{"id":90051,"date":"2026-05-26T07:00:00","date_gmt":"2026-05-26T11:00:00","guid":{"rendered":"https:\/\/paulsyng.com\/blog\/?p=90051"},"modified":"2026-05-26T07:00:00","modified_gmt":"2026-05-26T11:00:00","slug":"product-brand-slop","status":"publish","type":"post","link":"https:\/\/paulsyng.com\/blog\/product-brand-slop\/","title":{"rendered":"“Product > Brand” slop"},"content":{"rendered":"

\u265b
26.05.2026<\/p>\n

\nThe Best Product Does Not Win<\/strong><\/h2>\n

Read Time: 6\u00a0Minutes<\/em><\/p>\n

Hello<\/p>\n

Before we get into today’s topic, two pieces from this week are worth a few minutes of your time.<\/p>\n

The first is a positioning read of the SpaceX S-1, three weeks before the largest IPO in history. Five customer bases describing five different companies, and one noun the label has never used. Read it here.<\/p>\n

The second is Part 3 of The Autopilot Corporation series. The first two pieces tracked drift inside a single company. This one tracks the same drift across a whole economy, and what fourteen years of cheap capital did to the question of who the marginal decision-maker inside a public company actually was.<\/p>\n

1. The Noun Elon Musk Left Out of the SpaceX S-1<\/strong><\/p>\n

SpaceX is three weeks from the largest IPO in history, and not one of its five customer bases describes the same company. A NASA Administrator talks about American sovereignty. A South Korean defense ministry talks about independent eyes overhead. Iridium’s CEO talks about a financial transformation. Anthropic talks about compute that is not also a competitor. A grandmother in Wisconsin calls it a lifeline.<\/p>\n

Five voices. Five registers. One transformation pattern none of them coordinate on.<\/p>\n

The S-1 names the destination (Mars, multiplanetary civilization) and leaves the transformation silent. The capex, the contracts, the refusals, and the compensation grant have been writing the noun for 24 years. The label has never used it.<\/p>\n

If the analyst class picks the wrong word on June 12, the wrong word sticks for years.<\/p>\n

Read the positioning analysis<\/a><\/p>\n

2. The Autopilot Corporation: When the Whole Economy Forgets the Customer<\/strong><\/p>\n

Part 3 of the series. The first two pieces were about drift inside one company. This one is about drift across a whole economy.<\/p>\n

Fourteen years of near-zero rates rewired who the marginal decision-maker inside a public company actually was. It stopped being the person with a wallet and became the institution with a metric. ESG ratings. DEI dashboards. AI strategy slides. Three asset managers became the largest shareholder in 88% of the S&P 500 and built a structural incentive to defer to whatever management was already doing.<\/p>\n

Then the greenium collapsed from six basis points to zero, and the bond market issued its own quiet verdict on the whole arrangement.<\/p>\n

Capital re-pricing creates the incentive to refocus. It does not create the capability. The companies that come through the next five years intact are the ones that already had customer obsession built in, or the ones whose new leadership is willing to do what Mulally did at Ford and sell off the parts that do not connect to the answer.<\/p>\n

Read Part 3<\/a><\/p>\n

Now, today’s topic.<\/p>\n

\"\"<\/p>\n

Brand is Not the Opposite of Product<\/strong><\/p>\n

I have spent the last year watching the same argument play out across LinkedIn and X. Two slogans keep recycling, treated as fresh insight, rewarded with engagement.<\/p>\n

The first slogan: product is greater than brand.<\/p>\n

The second slogan: the best product wins.<\/p>\n

Both are wrong in ways that matter, and both are wrong in ways anyone running positioning work can demonstrate without much effort. The slogans persist because they preserve a flattering story for the people who tell them, because the platforms reward confident generalization more than diagnostic accuracy, and because the cases that would falsify them get forgotten faster than the cases that confirm them.<\/p>\n

I work in this terrain every week, running positioning audits for companies whose brand health and product health have drifted apart. What follows is what I see in the actual diagnostic data, before any company has had time to dress it up for an investor update.<\/p>\n

Today’s essay does three things. It shows why “product versus brand” is a category error. It shows why “best product wins” is empirically false. And it shows why both slogans survive despite the evidence, and what stops smart founders from seeing through them.<\/p>\n

Brand is the Residue of Signals<\/strong><\/p>\n

Markets are not adjudicating a contest between product and brand. They are reading a continuous stream of costly signals and leaving the residue behind. The residue is what people call brand.<\/p>\n

