Marketing Insights
Two marketing leaders. Same company stage, similar market, comparable budget. One spent a decade building startup marketing functions from scratch. The other led demand generation at two enterprise software companies before making the jump.
Two marketing leaders. Same company stage, similar market, comparable budget. One spent a decade building startup marketing functions from scratch. The other led demand generation at two enterprise software companies before making the jump.
You give them the same brief: "Build our first content marketing programme."
The startup-native leader sits down and asks: "What can I actually do myself? What do I know about our buyers? Who in my network could help?" Within a week, they've published two LinkedIn posts, launched a scrappy blog, and started measuring what gets engagement. No strategy deck. No editorial calendar. Just action and adjustment.
The corporate-transplant leader opens a spreadsheet. They map the competitive environment, build a channel-by-channel analysis, draft a three-month editorial plan, and present it to leadership for sign-off. It's thorough, well-structured, and takes three weeks before anything gets published.
Neither is wrong. But the cognitive processes driving each approach are fundamentally different. The differences go deeper than you might expect. When I decomposed the decision-making of thirteen marketing leaders into individual cognitive processes, I found eighty-four distinct types distributed across five domains. The distribution wasn't random. It followed developmental context. The structural environments in which leaders built their careers reshape how we should think about hiring, team composition, and what "marketing capability" actually means at the individual level.
Key Concepts Introduced in This Article
Cognitive Heterogeneity: An established concept in management research (e.g., Bhansing et al., 2012; Helfat & Peteraf, 2015), applied here to marketing leadership by developmental context. Systematic differences in cognitive processes; how leaders perceive resources, formulate problems, search for solutions, and evaluate outcomes; between marketing leaders who developed in different structural environments. Not personality types. Not management styles. Patterns of cognition shaped by career context.
Startup-Native (SN): A marketing leader whose career developed primarily in startup or early-stage environments. Characterised by identity-anchored means assessment, experiential search as the default learning mode, affordable loss reasoning, and rapid prototype action.
Corporate-Transplant (CT): A marketing leader with significant experience in corporate or enterprise environments before joining a startup or constrained context. Characterised by structure-based means assessment, sustained cognitive search capability, analytical problem formulation, and systematic evaluation.
Experiential Search: Learning by doing: iterating from direct experience, adjusting based on real-time feedback, building knowledge through action rather than analysis (Gavetti & Levinthal, 2000). The dominant search mode for twelve of thirteen participants in this research.
Cognitive Search: Learning from mental models: hypothesis-driven analysis, structured competitive reasoning, deliberate evaluation against predetermined frameworks (Gavetti & Levinthal, 2000). Sustained cognitive search was observed in five of thirteen participants, four of whom were Corporate Transplants.
The research identified consistent cognitive differences between leaders who built their careers in startup environments (Startup Natives) and those who transferred from corporate or enterprise contexts (Corporate Transplants). These aren't personality types or management styles. They're patterns of cognition. How leaders perceive resources, formulate problems, search for solutions, and evaluate outcomes. These patterns are shaped by the structural environments they developed in.
This matters because most hiring conversations frame the choice as "startup experience vs. corporate experience." It's as though the difference is in what leaders know. The data suggests the deeper difference is in how they think. And that's harder to see in an interview, harder to manage once hired, and more consequential for what happens in the first six months.
When a Startup Native assesses their means, they start from constraint. The instinct is to map limitations before possibilities. They anchor that assessment in personal identity rather than structural opportunity.
Consider one startup-native leader. The sole marketing person at a fifteen-person software company with a monthly budget closer to zero than to ten thousand. Their first move wasn't to request more resources or draft a channel strategy. It was to take stock of themselves. "What is my skill set? What can I do reasonably on my own without additional help? And then what's the least expensive way to do all the other things that I probably can't reasonably do?" That self-audit led to a specific set of decisions. They taught themselves web development through ChatGPT. "Challenging but fun," they called it. They handled SEO research, content creation, and site management personally. For the one capability they genuinely couldn't build themselves, link building, they hired a contractor at a thousand dollars. The means audit here wasn't a spreadsheet exercise. It was an identity-grounded assessment of what one person could actually do. This produced a resource map where personal curiosity and willingness to learn counted as deployable assets.
