The most revealing observation about how a company hires is rarely found in the human capital file; it surfaces in the running order of the weekly management meeting. The segment covering open roles tends to sit near the end of the agenda, expanding or contracting according to whatever time remains, even though the sales target and the delivery schedule discussed earlier in the same meeting depend directly on those roles being filled. That a position has been open for three months usually emerges not from a tracker but from a verbal reminder by the relevant line manager, and in weeks when no reminder is offered, the position drops off the agenda entirely. Asked where candidates stand, the answer given describes quality rather than dates: there is a strong candidate, a second interview is pending, the founder will see them over the weekend.
The second and more consequential observation concerns where the final decision is actually made. In most mid-sized companies whose founder remains operationally active, no candidate who has reached offer stage begins work without meeting the founder — a rule that derives from no written policy and is questioned by no one, because it has worked for a long time. The founder's hit rate on candidates may well be genuinely high, having accumulated over years an understanding of what work the company does at what tempo, which temperament sustains that tempo, and which technical gap closes at which cost. The difficulty lies not in the existence of that accumulated judgment but in where it resides: so long as the knowledge sits inside one person's assessment, it registers as an asset on the founder's calendar rather than on the company's balance sheet.
What sustains this arrangement is the internal coherence of the short-run cost calculation. Formalising a hiring process — decomposing the role definition into assessment criteria, fixing the interview stages, specifying who examines what at each stage, recording the grounds for rejection — substitutes a procedure consuming several people's time for a decision the founder renders in fifteen minutes. The first outputs of that procedure will probably be weaker than the founder's own calls, since the system is not yet calibrated and calibration becomes possible only as data accumulates. The picture facing the decision-maker is therefore straightforward: building a slower and less accurate process today is the price of owning a transferable one tomorrow. Because the near-term cost is concrete and immediate while the long-term gain is abstract and deferred, the preference forms predictably in favour of leaving the existing arrangement intact.
The same mechanism takes a different shape on the documentation side. Most companies do in fact hold a set of hiring documents — role descriptions, a posting template, perhaps an interview form — but these were produced to satisfy an audit or certification requirement rather than to serve the flow of work. The gap between a document existing and a document being used is among the first things a reviewing party measures: asked in how many of the roles filled over the past twelve months that interview form was actually completed, the answer tends to come not from the form itself but from a manual sweep of the files, and more often than not it does not come at all. A documented but unpractised process yields a weaker signal in review than an undocumented one, since it evidences not merely a gap in capability but the normalisation, inside the company, of distance between what is written and what is done.
Measurement is the most common void in this area, and the reason for the void is not indifference but the perception of hiring as an event rather than a process. Events are not timed; processes are. The date the requisition was approved, the date the posting opened, the date of the first interview, the date the offer was extended and accepted — absent a record of those four intervals, everything that can be said about a company's hiring capacity is estimation. The second measure that follows is offer acceptance rate, which indicates simultaneously where the salary band sits relative to the market, how much of the effort expended up to offer stage is being wasted, and how the company reads to candidates. The third is attrition within the first twelve months, which reports miscalibrated selection criteria far more reliably than time-to-hire ever will.
Ownership, in turn, is answered not by reference to the organisational chart but by locating three authorities: the authority to open a requisition, the authority to set the salary band, and the authority to approve an offer. Where all three sit with one person — as they typically do in founder-active companies — the human capital function is a coordination unit rather than a decision centre, publishing postings, building calendars and collecting paperwork while exercising independent judgment at no threshold of the process. The configuration is easy to detect: asked whether they could extend an offer within their own budget to a candidate they consider suitable, a line manager gives an answer that departs from what the chart implies. Where ownership remains undistributed, hiring velocity is indexed to the availability of one calendar, and that index appears nowhere in the growth plan.
The institutional cost accumulates not in the cost line of recruiting but in the credibility of the growth scenario. A three-year projection presented in an investment review almost always embeds an assumption about headcount growth, tying the increase in sales volume to a defined number of field staff, production growth to a defined number of skilled operators, and services revenue to a defined number of specialists. Testing that assumption, the reviewing party interrogates not the cost of those hires but their obtainability: how many people the company recruited last year, in how many weeks on average a comparable role was filled, and by what multiple that cadence can be scaled before it degrades. Where those three questions have no answer supported by a record, the headcount line in the projection is classified as an aspiration rather than a plan, and the discount attaches not to payroll but to the entire revenue block resting on that line.
This assessment usually reaches the transaction through the payment architecture rather than through headline price. In companies unable to evidence headcount growth capacity, a portion of consideration migrates into an earn-out with a trigger keyed directly to revenue, which is the most economical way of leaving staffing supply risk on the seller's side. A second frequent arrangement converts key-person retention undertakings into conditions precedent and binds named individuals contractually — an arrangement that is itself founder dependency in priced form. A third defines the budget for roles planned within the first year as a discrete working capital reserve and subjects its release to approval. All three issue from the same judgment: the company may well be capable of producing headcount, but until it can demonstrate as much, no buyer assumes that capacity free of charge.
The intervention that neutralises this pattern is not the founder's withdrawal from interviews, which in most cases is neither feasible nor correct. It is the conversion of the founder's judgment into a record. In the work we run, the first mechanism established is that the decision record is kept at the moment of proposal rather than the moment of approval: when a candidate is advanced or declined, the reasoning is written under three fixed headings — technical sufficiency, fit with the tempo of the work, compensation expectation — and that record is later compared against who stayed and who left six months on, calibrating the selection criteria against observed outcomes. The second mechanism is a minimum hiring register composed of four date fields, requiring no additional software, without which any statement about time-to-hire remains unverifiable. The third is the separation of requisition, salary-band and offer-approval authority, with at least one of them devolved to the line manager.
The fourth component, and the one that meets the most resistance, is the establishment of rhythm. Moving open roles out of the variable slot at the end of the agenda and into a fixed heading with a fixed format — role, opening date, weeks elapsed, stage, owner — does not by itself shorten time-to-hire; it does make it impossible for a role to stay quietly open for three months. What we principally record while running that rhythm is not how long a given role has been open but at which stage the elapsed time accumulates: if the candidate pool never forms, the difficulty lies in the posting and the sourcing channels; if a pool forms but offers are never reached, the assessment criteria are unrealistic; if offers are declined, the salary band has separated from the market. Those three diagnoses call for entirely different remedies, and until they are disentangled the company spends resources solving the wrong problem.
None of these mechanisms is designed to raise hiring quality in the near term, and none displaces the founder's instinct. Their single function is to translate the decisions that instinct produces into a language the company can read, and that translation is precisely what is valued in an investment review. A twelve-month register exists if it was begun a year earlier; it cannot be reconstructed retroactively once a review has started, and any attempt to do so surfaces as date inconsistency at the first cross-check. Institutionalising the hiring process is therefore among the items in pre-sale preparation that permit the least delay: many other gaps can be closed before signing, whereas evidence of the capacity to produce headcount accrues only with elapsed time.
There is a single question worth asking about any company's hiring practice: were the founder absent from the room for three months, could a role of comparable specification be filled in comparable time and with comparable accuracy. If the answer is affirmative, hiring is a capacity of the company and the headcount line in the growth plan rests on a defensible assumption. If it is negative, what the company holds is not a process but the undelegated judgment of one person accumulated over many years — an asset that cannot be purchased, and a dependency carrying the risk of disappearing at closing.