Product is one of those signals. Service is another. Founder behaviour is another. So are pricing decisions, hiring choices, partnership terms, which customers a company chases and which it walks away from, how the company handles outages and recalls and bad press, and the consistency of all of these over time.<\/p>\n

The buyer reads each signal independently. The accumulated reading produces a pattern in the buyer’s memory. The pattern is the brand. Brand is downstream of the signal stack. Product is one of the upstream contributors.<\/p>\n

Arguing “product is greater than brand” is like arguing that a tree is greater than the forest. The categories do not compete because one is part of the other. The lived experience of the buyer is walking through the whole forest, and what they remember is the forest.<\/p>\n

The frame is older than the slogans defending it. On May 13, 1931, Procter and Gamble’s Neil McElroy wrote a three-page memorandum to R.F. Rogan proposing that each P&G brand should be run by a dedicated team, each with its own marketing budget, audience and mental availability slot with a shared back-office and independent positioning. McElroy was twenty-seven. He went on to become president of P&G in 1948 and Secretary of Defence in 1957. The memo is archived at Harvard Business School as the founding document of modern brand management.<\/p>\n

The literature has been continuous since. Byron Sharp at the Ehrenberg-Bass Institute has published forty years of empirical work showing that market share is driven by mental availability (does the buyer think of you in the buying moment) and physical availability (can the buyer actually get you). Both are produced by accumulated signals, not by individual product attributes. The economic foundation is Michael Spence’s 1973 paper on costly signalling, which earned the Nobel and proved formally that observable costly actions can communicate quality in markets where buyers cannot verify it directly. A low-quality firm cannot profitably mimic a costly signal forever.<\/p>\n

If you accept the frame, the slogan does not just fail. It stops making sense as a position. Asking whether product or brand matters more is asking whether the input matters more than the output.<\/p>\n

Gravity, Glitter, and the Illegibility Defence<\/strong><\/p>\n

The next move in the diagnostic is to ask which signals in the stack are visible to competitors and which are not.<\/p>\n

Most companies overspend on the visible layer. Messaging, campaigns, visuals, slogans, taglines, refreshes. I have called this glitter for years. It is the surface a competitor can read in an afternoon and copy in two weeks.<\/p>\n

The signals that actually carry the brand are gravity. Structural commitments. Capital allocation. The product roadmap. Who does the company say no to? Where seventy percent of resources go. The decisions are baked into hiring, compensation, partnership terms, customer qualification, refund policies, and escalation paths. These signals are not visible from outside the company, and they are not copyable on a sprint cycle. They take years to install and longer to dismantle.<\/p>\n

Scott Stevenson<\/a> has been writing about a closely related distinction under the heading “legibility”. A legible system can be summarized, enumerated, and centrally comprehended by people outside it. A wild forest is illegible. A planted commercial forest is legible. The planted forest is also more fragile, because the legibility came at the cost of the resilience the wild forest had built into its messiness.<\/p>\n

Stevenson’s diagnostic for competitive advantage is sharp. If a competitor can fully understand and copy your advantage by reading your website or skimming your case studies, the advantage is glitter, not gravity. Real gravity requires years inside the operation to see the web of decisions that produced it.<\/p>\n

The single most useful line in his framework: if your competitive advantage can be summarized in one clean sentence, you probably do not have one. Illegibility is the defence.<\/p>\n

The legibility gradient maps onto the four-level positioning canvas I have used for years.<\/p>\n

Saying It <\/strong><\/em>is maximally legible. Anyone can read your homepage. Pure glitter.<\/p>\n

Proving It <\/strong><\/em>is mostly legible. Case studies, numbers, references. Structurally similar competitors can mimic the pattern.<\/p>\n

Living It <\/strong><\/em>is partially illegible. Outsiders can see individual costly decisions: no freemium, brutal qualification, weird pricing. They cannot see the full rationale behind them.<\/p>\n

Being It<\/strong><\/em> is maximally illegible. The position is built into org architecture, hiring discipline, ritual, compensation, product architecture, and partnership terms. Competitors can see the outcome. They cannot see the recipe. They can copy your homepage in an afternoon. They cannot copy the bottom two levels in five years.<\/p>\n

Companies winning their categories operate at Levels 3 and 4, while companies arguing about Level 1 lose.<\/p>\n

Signal-Stack Fit and the Graveyard<\/strong><\/p>\n

The second slogan, “the best product wins,” fails on first principles before any case study is required.<\/p>\n