This identity-anchored assessment produces a specific downstream effect. Startup Natives formulate problems in terms of what's buildable with existing means, not what's ideal in the abstract. Their problem formulation favours tangible prototyping: "let's try this and see." It favours this over analytical decomposition. Another startup-native leader, a solo founder operating on one to two thousand euros a month, constructed a complete video production capability from scratch. They wrote the scripts themselves, bought a subscription to an AI voiceover platform, and hired four freelance video editors, giving each one script to see who produced the best result. Total cost: roughly one hundred euros per video, against an industry norm of five to ten thousand for equivalent production. The quality wasn't perfect. But the capability was real: repeatable video production built from the combination of founder writing skill, AI voice generation, and freelancer arbitrage. No single component would have worked alone. The value emerged from the bundle.
The search mode is predominantly experiential: learning by doing, adjusting based on feedback, iterating in real time. Across the full dataset, twelve of thirteen participants primarily rely on experiential search as their dominant learning mode. Five of those thirteen also demonstrate cognitive search capabilities alongside the experiential default. For Startup Natives, experiential search isn't a fallback. It's the default cognitive mode. As that solo founder put it when explaining why they chose low-cost experimentation over benchmarks: "Benchmarks could also habe some bias, either intended or unintended. Experimenting yourself with low-cost campaigns gives you raw data which is irreplaceable." They try things, watch what happens, and refine. Formal hypothesis-testing and structured competitive analysis are rare. Not because these leaders lack analytical capability, but because their developmental environments rewarded rapid action over deliberate planning.
The strength of this pattern is speed, adaptability, and realistic resource expectations. The risk is under-investment in planning, potential blind spots from narrow information ecosystems, and a tendency toward what one participant called "living in a bubble." The information sources you rely on, primarily LinkedIn, peer networks, and personal experience, reflect your existing assumptions back at you without challenge. This is a specific form of the experience paradox. The very information ecosystem that enabled rapid learning in one context becomes a constraint on learning in the next.
Corporate Transplants assess means differently. Where Startup Natives start with personal capability, Corporate Transplants start with structural opportunity. Their audit language is about budget reallocation, market sizing, and competitive positioning.
One corporate-transplant leader illustrates the pattern vividly. With a ten-person marketing team and a budget between ten and one hundred thousand euros monthly, their approach was architecturally different from the SN leaders described above. They maintained a rolling five-quarter plan. "I'm always looking at the data, so I want to know what marketing is going to be measured on come the end of the year, how we're supporting the commercial function." When a new opportunity appeared mid-cycle, the response wasn't to prototype. It was to restructure: "What is a nice to have and if we are going to execute on this because it's really a great opportunity, what am I dropping or what needs to be redeployed in order to take advantage of that?" The logic is structural. Evaluate against existing commitments, reallocate resources, document the trade-off. Another participant describes a three-tier triage that's a miniature corporate resource management framework: "Reprioritise our own bucket, look for that wallet from someone else, do low-hanging stuff which we can still do in house."
Problem formulation follows a different track. Where Startup Natives prototype, Corporate Transplants analyse. That same leader applies Pareto reasoning instinctively: "What are the three activities that are going to give us 80% of what it is that we need?" They bring explicit structured methodology. "Build, measure, learn. I come from a technical background. Agile marketing approach." But note that the methodology comes from a technical background, not from startup iteration. The build-measure-learn loop here is a deliberate analytical framework imported from software development, not the organic experimentation-by-necessity that characterises the SN pattern.
The Corporate Transplant search mode includes a cognitive component that's largely absent from the Startup Native pattern. Only five of thirteen participants show sustained cognitive search: comparative analysis, market research, deliberate competitive reasoning. Four of those five are Corporate Transplants. This isn't because Corporate Transplants are "more analytical." It's because they developed in environments where structured analysis was rewarded, resourced, and expected. They have cognitive infrastructure for hypothesis-testing that was built through years of operating in contexts that supported it. The relationship between developmental context and search mode is one of the most consistent findings in the data.
