Slack vs Glitch Behavior Fit #

Jason Hreha· Updated July 4, 2026

Key Result (company-reported): Slack reported more than 1 billion messages sent during the week ended January 31, 2019, alongside 143% net dollar retention at IPO scale (SEC S-1 filing). BS-0006

Background #

Slack is the rare natural experiment in behavior selection: the same founders, much of the same team, and largely the same technology produced one of the fastest enterprise-software failures and one of the fastest successes, back to back. The variable that changed was not talent or execution capacity. It was the behavior the product asked of its users.

Tiny Speck spent 2009-2012 building Glitch, a whimsical massively multiplayer game centered on collaborative creation. It raised significant venture funding, earned press affection, and shut down in late 2012 because not enough people would do the thing the product needed them to do. The team’s internal chat tool - built to coordinate their own work - became Slack, launched publicly in 2014 and IPO’d in 2019.

What actually drove adoption #

Glitch asked users to create collaborative games together: a behavior requiring a niche identity (playful creator), substantial skill, and coordinated blocks of free time. Almost no segment of the market could pass all three filters at once. No amount of polish fixes a behavior the population cannot and will not perform.

Slack asked teams to send messages to each other about work. Workers already did this constantly - over email, SMS, IRC, and hallway conversation. Slack did not have to create a behavior; it had to become the best place for an existing one. That is Behavior Market Fit, and it converted the adoption problem from persuasion into switching.

The design choices that mattered were enablement, not motivation:

One-link team invites removed the setup friction that kills multi-person adoption before network effects can start.
Channels turned scattered updates into persistent, searchable knowledge, so value compounded with use.
Integrations surfaced events (pull requests, tickets, deploys) where teams already acted, embedding Slack into existing workflows rather than demanding new ones.

The main behavioral constraint was that messaging is a team behavior: value arrives only when a whole group activates, so the product’s early funnel was engineered around team-level, not individual, adoption.

Case facts
Company / system	Slack (Tiny Speck)
Industry	Enterprise Software
Population	Workplace teams
Target behavior	Send and read team messages in channels as part of daily coordination
Window	First 30-90 days post-activation; then ongoing retention
Denominator	Teams invited/activated
Key metric	1B+ messages in the week ended Jan 31, 2019; 143% net dollar retention at IPO (SEC S-1)
Behavior fit	Identity: High (messaging colleagues is already part of every worker's self-concept) Capability: High (typing a message requires no new skill) Context: High (always-on workday communication across projects and functions) Fit ratings are analyst assessments unless linked to direct measurement.
Confidence	Working
Evidence	BS-0006

Behavior Fit Assessment #

These ratings are analyst examples; the point is the relative profile between the two behaviors the same company bet on.

Target behavior	Identity Fit	Capability Fit	Context Fit	Why it matters
Glitch: “create collaborative games”	Low	Low	Low	Niche identity + high skill + coordinated time blocks required
Slack: “send team messages in channels”	High	High	High	Universal workplace behavior + no skill barrier + always-on context

A behavior ranking exercise in 2009 would have flagged Glitch’s target behavior as a long-odds bet before a line of code was written. That is the practical value of assessing fit before building.

Results #

More than 1 billion messages were sent in Slack during the week ended January 31, 2019 (SEC S-1 filing). BS-0006
Net dollar retention was 143% at IPO, meaning existing teams expanded usage over time (SEC S-1 filing).
Roughly 10% of FY2018 revenue and 8% of FY2019 revenue came from organizations that converted from the free tier to paid (SEC S-1 filing); note this is a revenue-mix disclosure, not a user conversion rate.
Adopting teams reported a 32% reduction in internal email volume (company-reported survey).
A widely cited activation threshold - roughly 93% retention once a team exchanges 2,000+ messages - comes from founder interviews and press coverage, not the S-1, and should be treated as indicative.
Glitch, the predecessor product, shut down in 2012 after failing to sustain an audience for its target behavior.

Limitations #

Most adoption metrics here are company-reported and milestone-dependent, so treat thresholds as indicative rather than universal. Team culture, organizational communication norms, network effects, and freemium dynamics all confound clean attribution. The comparative claim, however, does not rest on any single metric: one behavior selection died and the other reached IPO with the same team behind both.

Lessons #

Behavior selection dominates execution quality. The same team executed both products. The product built on a universal existing behavior thrived; the one requiring a rare new behavior died.
Prefer becoming the best venue for an existing behavior over creating a new one. Slack’s growth problem was switching, not persuasion - a categorically easier problem, and the essence of behavior matching.
Match the adoption unit to the behavior’s unit. Team messaging is performed by teams, so activation, onboarding, and measurement had to operate on teams. Getting the denominator right was itself a strategic decision.