboards condescending) or accepting that the platform tradeoff had an activation cost he would need to close through operator effort.

The 30-member manual test

Before building anything automated, Sam ran a manual test on the next 30 members who joined. For each one, he sent a personalised DM within two hours of them appearing in the workspace. The message was not a welcome message. It was a first-action ask tied to their PLG context:

“Hey [name] — welcome to PLG Practitioners. Given that you’re working on [goal from signup question], the most useful first move here is usually to drop one data point in #plg-experiments: what’s one metric from your self-serve funnel that’s surprising you right now? Even a single number with no context generates the most discussion in that channel. What are you tracking?”

The framing was deliberate. The ask was not an introduction (too low-stakes to generate replies from a PLG practitioner audience that values precision) and not an open question (too high a bar for a first post — practitioners uncertain about the quality bar will write nothing rather than risk a shallow post). It asked for a data point, which every PLG practitioner had, and which required almost no framing to share. A metric is a unit of contribution that does not require the member to reveal uncertainty or explicitly ask for help.

The results from the 30-member test: 22 of 30 replied to the DM within 48 hours. 19 posted in #plg-experiments within 72 hours. His baseline without the DM had been approximately 6 of 30 posting in the same window — a 20% rate. The DM produced 63% activation in 72 hours, a 43-point lift from a single two-minute message.

The day-3 nudge was sent only to the 8 members who had not replied after the day-0 DM. Sam deliberately did not repeat the day-0 ask. Instead, he sent something that referenced a specific real contribution from an existing member:

“Hey [name] — three days in, figured I’d check. Marcus posted an interesting data point in #plg-experiments this week: 23% PQL conversion at freemium tier vs. 41% at paid trial. Curious whether your funnel is tracking something similar or diverging. Worth a note in the thread if it resonates.”

The key differences from a generic nudge: he named a specific community member (Marcus), cited a specific real number from a real thread, and framed the invitation as “does this resonate?” rather than “we haven’t heard from you.” The message told the non-poster that something specific had happened in the community since they joined, and that their perspective on that specific thing was worth adding — not that they had failed to do something. Of the 8 non-posters at day 3, 4 replied to the nudge and 3 posted in #plg-experiments within 24 hours.

The day-7 review was Sam’s own manually-run scorecard: which members had posted, which had not, who had replied to the day-3 nudge but not yet posted publicly, and which members were approaching their first renewal date with no visible community activity. At the day-7 checkpoint, 26 of 30 pilot members had either posted in a channel or replied to one of the two DMs. His pilot week-one activation rate: 87%.

The 90-day results after full implementation

Sam automated the day-0 and day-3 messages using Foothold and kept the day-7 review manual — a 15-minute weekly read of the activation scorecard, followed by personal DMs to any member approaching their renewal date with no channel activity. By the 90-day post-implementation checkpoint:

• Week-one activation rate: 31% → 68%. In the 90 days since implementing the sequence, 83 of 122 new members who had joined posted within their first 7 days. His pre-implementation baseline had been running at 31–34%.
• Month-two renewal rate: 64% → 81%. Activated members — those who had posted within 7 days — renewed at 88%. Non-activated members renewed at 51%. The 37-point renewal gap between activated and non-activated members meant that every additional activation was worth an expected $129/month × approximately 4 additional months in the community.
• Weekly active poster rate: 22% → 51%. The percentage of paying members posting at least once per week nearly doubled. On Skool at month four, the comparable metric had been 61% — the 10-point gap to Skool’s gamification-driven engagement level had closed to within reach rather than the 39-point gap he had started with.
• Conversation depth (subjective measure): Threads with five or more replies per week went from 3 (month one, no activation sequence) to 11 (month four, with sequence). The goal-keyed first-action prompt had seeded #plg-experiments with real experimental data, and data-seeded threads drew replies from practitioners tracking comparable metrics. The peer-conversation quality Sam had moved to Slack for was present — but only after the activation layer was in place to generate the seed posts that made the conversations possible.

What Sam says about the Skool vs Slack tradeoff in retrospect

The activation gap is real. Skool’s gamification closes it by default; Slack operators close it explicitly. The gap is not a reason to stay on Skool if your ICP is a professional practitioner who already uses Slack for work — but it is a concrete operational cost that Slack operators take on that Skool operators do not.

Sam’s view at month six post-migration: the conversation quality difference between Skool and Slack, once the activation layer was in place, was material and sustained. The depth of threading, the DM culture, and the ability to share a live Mixpanel dashboard or a Notion spec in a thread and discuss it without leaving the interface produced practitioner relationships his Skool community had not reached. By month five post-migration, three members had co-written a PLG case study together. Two had hired someone they had met in the workspace. Four had formed a small weekly accountability group. None of this would have happened at the conversation depth Skool’s platform had supported for his specific use case.

The operational cost: he had spent approximately 15 minutes per day for the first 30 days manually writing DMs, monitoring the day-7 scorecard, and following up on non-posters. That investment would have been zero on Skool — the gamification had done it for him. Once automated with Foothold, it was about 5 minutes per week of reviewing the scorecard and writing occasional day-7 DMs to renewal-risk members. The ongoing cost was low. The one-time setup cost — designing the goal-keyed DM, testing the day-3 nudge format, building the day-7 review discipline — was about 20 hours spread across the first two months.

His recommendation for operators evaluating the choice: if your community’s value is structured content delivery plus accountability progression, Skool’s gamification is a genuine activation asset and the conversation-depth tradeoff is acceptable for that use case. If your community’s value is professional peer conversation in the interface your ICP already uses for work, Slack is the right platform — and the activation gap is a concrete known cost of that choice, not a reason to avoid it. The cost is closeable with a three-touch sequence. The peer-conversation upside, for practitioner communities, is not replicable on a course platform.

For the full eight-dimension comparison between Skool and Slack — marketplace discovery, member interface, course delivery, gamification, peer-conversation quality, payment, analytics, and toolstack complexity — see Skool vs Slack community: which platform fits professional paid communities in 2026.

Related: Skool vs Slack community — eight-dimension comparison for paid community operators · How to write a paid community welcome DM that gets replies · What is a paid Slack community · Paid Slack community tools — the four-layer stack · Onboarding Health Check — 2 minutes, free.