The strength of this pattern is systematic analysis, stakeholder communication, and the ability to build structured processes that scale. The risk is overconfidence in pattern transfer. This is the very phenomenon explored in the Experience Paradox. The analytical frameworks that worked in a five-hundred-person company with a twenty-person marketing team don't port directly to a context where you are the entire team. Corporate Transplants can be slower to calibrate: spending weeks on a strategy deck while the Startup Native has already shipped, learned, and iterated.
There's a specific risk around brand conviction that surfaces in the data. Corporate Transplants tend to show stronger explicit brand conviction. That same leader states: "Brand is key in terms of sustainability. What's going to distinguish or separate us is brand." This is sophisticated strategic thinking. But it creates a particular vulnerability when they enter startup contexts where brand investment is structurally infeasible. As they put it: "The minute you speak to anyone outside of marketing about brand, they glaze over, they check out emotionally." So their pragmatic response is to build the commercial engine first and defer brand. They know the whole time that brand is what matters for long-term sustainability. The knowledge is correct. The environment won't accommodate it. And as the Metacognitive Paradox predicts, knowing the right answer doesn't help when the structure prevents acting on it. This frustration is also a form of structural causation. The environment shapes what's possible regardless of individual capability.
One participant occupies a distinct third position: the Advisory profile. Rather than identity-based or structure-based means assessment, this participant takes a research-based approach. "Really start looking in depth on research. Qualitative and quantitative. Paid research and also our own desktop research, AI research." The Advisory pattern offers proof-of-concept emergence as a strategic approach: grow proof points through low-budget execution until results attract formal resource allocation. This is a hybrid logic. Effectual in its means-driven starting point, but systematic in its evidence-gathering orientation. With a single participant, this is an observation, not a pattern claim. But it suggests the SN/CT binary may be a simplification of a broader cognitive spectrum.
The research produced a granular map of how these cognitive differences manifest. One hundred coded passages across all thirteen participants were decomposed into eighty-four distinct cognitive process types, organised across five domains.
Means assessment (12 types). How leaders evaluate what they have to work with. SN participants emphasise constraint recognition and personal capability audit. CT participants emphasise budget structures and market opportunity sizing. Both identify network resources, but SN leaders see networks as capability extensions: "who can help me do this?" CT leaders see them as information channels: "what does the market data say?" The difference in how networks function for each cohort is explored in depth in the network-mediated capability analysis.
Problem formulation (18 types). How leaders frame what needs solving. This is where the SN/CT divergence is particularly visible. SN leaders formulate problems as design challenges: "given what I have, what can I build?" CT leaders formulate them as analytical puzzles: "given the market conditions, what should we prioritise?" The former produces faster time-to-action. The latter produces more thorough strategic rationale. The data doesn't measure which produces better outcomes. It shows only that the cognitive approaches are structurally different and carry different implications for speed and rigour.
Constraint reasoning (15 types). How leaders work within and around limitations. One of the distinctive SN capabilities is affordable loss calculation. This means determining what you can afford to lose rather than what you expect to gain. Three participants demonstrate this explicitly. One described scraping together funds from different budget lines: "came up with a figure that we were willing to lose and basically bet that on initiating some marketing efforts." Another frames every decision through minimum viable cost: "What's the least expensive way to do all the other things that I probably can't reasonably do?" A third evaluated an affiliate marketing platform at twelve thousand dollars annually and passed. Not because they lacked the budget, but because "the ROI is not clear enough for me to justify the cost" in a domain where they had no prior experience to anchor assumptions. All three are Startup Natives. The affordable loss heuristic, a core element of effectuation theory (Sarasvathy, 2001), appears to be SN-characteristic in this dataset.
Decision-making under uncertainty (22 types). The largest domain, reflecting the centrality of uncertainty in startup marketing. Both cohorts make decisions under uncertainty, but they manage that uncertainty differently. SN leaders tend toward rapid experimentation and feedback-driven adjustment. CT leaders tend toward comparative reasoning and structured evaluation. Neither approach eliminates uncertainty. They manage it through different cognitive strategies, with different implications for speed, accuracy, and the types of errors they're likely to make.
