Bought capacity is delivery work a company sells under its own name and pays someone else to perform: subcontractors, partner firms, labor-only crews, white-label installers. It is the only response to a capacity shortfall that can be live next week rather than next quarter. The price of that speed is not in the rate card. It shows up in coordination hours nobody budgeted, in a warranty promise that outlives the contract behind it, and in the part of the customer relationship the company has quietly stopped controlling.
This guide is for portfolio executives and sponsors in services and multi-site businesses deciding whether to buy capacity, how much, and for which work. It covers the margin arithmetic beyond the rate card, which jobs can be subcontracted without the customer minding, warranty and callback exposure, brand and reviews when the van is not yours, what the customer is told, the rate card as a ceiling on repricing, and when bought capacity should convert to hired. Selling into a capacity ceiling sets out the four honest responses to a live ceiling and treats adding capacity as a hiring decision measured in months. Buying it is the version that acts in weeks, and this is the build-out of that one option. Contract terms, notice periods and who holds the accounts when a supplier leaves sit in marketing vendor consolidation, and the mechanics of raising prices in a pricing reset.
The comparison most management teams run is the subcontract rate against an hourly wage, and it is decided before it starts because the two sides are not loaded the same way. The subcontract rate is a finished price. The wage is one line of a cost that also carries payroll taxes, insurance, vehicle, tools, training, supervision and the hours the crew is paid for and not working.
Load both sides, and charge the central costs to the jobs that caused them. Four lines are usually carried centrally and never charged to bought volume, which is why bought volume looks profitable per job and thins the margin by the month.
Charge those four lines where they belong and the real comparison is not about rates at all. An employed crew is cheaper per unit only if you can keep it busy, because its cost does not fall when the week is empty, and bought capacity costs more per unit and nothing at all on an empty week. The choice is a judgment about how durable the demand is. A business certain of its volume for two years and still buying that volume is paying a premium for optionality it no longer needs.
The instinct is to sort by complexity and keep the hard jobs in house. That is the wrong cut: plenty of simple work is commercially load-bearing, and plenty of complicated work is a specification someone else can follow exactly. The cut that holds is whether the job creates the relationship or spends one that already exists.
Keep in house the jobs where someone explains what is wrong and what it will cost, because that conversation is where the customer decides whether to call again. Keep first visits, diagnostic calls where the recommendation sets the price, warranty returns, anything on a named account, and anything that habitually produces a referral. Buy the work whose scope is settled before anyone arrives: single-visit installs to specification, routine repeat work in overflow, and jobs at the geographic edge where your own crew would spend two hours driving.
This cut matters more than complexity because the moment the customer forms a view is rarely the moment the work is performed. It is the moment someone stands in their property and tells them what it needs. Sell that moment and you have subcontracted the part of the job you were being paid for. In one Claymore engagement we found 38 percent of new business arriving through word of mouth that no system was recording, which sharpens the point: the jobs that generate referrals are the least safe to buy, and because the referrals are unrecorded, nothing in the monthly pack shows the damage.
The geographic edge is the best use of bought capacity there is, because it lets a business serve a new area before it staffs one, which is how a footprint gets tested rather than assumed. In one engagement, commercial work of that kind supported 82 percent footprint growth in 15 months. Service-area expansion covers the evidence that decision needs.
Two obligations exist on every bought job and they are not the same shape. The customer holds a promise from you, for the term you publish, against a company that plainly exists. You hold whatever recourse the subcontract gives you, for whatever term it states, enforceable only while that partner is trading and solvent. The gap is the exposure, and it is almost never sized.
Size it as you would any other reserve: the callback rate on bought work, times the average cost of a callback, times the volume bought, against the margin that volume earned. If the recourse term is shorter than the promise, the company is self-insuring the difference, and that is a cost belonging in the arithmetic above rather than a risk belonging in a footnote.
Callbacks on bought work also cost more. Your own crew can be redirected this afternoon; a partner returns on a schedule you do not control, so either the customer waits twice or your crew does the job again at full cost. Insurance certificates, indemnities and worker classification are matters for counsel, and classification should go to counsel and HR before volume grows rather than after. The commercial decision is narrower: price the gap, or close it in the contract.
A customer who books with you reviews you. The rating lands on your listing whoever performed the work, which means bought capacity is a marketing input whether or not anyone has classified it as one. Two things move, and they move independently.
Quality is the obvious one. The less obvious one is capture. Partner crews rarely ask for reviews, because asking is a habit built by a manager who is not theirs, so review volume falls on bought work even where the work itself is fine. Falling volume drags the location average and the local ranking that average feeds, so a platform can lose local search position on the strength of good work nobody was asked to describe. Track rating and review volume per crew, bought crews included, which is a small extension of treating reviews as an operating metric. Reported only as a location average, the dilution is invisible until the average moves, and by then the reviews are published and permanent.
