Section 1
The five challenges at a glance
Diversification at small scale fails through five mechanisms: management bandwidth splits before revenue does; capability and brand dilute; capital that should compound in the core is scattered across experiments; the second business cross-subsidizes its losses invisibly; and the firm becomes harder to sell because buyers cannot tell what it is. Each mechanism has an evidence base, some from large-firm research that must be translated carefully to small-firm scale, which the table flags. The deeper analyses follow.
Section 2
Challenge one: the Bain evidence, sustained growth comes from the core
The most comprehensive evidence on focus versus diversification comes from Bain & Company's long-running growth studies, synthesized in Chris Zook and James Allen's Profit from the Core. Drawing on a decade of data covering thousands of companies (the underlying database spans roughly 8,000 firms, with a core ten-year study of about 2,000), the research found that only around one company in nine achieved sustained, profitable growth, typically defined as 5.5%+ real growth in revenues and profits over a decade while earning the cost of capital, and that roughly nine out of ten of those sustained value creators built their performance around one strong, focused core business rather than diversification (Zook & Allen, 2001/2010; Bain & Company). The corollary finding matters just as much: most growth initiatives launched outside the core failed outright or destroyed value, and the companies that expanded successfully did so through 'logical and reinforcing adjacencies', moves one step from the core that reused existing customers, channels, or capabilities. Translating to sub-$10M scale strengthens rather than weakens the conclusion. Large firms diversify with separate management teams, dedicated capital, and portfolio governance; a small firm diversifies with the same founder, the same senior staff, and the same bank account. Every structural buffer that occasionally makes large-firm diversification survivable is absent at small scale. The small firm's version of a growth portfolio is sequential, not parallel: dominate a niche, extract its full potential, then take one adjacent step.
Section 3
Challenge two: the diversification discount, real, contested, and instructive
Corporate-finance research quantified the cost of spread decades ago. Berger and Ofek, comparing diversified firms' actual values against the imputed stand-alone values of their segments, found a 13-15% average value loss from diversification during 1986-1991, driven substantially by overinvestment and cross-subsidization, strong segments' cash funding weak segments' losses (Berger & Ofek, Journal of Financial Economics, 1995). Intellectual honesty requires the caveat: this literature is contested. Later work argues part of the discount reflects selection (already-discounted firms choose to diversify), and recent research contends the standard measurement matches large, old diversified firms against small, young focused ones, manufacturing a discount from valuation mechanics (Campa & Kedia, Journal of Finance, 2002; Hund, Monk & Tice, Critical Finance Review, 2024). What survives the methodological fight is the mechanism, and the mechanism is what small operators should fear: cross-subsidization is invisible without segment-level P&Ls, and overinvestment in the weak line is the default because struggling ventures demand attention and cash in a way that healthy ones do not. A $3M agency running a $400K side product rarely knows the product's true loaded cost, because founder hours, shared staff, and brand equity are never charged to it. The research translation for sub-$10M firms is not 'diversification destroys exactly 14% of value', it is that without ruthless segment accounting, the strong business will quietly finance the weak one until both are mediocre.
Section 4
Challenge three: specialization economics and the reinvention counterargument
The services-specific evidence reinforces focus. Hinge Research Institute's 2025 High Growth Study, vendor research from a marketing firm, flagged as such, found high-growth professional services firms grow roughly four times faster and are about 30% more profitable than peers, with specialization a recurring differentiator: specialists command premium fees, convert referrals at higher rates, and standardize delivery in ways generalists cannot (Hinge, 2025). The honest counterargument comes from the corporate-reinvention literature. PwC's 29th Global CEO Survey reports that 42% of CEOs say their company has begun competing in at least one new sector in the past five years, and that companies deriving substantial revenue from new sectors report stronger margins, PwC calculates a meaningful performance premium for firms that reinvent their business models well (PwC, 2026). Doesn't that argue for diversification? Only with scale-translation discipline. The PwC respondents run companies with dedicated divisional leadership and balance sheets that can absorb failed experiments; their 'new sectors' are usually capability-adjacent moves executed by hired teams. The small-firm equivalent of reinvention is not adding a second business, it is evolving the core: repositioning the same capability for a higher-value problem, productizing the method, or following the niche's changing needs. Zook's later work locates most successful renewal in 'hidden assets' already inside the firm, underused capabilities, customer insight, platform potential, rather than in new markets (Zook, Unstoppable, 2007). For sub-$10M firms, reinvention and focus are the same move, executed in sequence.
