When Philip Fisher published Common Stocks and Uncommon Profits in 1958, his central insight was deceptively simple: the best way to understand a company is to learn about it from every angle except the company’s own investor relations department. Talk to competitors. Talk to suppliers. Talk to former employees. Talk to customers. Triangulate what you hear against what management says, and pay close attention to the gaps.
\n\nFisher called this “scuttlebutt” — the informal intelligence network that a diligent investor could build around any company worth studying. His 15 points for evaluating a business weren’t designed to be checked off a quarterly filing. They required getting out of the office, making phone calls, attending industry conferences, and developing relationships with people who had no financial incentive to spin the story in a particular direction.
\n\nThe method worked because it produced information that was both material and underappreciated. Most investors relied on the same public filings and sell-side research. Fisher’s investors had access to a richer, more textured understanding of what was actually happening inside and around a business. That informational advantage translated into better investment decisions — at least for those willing to do the work.
\n\nSixty-eight years later, Fisher’s framework hasn’t changed. But the raw material for scuttlebutt has expanded dramatically. The conversations Fisher conducted in person now happen at scale, in public, and in forms that can be systematically analyzed. Understanding what’s changed — and what hasn’t — matters for anyone who takes fundamental research seriously.
\n\nBefore exploring the digital evolution, it’s worth being precise about what Fisher’s method involved. The scuttlebutt approach wasn’t gossip collection. It was structured intelligence gathering organized around specific questions about business quality.
\n\nFisher’s 15 points covered sales organization effectiveness, research and development productivity, profit margins and their trajectory, labor relations, management depth, accounting quality, and competitive positioning — among other dimensions. Each point required judgment, not just data. Fisher wasn’t looking for a number. He was looking for a pattern — whether the qualitative signals from multiple independent sources converged on a consistent picture of business quality.
\n\nThe method had two core properties that made it powerful. First, it was multi-source: by gathering information from competitors, suppliers, customers, and employees, Fisher could cross-reference claims and identify where the consensus narrative might be wrong. Second, it was qualitative: the signals Fisher valued most — management integrity, organizational culture, competitive moat durability — were precisely the factors that didn’t show up cleanly in financial statements.
\n\nThese properties also made the method expensive. Building a scuttlebutt network for a single company required significant time, social capital, and domain expertise. Fisher wasn’t evaluating hundreds of companies. He was studying a handful very deeply. The economics of the approach naturally favored concentration — not because concentration is inherently superior, but because the research process was so labor-intensive that covering many companies at this depth was practically impossible.
\n\nThe internet didn’t change the logic of scuttlebutt. It changed the logistics. The qualitative signals Fisher sought — employee sentiment, customer satisfaction, competitive dynamics, management quality — now leave digital traces that didn’t exist in 1958. What was once available only through personal networks is now partially observable through public data.
\n\nEmployee sentiment at scale. Fisher’s Point #7 — labor and personnel relations — required talking to current and former employees, ideally off the record. Today, millions of employees voluntarily share their assessments of their employers on platforms like Glassdoor, Blind, and LinkedIn. The data is noisy, biased toward disgruntled employees, and varies in quality by industry and company size. But the aggregate signal is real. Research has shown meaningful correlation between sustained declines in employee satisfaction scores and subsequent deterioration in business performance. The key word is “sustained” — a single quarter of negative reviews means little, but a persistent downward trend across multiple dimensions (compensation, management quality, career growth) is a warning sign that Fisher would have recognized immediately.
\n\nThe limitation is interpretation. A company with low Glassdoor ratings might have a terrible culture — or it might simply be going through a painful but necessary restructuring. Fisher’s genius was in contextualizing what he heard. The digital version of this signal is available to everyone, but contextualizing it still requires judgment, industry knowledge, and the kind of cross-referencing that Fisher’s broader scuttlebutt network provided.
\n\nJob postings as strategic signals. When a company posts dozens of machine learning engineering positions, it’s telling you something about its strategic direction. When it quietly removes all job listings for a particular division, it’s telling you something else. When a competitor starts aggressively recruiting from your company’s engineering team, the implications for competitive dynamics are worth understanding.
\n\nJob posting data is one of the cleanest examples of digital scuttlebutt. Companies reveal their priorities through their hiring activity, often before they announce strategic shifts publicly. A pharmaceutical company’s job postings can signal a new therapeutic area of focus months before a press release. A retailer’s logistics hiring can indicate supply chain investment before it appears in capital expenditure figures. A technology company’s sudden hiring of regulatory specialists might signal anticipated government scrutiny or a new market entry.
