The work between the specialists belongs in the price.

Consider a hypothetical manufacturer buying a weekly inventory report. An analyst reconciles stock records. A developer schedules the import. An operations specialist identifies which material another plant can actually use. On Friday morning, the report should help someone decide what to move.

On Thursday, the export format changes. The developer is away. The analyst can see that the import failed but cannot repair it. By Friday morning, the client is arranging a call between three capable people while the transfer decision waits.

The specialists may each have delivered what their individual agreements required. The client still has an unfinished job.

This is the question I would put ahead of hourly rates when evaluating a collective: who is responsible for the work that falls between its members?

What the fee has to carry

Someone has to translate the request, resolve disagreements about the data, check the output, arrange cover and deal with changes. A client can retain those duties. A specialist can absorb them. A service provider can take responsibility for them. The proposal should make that choice visible.

Otherwise, a low quote can conceal a substantial assignment for the buyer. The operations manager becomes the project manager, translator and emergency support desk. Or a specialist spends unpaid evenings keeping the arrangement together. The apparent saving comes from leaving someone's work out of the comparison.

For a recurring service, I would compare the fee alongside the client's remaining hours, the cost of delayed decisions and the work required when something fails. Some of those costs will be uncertain. Naming them still produces a better comparison than treating every proposal as the same service at a different price.

A collective's margin has work to do. It may need to fund documentation, review, backup capacity and time spent resolving problems across assignments. If every available hour is sold as direct production, the promise of continuity has no capacity behind it.

There is room to improve the deal for both sides. Shared tools can eliminate repeated setup. Familiar colleagues can need less explanation. Several clients may support a specialist whose expertise none needs full time. Those efficiencies must cover the additional coordination before they can support better pay, a lower client cost or both.

Better tools change the work a team can carry

The June 2026 publication of The Cybernetic Teammate gives a reason to reconsider how expertise is combined. In a randomized product-innovation experiment at P&G, individuals using AI produced proposals comparable in quality to those of two-person teams without it. The study also found that AI helped participants draw beyond their functional specialties. It examined a day of product-innovation work; it did not test running a continuing operation with fewer people. Dell’Acqua and colleagues, Organization Science

The broader business evidence asks a different question. The March 2026 revision of Firm Data on AI, surveying nearly 6,000 senior executives across four countries, reports that 89% saw no productivity effect from AI over the preceding three years. These are executives’ assessments, rather than a controlled measurement of what AI caused. Yotzov and colleagues, NBER working paper

Neither result supplies a staffing formula. My conclusion for service design is to evaluate the whole responsibility: preparation, judgment, coordination and completion. A tool may let one person cover more of it. The remaining duties still need time, authority and someone accountable for them.

Put the awkward cases in the agreement

In the inventory example, the agreement could assign the collective responsibility for a checked report by Friday morning, including routine corrections to the import and a named backup. The client would supply the agreed source by a stated deadline and retain the decision to move material.

A renamed column might fall within normal maintenance. Replacing the source system might require new work. An unresolved stock discrepancy might require the report to identify the affected material and route it to a client decision-maker before anyone relies on it.

These details determine what the service is. They also determine whether the people delivering it have a workable job. Availability has to be agreed and paid for. Scope changes need a way to become paid assignments. A specialist should know who can accept a request, what counts as completion and when payment is due.

I have dealt with coverage as a practical operating problem. As my planning team went from nine to five while retaining its scope, I rewrote roles, established progression and succession, and documented backup assignments. Two planners moved into senior roles. That experience informs my view here: a smaller group needs deliberate provision for the work that continues when someone is unavailable.

Where an internal hire can be better

Some work depends on constant access to context, competing priorities and authority inside the business. An experienced employee may resolve an issue in a short conversation that takes an outside group several handoffs to understand. Splitting that work across specialists can increase the coordination burden.

A collective can also create its own bottleneck. If every request and exception must pass through the founder, the client has acquired another approval queue. Knowing talented people does not settle who can commit their time, resolve a conflict or authorize a change.

I would therefore examine the decisions before choosing the arrangement. Which can a provider make within agreed limits? Which belong to the client? How often does the work cross that boundary? Frequent, consequential crossings may favor keeping the role inside the company.

For work with a dependable boundary, a collective deserves consideration. I would ask it to walk through the ordinary week and then the Thursday when the source changes and someone is away. The answer should identify who acts, what capacity is available, what the client still owes and what happens to the price. That is where the proposed service becomes concrete enough to buy.