In the summer of 2013, Stewart Butterfield was trying to salvage something from the wreckage of a failed video game. Glitch had shut down. The team had scattered. But internally, Tiny Speck had been using a homemade messaging tool to coordinate across time zones, and that tool worked unusually well. Butterfield decided to productize it.
The feedback he got back was consistent and, on the surface, reasonable. Enterprise communication was a graveyard. Microsoft had poured resources into Lync and barely moved the needle. HipChat existed and was going nowhere fast. Email was entrenched. IT departments controlled purchasing decisions, and IT departments did not buy chat tools from video game companies. The market had spoken, multiple times, to multiple people with more resources.
Butterfield ignored essentially all of it. Slack launched in August 2013, and within 24 hours had 8,000 sign-ups. Within two weeks, 15,000. The company eventually reached a valuation north of $26 billion when Salesforce acquired it in 2021.
So what did Butterfield know that all those reasonable people did not?
The Setup
The feedback Butterfield received was not wrong about the facts. Enterprise chat had failed. IT departments were the wrong buyers. Email was entrenched. Every data point was accurate.
The problem was that the feedback was answering a different question than the one Butterfield was actually asking. Critics were asking whether the enterprise communication market had room for another product competing on the same terms as existing products. Butterfield was asking whether there was a group of people whose daily work lives were genuinely worse because of how communication tools worked, and whether he could reach them directly before any gatekeeper got involved.
Those are not the same question. The first question had a clear answer: no. The second question had a different answer entirely.
This distinction matters enormously, because most market feedback is shaped by the frame of whoever is giving it. When early users, investors, or competitors tell you a market is saturated or the problem is already solved, they are usually describing the world as it exists from their vantage point. They are not necessarily describing the world as it would exist if you executed on a genuinely different hypothesis.
What Actually Happened
Slack’s early go-to-market strategy was almost aggressively indirect. Rather than targeting IT buyers or pitching to enterprise procurement departments, the team got individual teams inside companies to try the product. The unit of adoption was the team, not the organization. The buying decision was made by people whose daily experience the product was actually improving, not by people evaluating compliance requirements.
This mattered because it meant the product got to prove itself before anyone with institutional reasons to reject it had a chance to weigh in. By the time enterprise IT departments were evaluating Slack, thousands of employees inside those companies were already using it and already dependent on it. The typical enterprise sales problem, which is getting anyone to take a risk on something unproven, was inverted. The risk now was removing a tool people had built their workflows around.
Butterfield had identified a specific mechanism by which the usual rules did not apply: the rise of software-as-a-service meant that individual teams could adopt tools without central approval. The feedback he was getting was calibrated to an older model of enterprise software adoption. He was operating in a different one.
Why Most Founders Get This Wrong
The problem with feedback is not that people give you bad information. It is that feedback arrives in a form that feels authoritative, and most founders lack a clear framework for deciding which authority to trust.
There are roughly two categories of negative feedback worth distinguishing. The first is feedback that tells you something is hard or competitive or uncertain. This is almost always true and almost never a reason to stop. The second is feedback that reveals a specific, structural reason why your particular approach will not work, grounded in something you have not accounted for. This kind is rare but worth listening to carefully.
The feedback Butterfield received was entirely in the first category. Yes, enterprise chat was hard. Yes, it was competitive. Yes, email was entrenched. None of that constituted a structural argument against the specific bet he was making, which was that bottom-up SaaS adoption had changed the rules of how communication tools spread inside organizations.
When you look at founders who successfully ignored negative feedback, there is usually a version of this pattern. The feedback is accurate about the general territory but wrong about the specific path. Brian Chesky heard that no one would rent out space in their home to strangers. That was true about a generic home-rental product marketed to generic users. It was not true about a carefully designed experience with reputation systems, photography support, and a community built around trust. Travis Kalanick heard that city transportation was locked up by regulators and incumbents. Accurate about traditional taxi licensing. Not accurate about whether a sufficiently useful service could create political pressure faster than regulators could act. (Whether Uber’s particular execution was worth it is a separate, legitimate argument.)
The founders who ignored feedback badly, and there are many of them, tended to hear feedback in the first category and mistake it for noise because they wanted to. Theranos received structural warnings, not just competitive skepticism. WeWork’s investors raised genuine unit economics concerns years before the collapse. Ignoring feedback is not a virtue. It is sometimes the right call and sometimes catastrophic, and the difference is whether you have actually understood what the feedback is telling you before deciding it does not apply.
The Framework Worth Borrowing
The practical question is not “should I listen to market feedback” but “what is this feedback actually measuring.” When someone tells you a market is too crowded or a problem is already solved, ask what assumptions are embedded in that claim. Usually the assumption is that you will compete on the same terms as existing players. If your hypothesis is that a different mechanism of distribution, adoption, pricing, or positioning changes those terms, the negative feedback is probably not about you.
If, on the other hand, the feedback points to something structural, a regulatory barrier you cannot move, a unit economics problem that does not improve with scale, a customer behavior that is genuinely fixed rather than just habitual, that is worth sitting with. The distinction between charging too little and pricing strategically matters here too, because feedback about pricing is often structural in a way that feedback about market size is not.
Butterfield’s actual insight was not that enterprise chat feedback was wrong. It was that the feedback was measuring a world where individual teams could not adopt tools without IT approval, and that world had quietly changed. He had evidence from inside his own company that the tool worked. He had a thesis about why the adoption mechanism was different. He had a specific population of early users who were already sold.
That combination, working product, a clear thesis about why the usual rules did not apply, and a real group of people whose behavior confirmed the thesis, is the bar. Without all three, ignoring market feedback is not boldness. It is just stubbornness with a better story attached.