There is no single scalar called product quality. Products are multi-attribute bundles. Different buyers weigh attributes differently. Different buying contexts privilege different attributes. The compression of all of this into one ranking that picks a winner is the engineer’s fantasy.<\/p>\n

What markets reward is signal-stack fit. The alignment between the entire bundle of signals the company sends and the jobs, identities, constraints, and fears on the buyer side. Distribution fit. Network and ecosystem fit. Pricing and business-model fit. Category narrative legibility. Founder credibility. Timing relative to platform shifts and regulation. All of these get weighed by the buyer simultaneously, often subconsciously, often through committees with incentives the engineer never sees.<\/p>\n

The graveyard of measurably superior products that lost is large enough to make the slogan empirically untenable.<\/p>\n

Betamax had a better picture standard than VHS when both launched in the mid-1970s. Smaller cassettes, higher recording quality. Sony restricted Betamax to its own hardware. JVC licensed VHS to anyone willing to ship a player. VHS could hold a full movie on one tape before Betamax could. By 1987, VHS controlled the United States home-video market, and Sony was conceding.<\/p>\n

HD-DVD had a stronger initial picture standard than Blu-ray and outsold Blu-ray players through most of 2007. Sony bundled Blu-ray into every PlayStation 3. Warner Bros announced Blu-ray exclusivity on January 4, 2008. HD-DVD was dead within months.<\/p>\n

BlackBerry had the best physical keyboard, the most secure push email infrastructure, and the strongest enterprise IT integration in the smartphone market. The iPhone App Store, launched in July 2008, created a developer network effect that BlackBerry could not match. RIM peaked at 19.9 percent of the global smartphone market share in 2009 and was at 0.048 percent by the end of 2016.<\/p>\n

Lotus 1-2-3 dominated DOS spreadsheets for nearly a decade and then bet on IBM’s OS\/2 over Microsoft Windows. The Windows version shipped over a year late and never recovered. Excel took the category. WordPerfect lost the same way against Word.<\/p>\n

Friendster predated Facebook with similar features and sometimes took forty seconds to load a profile at peak. Google Plus launched with better privacy primitives than Facebook and never developed the network gravity. Quibi raised one-point-seven-five billion dollars from Disney, WarnerMedia, Alibaba, and Hollywood studios. Premium short-form video, mobile-only, at the moment of a stay-at-home pandemic, with AirPlay and Chromecast deliberately disabled. Seven months from launch to shutdown. Color raised forty-one million dollars from Sequoia, Bain Capital, and Silicon Valley Bank pre-launch and solved a problem users did not have. Sold for talent within eighteen months. Webvan burned through more than eight hundred million dollars in nineteen months and filed Chapter 11 on July 9, 2001, because the warehouse economics never worked.<\/p>\n

Each of these products was superior to whatever beat it on the metric the engineers cared about most. None of them survived.<\/p>\n

The buyer-behaviour insight underlying the pattern is one that Stevenson clearly states. Buyers in complex organizations are not optimizing for truth. They are optimizing for legible defensibility. They pick what they can explain to their committee, what will not get them fired, what falls inside the recognized frame the organization already knows how to absorb. The “better product” that wins is the one whose entire signal stack reads as legibly defensible to the buyer’s organization, not the one whose specs are highest on a feature comparison table.<\/p>\n

Why the Slogans Survive<\/strong><\/p>\n

If the slogans are this empirically wrong, the next question is why they persist.<\/p>\n

Three mechanisms are doing most of the work.<\/p>\n

The first is survivorship bias. We look at Apple, Tesla, Google search, Stripe, and we call them the best. They became known as the best because they won. The label is retroactive. Abraham Wald demonstrated the pattern during World War II when he advised the United States Air Force to armour the parts of returning bombers that showed no bullet holes, on the grounds that the planes hit in those places never came back. The visible cases are the survivors. The invisible cases are the failed ones, which are most of them, and which the slogan ignores.<\/p>\n

The second is what Stevenson has called the Fantasy Game pattern. Founders pursue structured dopamine hits that simulate progress while avoiding the actual illegible work. The enterprise versions are easy to name. Another homepage rewrite. Another competitive intelligence tool. Another repositioning off-site. Another marketing reshuffle. Each is comfortably legible. Each is also a way to avoid making the uncomfortable, costly, irreversible positioning decision that the company’s actual signal stack requires.<\/p>\n