Prototype action (17 types). How leaders test ideas in practice. SN leaders prototype through direct action: building, publishing, launching, and measuring. CT leaders tend toward more structured approaches with defined evaluation criteria. The SN approach generates faster feedback but risks premature termination. One participant describes killing an ad campaign after a single week because it "didn't deliver." The CT approach generates more systematic evaluation but risks slower time-to-action. It takes longer to design the test while the market window narrows. These different prototype rhythms directly shape how marketing routines emerge or fail to.
The most important thing about cognitive heterogeneity is what it isn't: a performance ranking.
The patterns described above aren't "startup native is better at startups" or "corporate transplant is better at scale." They're descriptions of different cognitive adaptations to different structural environments. SN patterns emerge from resource scarcity, network constraints, and the requirement for rapid feedback. CT patterns emerge from resource access, institutional infrastructure, and the requirement for stakeholder management. Both are rational responses to the environments that shaped them.
They become problematic only when the environment changes. The Corporate Transplant who applies enterprise analytical frameworks to a three-person startup will over-invest in planning and under-invest in action. The Startup Native who applies scrappy improvisation to a company that's just raised Series B and needs scalable systems will under-invest in structure and over-rely on personal heroics. The pattern that made you successful in one context is the pattern that will trip you up in another.
This is why hiring for "startup experience" or "corporate experience" misses the point. You're not hiring a resume. You're hiring a cognitive architecture. And the value of that architecture depends entirely on what your company needs right now and what correction mechanisms you can build around it.
To prevent confusion, it's worth being explicit about what cognitive heterogeneity is not. The world of people frameworks is crowded, and the differences matter.
It's not personality typing. MBTI, DISC, and similar instruments classify people by innate preferences. Cognitive heterogeneity describes developed cognitive patterns. They're shaped by structural environments over years of professional practice. A Startup Native doesn't think in means-driven, prototype-oriented ways because of an innate disposition. They think that way because that's the cognitive architecture their career environment built. The same person, developed in a corporate context, would show different cognitive patterns. The cause is developmental context, not personality.
It's not a performance ranking. The SN pattern isn't "better for startups" and the CT pattern isn't "better for corporations." Both patterns carry strengths and risks that are context-dependent. An SN leader in a scaling company may under-invest in structure. A CT leader in an early-stage company may over-invest in analysis. The question isn't who's better. It's which cognitive architecture fits the current structural conditions.
It's not management style. Transformational versus transactional leadership, servant leadership, situational leadership. These describe how leaders interact with teams. Cognitive heterogeneity describes how leaders think before they interact. How they perceive resources, formulate problems, search for solutions, and evaluate outcomes. Two leaders with identical management styles can have fundamentally different cognitive architectures.
It's not the "startup experience vs. corporate experience" hiring heuristic. That heuristic treats experience as a knowledge asset. What do they know how to do? Cognitive heterogeneity treats developmental context as a cognitive shaping force. How do they think about what to do? The distinction matters because it changes what you screen for. Knowledge can be transferred through briefing. Cognitive architecture can't be.
For team composition. SN + CT complementarity can be powerful, but only if managed deliberately. The risk is that different cognitive patterns produce misalignment rather than synergy if neither party recognises the other's approach as a legitimate adaptation rather than a deficiency. Productive complementarity requires explicit acknowledgment that both patterns are rational cognitive responses to different structural environments. It requires deliberate role design that leverages each pattern's strengths. A practical starting point: have each team member map their own default pattern across the five domains (means assessment, problem formulation, constraint reasoning, uncertainty management, prototype action). Where the team clusters in one pattern, that's a blind spot. Where patterns diverge, that's either a source of creative tension or a source of friction. The difference depends on whether the divergence is acknowledged.
For self-awareness. Knowing your cognitive profile is genuinely useful here. Not because awareness corrects the bias (the Metacognitive Paradox still applies), but because it helps you build correction mechanisms. If you're an SN leader who knows you under-invest in planning, hiring a CT-pattern operator for your team isn't admitting weakness. It's designing around a known cognitive tendency. If you're a CT leader who knows you over-invest in analysis, building forced launch deadlines into your process isn't abandoning rigour. It's correcting for a structural pattern. A diagnostic question: when you face a new marketing challenge, do you instinctively reach for your means (what can I do right now?) or your frameworks (what should the analysis look like?)? The answer tells you your default search mode. It points to where your correction mechanisms need to be.