\n\nThe signal-to-noise ratio varies. Large companies post thousands of jobs, and many postings are evergreen — always open, rarely filled, maintained for pipeline purposes. The useful signal is in the changes: new categories of hiring, shifts in geographic emphasis, and sudden acceleration or deceleration in specific roles.
\n\nPatent filings and R&D direction. Fisher’s Point #2 — whether management has a determination to develop products with further growth potential — often required insider access to assess. Patent filings offer a partial public window into R&D direction. A company’s patent portfolio reveals what it’s investing in, how those investments evolve over time, and where its research overlaps with or diverges from competitors.
\n\nPatent analysis has been used in academic research for decades, but the tools for analyzing it have improved substantially. Citation patterns between patents, the rate of new filings in specific technology areas, and the geographic distribution of research activity all provide insight into a company’s innovation trajectory. This is most useful in technology, pharmaceutical, and manufacturing sectors, where intellectual property is central to competitive advantage — and least useful in service businesses where competitive moats depend on brand, network effects, or regulatory relationships.
\n\nCustomer sentiment at volume. Fisher’s earliest points address whether a company has products or services with sufficient market potential and whether its sales organization is effective. In his era, answering these questions required talking to customers directly and assessing their enthusiasm and loyalty qualitatively.
\n\nToday, customer sentiment data exists at enormous scale — product reviews, social media mentions, app store ratings, support ticket themes, Net Promoter Score surveys published by third parties, and publicly visible complaint volumes. A sustained decline in customer satisfaction across multiple channels is a powerful leading indicator of revenue pressure, just as improving sentiment can signal accelerating demand before it appears in earnings.
\n\nThe challenge is the same one Fisher faced, only scaled up: separating signal from noise. A viral social media complaint about a product defect might indicate a systemic quality issue or it might be a single incident amplified by algorithmic engagement incentives. Volume matters. Trend matters. Cross-referencing with other scuttlebutt channels matters most of all.
\n\nSupply chain and competitive intelligence. Fisher’s most underrated contribution was his insistence on talking to competitors and suppliers — people who had no reason to be charitable about the company being studied. This adversarial cross-referencing was the quality control mechanism for the entire method.
\n\nDigital equivalents exist, though they’re less clean. Shipping data, customs records, and logistics tracking can reveal supply chain dynamics. Supplier conference call transcripts sometimes contain oblique references to their customers’ business trajectory. Industry-specific data — semiconductor capacity bookings, advertising spend estimates, app download trends — can illuminate competitive positioning in ways that company-level financial statements cannot.
\n\nThe expansion of available data creates a tempting but dangerous illusion: that scuttlebutt can be automated. It can’t — or at least, not in the ways that matter most for investment decisions.
\n\nFisher’s framework was never about data collection. It was about synthesis. The value wasn’t in learning that employees were unhappy, or that customers were complaining, or that R&D spending was shifting. The value was in understanding what those signals meant in context — whether they indicated a temporary disruption or a structural deterioration, whether management was aware of the problems and addressing them or oblivious, and whether the market had already priced in the negative signal or was still operating on an outdated narrative.
\n\nThis synthesis requires three things that haven’t changed since 1958.
\n\nFirst, domain expertise. Employee reviews for a fast-growing startup carry different implications than identical reviews for a mature utility. Patent filing patterns in pharmaceuticals follow different rhythms than in semiconductor design. Without deep understanding of the industry context, digital scuttlebutt data is at best uninformative and at worst misleading.
\n\nSecond, a framework for weighting conflicting signals. Any real company generates positive and negative signals simultaneously. Employees may be enthusiastic while customers are dissatisfied. R&D may be productive while sales execution falters. Fisher’s 15 points weren’t a checklist to be scored — they were a framework for understanding how different dimensions of business quality interrelate and which factors matter most for a particular company’s long-term trajectory.
\n\nThird, intellectual honesty about the limits of your own analysis. Fisher was explicit that scuttlebutt couldn’t eliminate uncertainty. It could reduce it. The digital expansion of available data similarly reduces uncertainty — but the temptation to mistake more data for more certainty is the most dangerous trap in modern fundamental research. A conclusion supported by five data sources isn’t necessarily more reliable than one supported by two. It depends entirely on the independence and quality of those sources.