The third is what David Sirlin called scrub mentality when he was writing about competitive games, and which Stevenson has applied to positioning. A scrub imposes self-invented rules the game does not recognize. In positioning, the scrub move is the belief that a good strategy must be complex, multi-pillar, hard to explain, and layered with sophistication. The market rewards clarity. Performative complexity is a social move inside the firm, not a winning move outside it.<\/p>\n

The slogans survive because they are legible answers. “Build a better product” is a Lego instruction. Snap the right pieces together, and the company should click. Reality is closer to woodworking. The grain runs where it runs. The wood splits where it wants. The work is granular, idiosyncratic, unforgiving, and never stops. The manual toil is the game.<\/p>\n

Most companies quit the woodworking somewhere between Levels 2 and 3 because the illegibility starts to feel like something is wrong with the strategy. The illegibility is not the problem. The illegibility is what a real strategy feels like in the wild.<\/p>\n

Tesla<\/strong><\/p>\n

The cleanest live test of the frame is Tesla. Tesla currently sells the best mass-market electric vehicle on the road by several measures that most buyers care about. The Model Y has the largest charging network behind it, the most aggressive over-the-air software updates, and a starting price under forty thousand dollars in the United States after the 2025 cuts. Edmunds and Kelley Blue Book rank it competitive on range against the Hyundai Ioniq 5 and decisively cheaper than the Mercedes EQS or the Porsche Taycan, both of which start above one hundred thousand dollars.<\/p>\n

Tesla also lost roughly thirty-five percent of its brand value in 2025, according to Interbrand. Brand Finance recorded a twenty-six percent decline the year before. The company fell from first to fourth in automotive brand rankings, surpassed by Toyota, Mercedes, and Hyundai. YouGov’s UK BrandIndex score for Tesla collapsed from 7.6 in April 2022 to negative 4.2 in January 2023, and was still negative 2.6 in April 2024. The Axios Harris Poll 100 ranked Tesla number eight in 2021 and number eighty-three by 2025.<\/p>\n

In January 2025, Tesla registrations in Germany fell 59.5 percent year over year while the broader German EV market grew 53.5 percent in the same period. BYD overtook Tesla in global pure-electric vehicle sales in 2024 and finished 2025 with roughly a four-hundred-thousand-unit lead.<\/p>\n

Tesla spent roughly one hundred fifty-five million dollars on marketing in fiscal 2024. Two-tenths of one percent of revenue. General Motors and Ford spent approximately two percent. BMW, Mercedes, and Toyota each spend between two and five billion dollars annually on awareness work.<\/p>\n

Elon posted “Product > Brand” to the platform he owns, on the same week the brand-tracking data above was already public.<\/p>\n

The charitable read is that the founder genuinely believes what he posted. Tesla loyalty is still the highest in the auto industry, with roughly seventy percent of Tesla owners buying another Tesla, according to S&P Global Mobility. The Supercharger network is the most valuable piece of physical infrastructure any car company owns.<\/p>\n

The data still does not support the slogan. Range improved. Charging got faster. Software got smarter. The brand collapsed anyway, on every tracker that exists.<\/p>\n

The product is one signal in the stack. The founder is another. The advertising spend is another. The handling of the X acquisition, the political alignment, the labour disputes, and the tweet behaviour. Each is a costly, observable signal. Tesla’s signal stack has lost coherence even as the product has improved. The slogan the founder posted is precisely the kind of legible answer that lets the company avoid auditing the illegible work that actually matters.<\/p>\n

Harvey AI and the Toaster Move<\/strong><\/p>\n

The cleanest counter-example is Harvey AI. Harvey was founded in 2022 by Winston Weinberg, a former attorney at O’Melveny and Myers, and Gabe Pereyra, a former research scientist at DeepMind and Google Brain. The underlying language models Harvey uses are the same models any competitor could access. OpenAI. Anthropic. Google. Nothing about the model layer is proprietary.<\/p>\n

In early 2023, when Harvey was four people working out of an Airbnb, the company signed Allen and Overy as its first major customer. The deployment covered three thousand five hundred lawyers across forty-three jurisdictions. By August 2025, Harvey reached one hundred million dollars in annual recurring revenue, thirty-six months from founding, with forty-five of the AmLaw 100 as customers. The March 2026 funding round valued the company at eleven billion dollars.<\/p>\n

The stack Harvey built was not a better model. The stack was founder credibility, the elite first customer, custom-trained legal embeddings on more than twenty billion tokens of legal text, premium seat pricing, security and audit certifications, a published partnership with Allen and Overy on agentic tools, and a refusal to compete on price.<\/p>\n