For hiring. Match the cognitive profile to the company's current structural conditions, not just the job description. A company operating under acute resource constraint needs someone who can operate with SN-pattern cognition. Means-driven, prototype-oriented, comfortable with affordable loss. A company that has secured resources and needs scalable systems benefits from CT-pattern cognition. Structured, analytical, capable of building processes that survive beyond any individual. The best hire isn't the "best marketer." It's the marketer whose cognitive architecture fits the structural environment they'll actually operate in. In the interview, try a constraint scenario: "You've just joined and have no budget for the next quarter. Walk me through your first two weeks." An SN-pattern candidate will describe action: what they'd build, test, and ship. A CT-pattern candidate will describe analysis: what they'd evaluate, plan, and propose. Neither answer is wrong. But one of them fits your context better than the other.
The cognitive heterogeneity finding sits at the intersection of several theoretical traditions that rarely speak to each other. The closest ancestor is the career imprinting literature (Higgins, 2005; Marquis & Tilcsik, 2013), which establishes that work experiences leave durable imprints on capabilities, connections, confidence, and cognition. The mechanism described in this research, developmental context shaping cognitive patterns, has clear lineage in that tradition. What the present research adds is a marketing-specific, micro-foundational operationalisation. Not just that imprinting happens, but exactly which cognitive processes vary, across which domains, and with what consequences for capability building under constraint. Three further traditions provide the analytical scaffolding.
Micro-foundations movement (Felin, Foss & Ployhart, 2015). The central argument of micro-foundations research is that organisational capabilities are not monolithic. They're composed of individual-level cognitive processes, social interactions, and structural enablers. The cognitive heterogeneity finding demonstrates this concretely: the same organisational capability (marketing under constraint) is composed of different cognitive processes depending on the individual performing it. Two leaders building "content marketing capability" are doing cognitively different things. The organisational outcome looks similar. The micro-foundational mechanisms producing it are distinct. This appears to be an uncommon empirical contribution: systematic cognitive heterogeneity in marketing micro-foundations, mapped by developmental context. Adjacent literatures on career imprinting (Higgins, 2005; Marquis & Tilcsik, 2013), managerial cognition, and effectuation have explored related ground. But the specific intersection documented here, cognitive process variation across marketing leaders shaped by startup versus corporate development tracks, is under-explored.
This has implications for how we think about capability transfer and replication. If capabilities are composed of individual cognitive processes, and those processes vary systematically by developmental context, then transplanting a person from one context to another doesn't transplant the capability. It transplants a different cognitive approach to producing capability. The capability that emerges in the new context will be shaped by the individual's cognitive architecture, not just by the organisational environment. This is why "hiring someone who's done it before" doesn't guarantee the same outcome. They've done it before. But the "it" they did was cognitively different from the "it" you need done now. The Effectual Orchestration framework provides the mechanism through which these different cognitive approaches produce different capability trajectories.
Managerial cognitive capabilities (Helfat & Peteraf, 2015). The distinction between SN and CT patterns maps directly onto Helfat and Peteraf's framework for managerial cognitive capabilities: the mental activities through which managers build, integrate, and reconfigure organisational resources. The data shows that these mental activities aren't generic. They're calibrated to specific environmental conditions. SN leaders have built cognitive capabilities for rapid-cycle means assessment and prototype action. CT leaders have built cognitive capabilities for structured analysis and stakeholder communication. Both are genuine managerial cognitive capabilities. They're just different ones.
Cognitive and experiential search (Gavetti & Levinthal, 2000). The search mode finding, twelve of thirteen participants primarily use experiential search, with cognitive search concentrated in Corporate Transplants, connects to Gavetti and Levinthal's fundamental distinction between learning from direct experience and learning from mental models. The data suggests this isn't a fixed trait but a developed capability. CT leaders show more cognitive search not because they're inherently more analytical, but because they developed in environments that resourced and rewarded structured analysis. The five participants who show sustained cognitive search (four CT, one SN) aren't cognitively superior. They're cognitively differently equipped based on developmental context. The full implications of this search-mode distribution are explored in Cognitive and Experiential Search.
