Distribution as infrastructure
How channel decisions quietly dictate who your users are, how you price, and what your company is forced to become.
Distribution as infrastructure
Founders tend to talk about distribution like it’s something you can keep tuning indefinitely. Channels, pricing, and funnels can be iterated until things work, right? That mindset used to be mostly fine, because go-to-market changes were painful but survivable, but we don’t seem to be living in that world anymore.
Distribution has become harder, noisier, and more path-dependent. Paid channels that used to deliver predictable ROI now require constant hand-holding just to stay flat. Self-serve funnels are less forgiving as buyers have more options and less patience. Outbound is flooded with AI slop and automation, with buyers conditioned to ignore anything that smells templated. At the same time, most categories have settled into a “normal” way software gets evaluated and purchased, which means you’re not simply picking a way to acquire customers, but also teaching the market how to buy from you.
Thus, distribution increasingly behaves like infrastructure. Once a motion starts working, the company begins reorganizing around it (through hiring, roadmap tweaks, and process-building) in subtle ways that feel reasonable in the moment. Over time, the motion stops feeling like a choice and starts feeling like how the company operates. Yes, that can compound momentum, but it also locks you into constraints you didn’t fully model when you were just trying to “get traction.”
The founders who handle this well treat distribution with the same gravity they treat other irreversible decisions, like picking a core customer persona or committing to a technical architecture. They know that switching distribution later rarely stays contained inside marketing and sales. It usually pulls product, operations, hiring, and customer expectations along with it.
I’ve seen at least three ways this goes wrong, and they all look like progress right up until they don’t.
1) When volume hides the problem
Some channels look incredible because they produce fast growth, and fast growth is intoxicating. The catch is that volume can drown out the thing that actually matters: the fact that channels filter for incentives and will ultimately select the “kind” of customer you get.
Groupon is a clean example because the mechanism was visible in real time. In the early daily-deals boom, Groupon could drive a wave of customers to a local business overnight. That felt like validation, so the machine scaled. The downside showed up in the economics. A meaningful share of merchants reported these promotions weren’t profitable and many said they wouldn’t run them again. Studies and analyses from the period also pointed to the same dynamic: deep discounts can attract customers who behave differently than the customers you actually want long term, and merchants often struggle to convert them into repeat, full-price buyers.
This is a trap founders and growth leaders fall into time and again because the motion looks textbook: you acquire customers, listen to feedback, improve onboarding, optimize activation, and the company gets better at serving the segment arriving through that channel. The sneaky problem is that the segment itself may be wrong for the business you want to build, and the better you get at serving it, the more committed you become.
An important question worth asking early: if this channel keeps working for the next five years, are these the customers you want your company built around? Not the customers you can close this quarter, but the customers whose expectations will shape product, pricing, and the operating model. If the answer is “I’m not sure,” it’s worth pausing before the org consolidates around a motion you’ll later spend years trying to unwind.
2) When you borrow from the future
The second failure mode is trickier because it often looks smart in the short term. Companies manufacture traction by subsidizing customer acquisition: underpricing to win deals, overserving to land logos, and customizing heavily to push adoption. The internal story is that it’s temporary- you buy your way into scale, then you clean up the economics later.
Sometimes that play works, but only under specific conditions. It works when the subsidy helps you secure something durable (like big switching costs, category dominance, pricing power, etc.) so that later you have leverage. The risk is that, while you’re subsidizing, you’re also teaching the market what to expect, and those expectations don’t reset just because the spreadsheet says it’s time.
Underpricing anchors customers, overserving sets a service baseline, and heavy customization teaches the team that exceptions are how deals get done. Meanwhile the roadmap starts bending toward the customers who show up under those terms, and hiring follows the work required to fulfill the promises. By the time you decide to “climb out of the hole,” it’s no longer a single adjustment. You’re undoing a set of habits and commitments that have come to define how the company operates.