Appearance is the third input and the cheapest to fix: an unmarked vehicle and an unfamiliar name on the paperwork tell the customer the work was bought, at the door.
The rule is about timing: before the visit, never at the door, never discovered from an invoice. A customer told at booking that an approved partner will carry out the work has been given information and can ask a question. One who meets an unfamiliar vehicle on the day has been given a surprise, and surprises come back as calls to the office while the crew waits.
Say who is accountable rather than apologizing for the arrangement, because accountability is what the customer is asking about: your name on the invoice, your warranty on the work, one number that reaches someone who can fix it. Commercial contracts frequently require consent to subcontract and some prohibit it, so read the contract before booking the partner.
This is the consequence most plans miss. Employed capacity gets cheaper per unit as utilization rises, because the fixed cost spreads. Bought capacity does not: the partner rate is a floor under cost that does not move with your volume, and it often moves against you once you become that partner's largest customer.
So the margin expansion a pricing reset is meant to produce is capped on bought volume at the distance between your new price and their rate, and that distance closes. Partners read the same market you do, and a price rise that holds is an argument for a rate rise in theirs. A reset underwritten across the whole book will under-deliver by roughly the bought share, so model the owned and bought halves separately.
The same structure is read closely at exit. A buyer looking at gross margin will ask what share of delivery is bought and whether those rates are contracted with terms or taken at spot. A large bought share on spot rates is not a cost question, it is a durability question about the margin itself, and it belongs in the evidence set described in the commercial evidence pack. S&P Global's 2026 research found 71 percent of GPs and 53 percent of LPs now prioritize operational value creation, so the composition of delivery capacity is a question a sponsor should expect.
Share of revenue is a tempting trigger and a poor one: it says nothing about whether the demand will outlast the fixed cost. Three tests settle it better.
Converting too late is the quieter failure and the more common one: the company pays the premium indefinitely while funding a partner who is building exactly the capability needed to compete for the same work. Converting too early buys a payroll against demand that has not committed.
Make the decision somewhere it recurs. Bought share of delivered jobs, callback rate split between owned and bought, and review rating and volume per crew belong as standing lines in the monthly commercial operating review, alongside the booked-out horizon that triggers the conversation. In one engagement, commercial work of this kind produced a 36 percent revenue lift over 24 months on flat marketing spend, and what made it durable was that every decision had a place it was read.
The most useful question each quarter is not how much capacity is bought. It is which of the bought jobs the company would be uncomfortable having the customer find out about. That list is not a disclosure problem. It is the conversion list, and it is shorter and more obvious than the argument about rates.
Bought capacity is delivery work a company sells under its own name and pays another party to perform: subcontractors, partner firms, labor-only crews or white-label installers. It is the only response to a capacity shortfall that can be live within weeks, which is why it is used when demand outruns the schedule. The commercial consequence is that the company keeps the customer relationship, the warranty and the review, while outsourcing the visit that forms the customer's opinion. That split is what has to be managed, and it does not appear on the rate card.
Usually not, once both sides are loaded properly. A subcontract rate is a finished price, while a wage is one line of a cost that also carries payroll taxes, insurance, vehicle, tools, training, supervision and paid hours that are not billable. Against that, bought capacity carries coordination time, inspection and rework, a warranty reserve at its own callback rate, and the cost of jobs performed twice. The real difference is structural rather than arithmetic: an employed crew is cheaper per unit only if you can keep it busy, and bought capacity costs nothing on an empty week. The choice is about how durable the demand is.
Sort the work by whether a job creates a customer relationship or spends one that already exists, rather than by how difficult it is. Keep in house the first visit, any diagnostic call where the recommendation sets the price, warranty returns, work on named accounts, and the jobs that habitually produce referrals. Work whose scope is fully specified before anyone arrives can be bought with little risk, as can routine repeat work in overflow and jobs at the geographic edge of a service area. The moment a customer decides whether to call again is the moment someone explains the problem and the price, so that moment is the one to keep.
Yes, and the timing matters more than the wording: at booking, never at the door, and never left to be discovered from an invoice. A customer told in advance that an approved partner will perform the work can ask a question; a customer who meets an unfamiliar vehicle on the day has been given a surprise, and surprises come back as calls to the office while the crew waits. Lead with who is accountable rather than with an apology, since that is what the customer is actually asking. Commercial contracts often require consent to subcontract and some prohibit it, so read the contract before booking the partner.
Not at a share-of-revenue threshold, which says nothing about whether the demand will outlast the payroll. Convert when the booked-out horizon in that area and job type has held above the agreed band for a full seasonal cycle rather than a single strong quarter, when the work being bought is the kind that creates customer relationships, and when the business can genuinely hold the utilization its fully loaded comparison assumed. Converting too early buys fixed cost against unproven volume. Converting too late pays the premium indefinitely while funding a partner who is learning to compete for the same work.
Have a revenue problem the board is asking about? Start a conversation.