What Harvey sold to the legal industry is what Stevenson calls a toaster. A toaster is a crisp, concept-owning noun. Customers know what it is for and when to reach for it. They do not have to assemble the use case themselves.<\/p>\n

The horizontal AI competitors are selling material heaters. A material heater is an abstract capability description that sounds versatile but lands nowhere. Heats diverse materials with configurable parameters. Generates content across multiple domains with adjustable tone. Nobody buys a material heater because nobody knows what it is for in their specific job.<\/p>\n

A lawyer who pays Harvey is not buying a model. The lawyer is buying every signal in the stack. The pedigree of the founders. The fact that Allen and Overy signed first. The legal-specific embeddings. The premium pricing that reinforces premium positioning. The security certifications that satisfy risk committees. Each signal was independently observable to a buying committee at the next firm in the prestige hierarchy. Together, the signals composed a brand that the next firm could trust.<\/p>\n

The Stanford RegLab study published in May 2024 found that Lexis Plus AI hallucinates on more than seventeen percent of legal queries, Westlaw AI on roughly thirty-three percent, and baseline GPT-4 on forty-three percent. The differences are not in the underlying model. The differences are in the workflow architecture, the training data, and the verification steps. The vertical AI companies that win are not winning because they have better models. They are winning because they made architectural commitments the horizontal alternatives cannot match without abandoning their position.<\/p>\n

Harvey is what woodworking looks like.<\/p>\n

Finally<\/strong><\/p>\n

The diagnostic that follows from the frame is unglamorous and operational.<\/p>\n

Audit your own signal stack. List every observable, costly thing your company does. The product. The service. The founder’s public conduct. The pricing. The hiring. The customer roster. The way you handled the last public mistake. The decisions about who to chase and who to walk away from.<\/p>\n

For each signal, ask whether it is coherent with every other signal, and whether it would be credible to the most skeptical professional buyer in your category. Ask which level of the canvas the signal lives at. Saying It <\/em>signals are legible and cheap. Being It <\/em>signals are illegible and expensive. The companies winning are spending most of their effort on the bottom two levels and letting the top two emerge from the work.<\/p>\n

If any two signals contradict each other on a buyer-visible dimension, the contradiction is your positioning problem. The contradiction is what the buyer is reading. The slogan is what you tell yourself to avoid the audit.<\/p>\n

The argument over product and brand is the wrong axis. The question is what your signal stack reads like to a buyer who has been burned before, what the legible parts of your stack are doing versus what the illegible parts are doing, and whether the position you claim on Level 1 is built or just announced.<\/p>\n

If your competitive advantage can be summarized in one clean sentence, you probably do not have one. The illegibility is the defence.<\/p>\n

DIGEST<\/strong><\/p>\n

Memo<\/em><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"

\u265b 26.05.2026 The Best Product Does Not Win Read Time: 6\u00a0Minutes Hello Before we get into today’s topic, two pieces from this week are worth a few minutes of your time. The first is a positioning read of the SpaceX S-1, three weeks before the largest IPO in history. Five customer bases describing five different […]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_coblocks_attr":"","_coblocks_dimensions":"","_coblocks_responsive_height":"","_coblocks_accordion_ie_support":"","_jetpack_feature_clip_id":0,"_jetpack_memberships_contains_paid_content":false,"footnotes":"","jetpack_publicize_message":"{title}\n\n{excerpt}\n\n{url}","jetpack_publicize_feature_enabled":true,"jetpack_social_post_already_shared":true,"jetpack_social_options":{"image_generator_settings":{"template":"highway","default_image_id":0,"font":"","enabled":false},"version":2},"_wpas_customize_per_network":false,"rank_math_title":"","rank_math_description":"","rank_math_canonical_url":"","rank_math_focus_keyword":""},"categories":[91],"tags":[],"class_list":["post-90051","post","type-post","status-publish","format-standard","hentry","category-digest-archive"],"jetpack_publicize_connections":[],"jetpack_sharing_enabled":true,"jetpack_likes_enabled":true,"jetpack-related-posts":[],"jetpack_featured_media_url":"","_links":{"self":[{"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/posts\/90051","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/comments?post=90051"}],"version-history":[{"count":0,"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/posts\/90051\/revisions"}],"wp:attachment":[{"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/media?parent=90051"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/categories?post=90051"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/paulsyng.com\/blog\/wp-json\/wp\/v2\/tags?post=90051"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}