Zenefits is a strong illustration of how quickly this can spiral. In the mid-2010s, Zenefits grew explosively by giving away HR software and monetizing primarily through insurance brokerage commissions, which made customer acquisition feel almost frictionless. That growth also put the company directly in the path of regulatory and compliance complexity, and by early 2016 the fallout was serious enough that founder/CEO Parker Conrad stepped down and David Sacks took over amid investigations and public scrutiny. The recovery wasn’t a light cleanup by any means. It involved layoffs, a reset of controls, and a reboot of the product and model under “Z2,” including charging for products that had previously been free and building compliance tooling into the platform.
The takeaway isn’t “never subsidize.” It’s that subsidized distribution creates obligations all across the board (to customers, internals, and sometimes even regulators) that don’t unwind on command. If the motion depends on aggressive concessions up front, it’s worth assuming those concessions will echo through pricing, sales behavior, customer expectations, and product direction for longer than you’d like. Keep that in mind.
3) When your channel rewires the company
The third failure mode has less to do with customer quality or unit economics and more to do with organizational gravity. Once a channel starts working, the company gets shaped to make that channel work better. The motion gradually becomes the operating system.
You can see it in the two most common “company shapes.” A product-led, self-serve motion pulls you toward low-friction onboarding, fast time-to-value, and an operating cadence built around activation and retention. An enterprise outbound motion pulls you toward procurement readiness, security posture, implementation rigor, and the kind of org design that can repeatedly win and retain large accounts. Neither is better on its own. The problem shows up when a team assumes it can jump between those worlds later without paying a serious price.
Slack is a useful example because you can trace the transition. Slack’s early growth was bottom-up: teams adopted it on their own, usage spread horizontally, and larger organizations often standardized it after it was already everywhere. That success shaped product and go-to-market around team-level delight and viral expansion. When Slack wanted to push harder into large enterprises, the constraint wasn’t sales headcount. Large organizations needed admin, governance, identity management, and compliance controls that made IT comfortable rolling it out broadly. Slack’s Enterprise Grid launch in 2017 was the clearest signal that moving upmarket required new infrastructure (not just a “louder” sales motion).
The takeaway here is it’s easy to underestimate switching costs. A channel shift looks like a go-to-market change on the outside. On the inside, it often becomes a shift in what the company is built to do well.
Choosing constraints on purpose
Under pressure, especially in the earlier stages of company building, it’s natural to ask what will produce traction fastest. The more useful question is what kind of company a channel forces you to become if it works. That framing keeps you out of tactical thrash and pushes you into constraints, which is where the real decisions live.
The fork between product-led self-serve and enterprise outbound makes this concrete. Self-serve tends to select for customers who can evaluate and adopt without human intervention, which means you’re committing to transparent pricing, intuitive onboarding, fast time-to-value, and a product that carries most of the adoption burden. Enterprise outbound tends to select for customers who expect consultative sales, custom pricing, and guided implementation, which means you’re committing to sales capacity, solution engineering, security and compliance readiness, and a product that supports complex workflows.
The common mistake is trying to keep both doors equally open for too long. In practice, you end up accumulating complexity rather than compounding a coherent advantage, and the constraints show up anyway. Honestly, they just show up by default.
A simple pressure-test, before you hard-commit to a motion, is to walk through five questions in order: who the channel tends to select for, what expectations it sets around pricing and service, what the company needs to become excellent at for the motion to work, what it will pull onto the roadmap, and what it would take to unwind if you’re wrong. If you can’t answer those cleanly, that’s usually a sign the motion hasn’t been thought through as a company design decision yet.
What it looks like when it works
Atlassian is a strong example of a company aligning the distribution motion with the company design. For years, they leaned into a high-velocity, low-touch model with minimal traditional sales and little price negotiation, supported by products that could be adopted by teams and expanded broadly over time. You can see that philosophy in their public filings and in how they describe their go-to-market model.
What’s interesting is how many companies try to imitate the surface mechanics without making the underlying commitments. They want the efficiency of self-serve while still building a product that requires a lot of human help to adopt. They want inbound without simplifying the buyer journey. Atlassian worked because the motion and the company were built to fit each other.
Distribution choices create constraints whether you think about them or not. The difference is whether you choose them deliberately, or wake up one day and realize you’ve been building around them by accident.