Section 5
Innovative solutions: adjacency discipline, hidden assets, and real options
The research suggests three disciplined alternatives to scattershot diversification. First, adjacency mapping, from the Bain playbook: list possible expansion moves and score each by distance from the core, shared customers, shared capabilities, shared channels, shared cost structure. Moves sharing three or four dimensions (a new service for existing clients; the same service for a directly adjacent client type) succeed at far higher rates than two-dimension leaps, and the Bain research found repeatability, using the same expansion formula again and again, the strongest predictor of successful growth beyond the core (Zook & Allen, 2001/2010). Second, hidden-asset extraction: before adding anything external, audit what the firm already owns but underuses, proprietary data, a delivery method that could be productized, a client community, training IP. Zook's research found renewal more often comes from these internal platforms than from new-market entry (Zook, 2007). Third, real-options sequencing: structure any genuinely new bet as a cheap, time-boxed experiment with explicit kill criteria, a paid pilot with three clients, not a hired team and a rebrand. Cap total experimental allocation at a fixed share of profit (10-15% is a common operator heuristic, a practice norm, not a research finding), require segment-level P&Ls from day one so cross-subsidy is visible immediately (Berger & Ofek, 1995), and pre-commit the kill date in writing. The goal is not zero exploration; it is exploration that cannot silently eat the core.
Section 6
Solution framework: the focus test for sub-$10M firms
Before any diversification move, run four gates in order. Gate one, core saturation test: is the core actually exhausted? Compute your share of realistically addressable niche revenue. Bain's research found most companies abandon their core long before its full potential is extracted (Zook & Allen, 2001/2010); if your share is under roughly 10-15% of the niche, the highest-return move is almost always deeper, not wider. Gate two, core strength test: diversify from strength, never from weakness. The core must be profitable at market-rate owner compensation, growing, and operable without daily founder involvement. Diversifying to escape a struggling core doubles the number of struggling businesses. Gate three, adjacency test: score the proposed move on shared customers, capabilities, channels, and economics. Three or more shared dimensions: proceed to gate four. Fewer: treat it as a new company and ask whether you would quit your current one to start it, because in attention terms, you partially are. Gate four, structure test: separate P&L, capped budget from profits (not operating cash), explicit success metrics and kill date, and a named owner who is not the founder if at all possible. Legitimate diversification triggers do exist: a single client above 25-30% of revenue, a structurally declining niche, or a repeatable adjacency formula already proven once. Concentration risk is the honest reason to diversify; boredom and stall-panic are the common ones.
Section 7
Evidence-based action plan
Days 1-30: see the firm as a portfolio. Build segment-level P&Ls for every distinct service line and market you currently serve, fully loading founder time and shared costs to each. Most multi-line small firms discover one segment funding the rest, the Berger-Ofek cross-subsidy mechanism operating in miniature (Berger & Ofek, 1995). Rank segments by gross profit per founder-hour. Days 31-60: prune and concentrate. Sunset or price-correct the bottom segment, redirect the freed capacity to the strongest core, and write a one-page core definition: who you serve, what problem, what offer, what proof. Test it against the Hinge specialization evidence, if your positioning would not let a referrer describe you in one sentence, it is too broad (Hinge, 2025). Days 61-90: extract before you expand. Run the hidden-asset audit, IP, data, methods, community, and pick one internal asset to develop into a core-reinforcing offer, applying the adjacency scorecard before committing (Zook, 2007). Quarterly thereafter: hold a capital-allocation review with three standing questions. Is the core saturated yet (measured, not felt)? Does any client exceed 25-30% of revenue (the legitimate diversification trigger)? Is any experiment past its kill date? Annually, restate enterprise value through a buyer's eyes: focused, systematized firms clear median multiples while mixed firms get discounted or parted out (BizBuySell, 2026). The compounding case for focus is that every quarter of concentration makes the core harder to compete with, and that is the asset you eventually sell. For adjacent evidence in this pillar, see [The Overhead Audit: What SaaS Sprawl Research Says About Reclaiming Margin](/blog/growth-overhead-audit-saas-sprawl) and [Outsourcing vs Hiring: The Research on Flexible Capacity for Service Firms](/blog/growth-outsourcing-vs-hiring-service-firms).