In a growth review, the line item that attracts the fewest questions is frequently the one growing fastest; as the rate of increase rises, the volume of interrogation directed at that line falls, and whatever questioning remains migrates toward the cost side of the page. In the same session, one observes that hiring approvals clear more quickly even though the signature chain governing them has not changed, and that the interval between a proposal draft and its authorised version narrows quarter over quarter. The approval mechanism remains formally intact while the weight it actually carries diminishes. Read together, these two tendencies describe not an organisation that has dismantled its controls, but one whose controls have fallen out of step with the tempo of the business they were designed to govern, which is a materially different diagnosis with materially different remedies.
A second pattern, less frequently noticed, appears on the documentation surface. In a company that has multiplied its delivery volume several times over a short interval, the document produced most quickly on request is the revenue breakdown, while the document produced last is the inventory of contracts actually in force; the question of how many of the last hundred agreements were executed against the standard form, which clauses were departed from, and on whose authority those departures were granted can rarely be answered from a single source. The answer is dispersed because the deviations were recorded not in a system of record but in the recollection of individual negotiations, distributed across the people who conducted them. That dispersion is not an incidental side effect of growth; it is the direct residue of the choice that made the growth possible.
The choice has a name — blitzscaling, the deliberate subordination of efficiency to scale while uncertainty persists — and under identifiable conditions it is entirely rational. In markets where increasing returns hold, where network effects or switching costs defend the position established first, and where whoever sets the standard captures the pricing power of the subsequent period, the cost of arriving second exceeds the cost of operating inefficiently. The same logic operates in capital-intensive settings where a scarce and sequential resource governs access: a place in the interconnection queue, the duration for which a permitting window remains open, the calendar threshold to which incentive eligibility is tied. Under such conditions speed is not profligacy but a form of option purchase, and the premium paid is defensible to the extent that it remains below the value of the position that lateness would forfeit.
The risk resides not in the decision to move quickly but in that decision persisting after the conditions justifying it have lapsed, and three mechanisms sustain that persistence. The first is the internal status system: in an organisation that rewards velocity, the person who slows a transaction is positioned as an obstacle rather than as a source of prudence, and over time that positioning raises the personal cost of objection until objections stop being raised. The second is measurement lag; to the extent that control systems report on a rhythm slower than the growth cycle itself, the signal arrives after the decision point and correction always operates one period behind. The third is the absence of a record: when the conditions justifying speed are not written down at the moment of decision, their disappearance cannot be observed either, and a strategy that worked once is treated as evidence that it will work again.
The price of speed is recorded not in the income statement but across three separate ledgers. The first is documentation debt — unsigned change orders, annexes presumed to be completed later, delivery scopes resting on verbal understanding. The second is process debt: approval thresholds that were never reduced to writing, pricing logic held in a single individual's recollection, exception procedures that remain undefined. The third is managerial-density debt, which accumulates as the ratio of experienced managers to newly joined staff declines and institutional behaviour begins transmitting through imitation rather than through documentation, with the consequence that behaviour transmitted by imitation departs when the person who modelled it departs. What these three ledgers share is that none of them is measured in the period that creates them, and all of them are measured in the period that follows.
At the valuation table, these ledgers translate into a question not of whether the growth is credible but of whether it is decomposable. An acquirer or a lender applies a discount not because rapid growth invites suspicion but because it cannot be separated into its constituent parts; absent cohort data distinguishing gross additions from net retention, there is no basis on which to establish that attrition is not being masked by acquisition. The same gap appears in customer concentration, in pricing discipline, and in whether recurring revenue genuinely recurs. What sustains a valuation is, more often than not, not the performance itself but the demonstrable proposition that the performance is repeatable independently of the founder and of individual negotiations; where that demonstration cannot be made, the difference is deducted from the multiple.
A second cost accumulates on the contractual surface. Concessions granted to accelerate signature appear minor when examined individually — a raised liability cap, a removed ceiling on liquidated damages, an extended warranty scope, a unilateral termination right conceded to the counterparty, or a side letter that steps outside the standard form — yet once they constitute a portfolio they surface collectively during the transaction. Every deviation that becomes a due diligence finding carries a concrete price: an increased escrow ratio, a lengthened list of conditions precedent, a narrowed scope of representations and warranties, an exclusion appended to the insurance policy, a portion of the consideration deferred into an earn-out structure. Speed is taken as an advantage at the moment of signature and repaid as an obligation at the moment of closing, generally at a rate the signatory never quoted.
A third cost sits on the cash and execution side, and it appears late because the accounting match breaks down. Growth consumes working capital; receivable ageing extends while supplier terms fail to flex at comparable speed, and expanding revenue rarely arrives accompanied by a contracting cash conversion cycle. On the execution side, the cost of quality is recognised one period after the revenue that generated it: rework, the punch list accumulating through commissioning, the nonconformities that emerge at site acceptance, and the schedule slippage those items produce. For this reason the true cost of speed becomes visible not in the quarter of peak velocity but in the quarter that follows it, by which point establishing the causal relationship between the two has become appreciably harder to argue in front of a board.
This tendency is neutralised through institutional architecture rather than individual prudence, and the intervention separates into four components. The first is a conditions record: the assumptions justifying speed — that the market produces increasing returns, that the position is scarce and sequential, that capital remains accessible — are written at the moment of decision and each is bound to a validity test, so that the expiry of a condition becomes observable rather than inferential. The second is deriving the speed ceiling from capacity rather than from demand, with the upper bound on growth taken from an internal indicator such as managerial density or commissioning throughput. The third is a deviation register, in which every departure from the standard form is recorded at signature, with its rationale and its approver, rather than at diligence. The fourth is a reversibility test, under which decisions taken quickly are treated as defensible to the extent that they can be unwound, and irreversible decisions are subjected to a separate threshold.
The same structure carries different meaning for the parties around the table. For the founder, the question is not whether speed should be constrained but which lines advance quickly and which advance slowly, deliberately separated — a market position may be established rapidly while the contract standard is established slowly and from a single centre. For the investment committee, the question is not the growth rate itself but whether the conditions under which that rate remains sustainable can be defended through a written thesis rather than an inferred one. For the senior secured lender, the question is covenant calibration; a covenant package that never triggers during expansion but becomes abruptly binding the moment growth stops typically serves none of the parties, and tends to convert a manageable slowdown into a restructuring conversation.
Intervention in this area generally begins with the construction of three mechanisms. The decision record is maintained at the moment of proposal rather than the moment of approval, since when the rationale, the alternative and the assumptions behind an acceleration decision are committed to writing while the outcome remains unknown, the subsequent assessment cannot be rewritten by the outcome itself. The deviation register is embedded within the contracting process, so that each departure from the standard form is tracked as a portfolio item rather than as an isolated negotiating moment, which makes collective remediation feasible during transaction preparation. Control gates are tied to thresholds rather than to the calendar: a review rhythm that engages when a defined volume, headcount or new-jurisdiction threshold is crossed, and stays dormant otherwise, renders the debt accumulated by speed visible without penalising speed itself.
The indicator that distinguishes speed as a strategic choice from speed as an institutional habit is how the organisation goes about deciding to slow down. Having chosen speed once does not amount to choosing it again each year, and the moment a company crosses the maturity threshold is not the moment it can defend its growth rate, but the moment it can show in writing which condition carries that rate and what it intends to do when that condition disappears. The distinction matters because the first capability is a narrative asset that expires with the cycle, while the second is a governance asset that survives it, and buyers, lenders and boards have become measurably more attentive to which of the two they are being shown.