The cognitive heterogeneity finding is the most empirically grounded element of the research programme:
Three empirical features are worth highlighting.
First, cross-domain consistency. The SN/CT patterns identified in means assessment reappear in structuring, bundling, and leveraging. A leader who conducts identity-based means audits tends to formulate problems as design challenges and leverage through rapid prototyping. A leader who conducts structure-based audits tends to formulate problems as analytical puzzles and leverage through systematised processes. The coherence across domains suggests these aren't domain-specific skills but general cognitive orientations. The specific downstream manifestations vary by individual and context rather than following a single deterministic chain.
Second, effectuation theory alignment. The SN pattern maps more closely to effectual logic (Sarasvathy, 2001): means-driven, affordable-loss-oriented, partnership-building, surprise-leveraging. The CT pattern includes more elements of causal logic: goal-driven, expected-return-oriented, competitive-analysis-based, prediction-seeking. This doesn't mean SN leaders are "effectual" and CT leaders are "causal." Both use both logics. But the emphasis differs. The default logic under pressure differs. When constraint intensifies, SN leaders default to effectual reasoning. CT leaders show more causal reasoning elements. The data suggests both cohorts ultimately operate in a mixed-logic space rather than occupying pure positions on the effectuation-causation spectrum.
Third, attention allocation patterns. Strong empirical support for Ocasio's (1997) attention-based view. SN leaders allocate attention through indicator chains: one participant focuses on "website activity, opportunities, revenue. Those are the three that I focus on." CT leaders allocate attention through measurement frameworks. Another participant specifies: "I want to know what marketing is going to be measured on come the end of the year." The SN pattern is activity-driven: tracking what's moving in real time. The CT pattern is framework-driven: defining what should be watched against predetermined criteria. Both are functional. They produce different blind spots.
A skill category emerged prominently in this data that the broader management literature has only recently begun to theorise: AI-augmented individual capability. Three participants, all Startup Natives, demonstrate capabilities that are fundamentally constituted by human-AI collaboration. One uses ChatGPT for web development learning, turning a knowledge gap into a deployable skill through AI-assisted self-teaching. Another leverages AI tools to replace designer and coder roles entirely, producing complete campaigns as a solo operator. A third uses AI-generated voiceover in video production, building a repeatable content capability at roughly one percent of the industry cost.
The concentration in SN participants is notable but not conclusive. It may reflect the SN tendency toward individual DIY orientation. If you're accustomed to doing everything yourself, AI is a natural force multiplier for individual capability. No CT participants in this dataset described equivalent AI-augmented individual capability. But the data doesn't tell us whether CT leaders integrate AI differently or simply didn't surface it in interviews. This is an observation requiring further investigation, not a confirmed finding.
What the data suggests is that the micro-foundations framework may need updating. The individual-level cognitive processes that compose organisational capabilities now include human-AI collaborative processes. The "individual" in micro-foundations is no longer a solo human agent. It's a human-AI system. This changes what "individual capability" means and how we think about capability transfer, fragility, and development. The implications for capability fragility, the dependency on tools you don't control, are explored in Marketing Routines Under Constraint.
If marketing leaders think differently based on their developmental context, and those cognitive differences shape how capabilities emerge, then the capabilities themselves will look different depending on who builds them. The next post examines this directly: how marketing routines emerge, persist, and fragment under constraint. It explores why the process through which a capability becomes a stable organisational routine is more fragile, more path-dependent, and more interesting than any strategy framework suggests.
Next in the series: Marketing Routines Under Constraint --- How Capabilities Emerge, Persist, and Fragment
This post is part of a 10-part foundation series exploring how marketing capabilities emerge under constraint. The cognitive heterogeneity framework draws on micro-foundations theory (Felin, Foss & Ployhart, 2015), managerial cognitive capabilities (Helfat & Peteraf, 2015), and cognitive/experiential search (Gavetti & Levinthal, 2000), grounded in original empirical research with 13 B2B SaaS marketing leaders. Browse all pillars.