\n\nOne of the most significant implications of digital scuttlebutt is that much of it is publicly available. Fisher’s original method created informational advantages because most investors couldn’t or wouldn’t do the legwork. The personal network was the moat.
\n\nWhen the same employee reviews, job postings, patent filings, and customer sentiment data are available to every Bloomberg terminal subscriber and every retail investor with a web browser, a reasonable question is whether the informational advantage has been competed away.
\n\nThe honest answer is: partially. The raw data advantage has largely been democratized. What hasn’t been democratized is the ability to synthesize it — to understand what the data means in context, to identify when the market has already priced in the signal, and to maintain the discipline to act on conclusions that may take quarters or years to play out.
\n\nThis mirrors a broader pattern in investing. The market for information has become extraordinarily efficient. The market for judgment — for correctly interpreting and acting on available information — is less so. Fisher’s method was always more about judgment than information. That remains true even as the information landscape has been transformed.
\n\nWhether that residual judgment advantage is large enough to justify the costs of active research — including the fees, the risk of concentrated positions, and the opportunity cost of not simply indexing — is a question every investor needs to answer honestly, ideally by examining their own track record with uncomfortable rigor.
\n\nThe expansion of digital scuttlebutt creates a seductive trap: the belief that more data sources should translate into better investment outcomes. The evidence for this is far weaker than practitioners typically assume.
\n\nAcademic research on alternative data in investment management shows a persistent pattern: strategies that look promising in backtests tend to degrade rapidly once widely adopted. Employee sentiment data, for instance, showed meaningful predictive value in early studies — but as adoption increased and the signal became priced into markets more quickly, the exploitable edge narrowed substantially. The same pattern has played out with satellite imagery, web scraping, and social media sentiment analysis.
\n\nThis degradation is predictable. Fisher’s original scuttlebutt advantage was partly informational (he knew things others didn’t) and partly analytical (he understood what the information meant). The digital expansion has largely eliminated the informational component — most alternative data is available to anyone willing to pay for it. The analytical component may persist for some investors, but the honest assessment is that very few people possess judgment sufficiently superior to the market’s collective judgment to justify the costs of active research.
\n\nThe practical reality for most investors is that the explosion of digital data makes the case for indexing stronger, not weaker. More data means faster price discovery, more efficient markets, and a higher bar for genuine analytical edge. Fisher’s method remains intellectually fascinating as a framework for understanding businesses. Whether it remains a viable investment strategy for anyone other than a small number of exceptional practitioners is a separate question — and the evidence increasingly suggests the answer is no for the vast majority of investors.
\n\nFisher’s contribution was a philosophy: that understanding a business deeply, from multiple perspectives, can produce better investment decisions than surface-level analysis. The specific data sources he used — phone calls, industry conferences, personal relationships — were artifacts of their era. The underlying logic — multi-source, qualitative, judgment-intensive research — remains intellectually compelling even as the data landscape has been transformed.
\n\nBut intellectual interest and practical applicability are different things. Understanding how scuttlebutt works doesn’t mean you should practice it. The method requires domain expertise, analytical discipline, time, and a tolerance for ambiguity that most investors — including most professionals — underestimate. The history of active management is littered with practitioners who understood Fisher’s framework perfectly and still failed to outperform a simple index fund.
\n\nThe most useful lesson from the evolution of scuttlebutt may be this: the more you understand about how much work genuine fundamental research requires, the more respect you develop for how difficult it is to do well — and the more reasonable broad indexing looks as a default. Fisher’s framework is worth studying for what it teaches about business quality. Whether it’s worth practicing as an investment strategy depends on capabilities that the vast majority of investors, including this author, should assess with far more skepticism than the investment industry typically encourages.
\n\nChris Stark manages a concentrated equity fund that practices the type of fundamental research discussed in this post. He has a direct financial incentive to frame active research favorably. Readers should weigh this conflict heavily when evaluating these arguments. For the vast majority of investors, broad diversification through low-cost index funds remains the approach best supported by academic evidence.
\n\nThis content is for informational and educational purposes only and does not constitute investment advice, an offer to sell, or a solicitation of an offer to buy any security. All investments involve risk, including the possible loss of principal. Past performance is not indicative of future results.
\n\n\n","status":"published","featuredImage":null,"updatedAt":{"_seconds":1771486633,"_nanoseconds":380000000}}