Betamax had a sharper picture, a sturdier tape, and a more compact cassette than VHS. It lost anyway. Sony kept a tight grip on licensing and priced its machines as a premium product. JVC did the opposite: it let anyone manufacture VHS decks, courted Hollywood for movie releases, and pushed recording times long enough to fit a football game on one tape. By the mid-1980s the technically worse format owned the living room, and the better one was a trivia answer. The engineering did not decide it. The distribution did.
I bring up a format war from the 1980s because product managers keep re-fighting it and keep losing it the same way. We fall in love with the artifact. We benchmark features against a competitor, close the gaps, add a few of our own, and assume the market will notice. It usually does not. The uncomfortable truth of product strategy is that the best product wins far less often than the best-distributed one, and most roadmaps are built as if the opposite were true.
The claim that should change how you plan
Peter Thiel put it more bluntly than most PMs are comfortable with. In Zero to One, he argues that superior sales and distribution by itself can create a monopoly even with no product differentiation, and that the reverse is not true: no matter how good your product, you still need a distribution plan that works. His line that sticks with me is that if you have invented something new but have not invented an effective way to sell it, you have a bad business regardless of how good the thing is. Poor distribution, he says, not a bad product, is the single most common cause of failure. You can read the full argument and it holds up better every year.
Most product teams treat distribution as someone else’s job. Marketing owns acquisition, sales owns the pipeline, and product owns the thing itself. That division of labor is exactly the trap. If distribution is the deciding variable, then how a product gets into people’s hands is a product decision, not a downstream one. It belongs on the roadmap next to the features, and often ahead of them.
What the data says about who actually wins
This is not just format-war nostalgia. The clearest modern evidence sits in how software companies grow. Product-led growth, where the product itself is the primary channel through free tiers, trials, and built-in sharing, has quietly become the default rather than the exception. Roughly 58% of companies now run some version of a PLG model, and among B2B software companies above 50 million dollars in annual recurring revenue, that figure climbs to about 91%. The reason is not ideology. It is the numbers: PLG companies grow revenue around 50% faster than sales-led peers while spending roughly 39% less on sales and marketing, according to the benchmark data.
Sit with that second half for a second. Faster growth on less spend. That is not a marketing efficiency. That is distribution engineered into the product, which means it is a product strategy that happens to show up on the growth chart. When the product does the selling, the company keeps the margin that a sales team would have consumed.
Dropbox is the canonical case, and the specifics matter more than the legend. Its referral program offered 500 megabytes of free storage to both the person inviting and the person joining. Over 15 months, signups went from 100,000 to 4 million, a 3,900% increase, and referrals drove roughly 2.8 times more signups than paid acquisition. Here is the part most retellings skip: before the referral program existed, about a third of Dropbox signups already came from word of mouth. The program did not manufacture that demand. It captured behavior that was already happening and gave it a rail to run on. You can read the full case study, and the lesson is that the growth mechanism was designed into the product, not bolted onto it afterward.
The mistake I watched play out in telecom
I spent years in Saskatchewan telecom and then two decades in IT operations, and I have watched the better product lose in person more than once. The one that still bothers me involved a genuinely superior technical offering: better reliability numbers, a cleaner install, lower long-term cost per seat. We could prove every claim. We lost the accounts anyway, and it took me too long to understand why.
The competitor did not win on the product. They won on how the product reached the buyer. They were already inside those accounts for something else, so their new offering arrived pre-trusted, bundled into an existing relationship and a single invoice. We arrived as a standalone decision that required someone to champion a switch, justify it to a committee, and absorb the risk of being the person who recommended the new vendor. Our product was better on the spec sheet. Their distribution was better in the room where the decision got made. The spec sheet does not attend that meeting.
What I would tell my younger self is that we were optimizing the wrong variable with total confidence. Every planning cycle went into closing feature gaps and widening our technical lead, because that felt like the honest, rigorous thing to do. Nobody in those rooms was asking the harder question: through what specific channel does this reach a buyer with less friction than the alternative? We had no answer, and a great answer to the wrong question is still the wrong answer.
Why feature parity is a strategy that quietly loses
The default posture of a threatened product team is to match the competitor feature for feature. It feels safe because it is measurable and it silences the sales complaints. It is also, more often than not, a slow way to lose.
Feature parity concedes the premise that the product is where the contest happens. If your rival already has the distribution advantage, and you spend your roadmap catching up on features, you have volunteered to compete on the axis where you are behind while ignoring the axis that actually decides it. Google+ is the expensive version of this lesson. It was, by several accounts, a perfectly capable social network, and in some respects a better-designed one than Facebook. It failed because it tried to be Facebook plus a little more, and a little more is not enough to overcome an incumbent’s network and the switching cost of leaving where your friends already are. Being better on features did not move people. There was no distribution wedge, only a nicer version of a place people had no reason to leave.
I have written before that your product strategy is probably a feature list when it should be a theory of how you win. Distribution is the missing half of that theory. So is counter-positioning, which is really a distribution and business-model bet disguised as a product one: you win not by having more features but by choosing a way to reach and serve customers that the incumbent cannot copy without damaging their existing business.
Putting distribution on the roadmap
The practical shift is small to describe and hard to do: treat the channel as a first-class design constraint, from the first planning conversation, not a launch afterthought.
A few questions I now insist on before a bet gets funded. Through what specific channel does this reach its buyer, and why is that channel cheaper or faster than the competitor’s? Does the product create its own distribution, the way a shared document or an invite or a public artifact pulls in the next user, or does every new customer require the same full-price acquisition effort as the last? If we win the first hundred customers, does customer 101 get easier or exactly as hard? A product where growth compounds through use is running a different strategy than one where growth is purchased one seat at a time, even if the two look identical on the feature comparison.
None of this argues that product quality is optional. A product people quietly abandon will not survive any distribution engine; the referral loop only worked for Dropbox because the underlying product was worth referring. The point is about sequence and weight. Quality is the price of entry. Distribution is frequently the thing that decides the winner among products that have all paid that price. Even the newer hybrid models make this explicit: McKinsey now describes the emerging default as product-led sales, where the product attracts and qualifies the buyer and a sales team closes, which is another way of saying distribution and product are being designed as one system rather than handed off between departments.
The next time your roadmap is a stack of features aimed at catching or beating a competitor, stop and ask what that competitor actually beat you with. If the honest answer is that they reached the buyer more easily than you did, then no amount of feature work fixes it, and the most strategic thing on your roadmap is not a feature at all. It is a channel. Betamax engineers understood video better than anyone at JVC. It did not save them, and it will not save you.
Related reading: pricing is product strategy and how product-market fit signals can fool you.