An enterprise's reward architecture is its real declaration of what it believes a team is, and most enterprises apply collective reward to work groups and individual reward to work teams — which is exactly the wrong way round.
The decision arrives as administration. Remuneration asks whether this year's incentive for a unit should be paid to individuals or shared, someone senior picks one, and the question closes inside a meeting with four other items on it. No rationale is recorded, because none was requested.
That decision is the most explicit statement the enterprise ever makes about what it believes the unit is. Every other statement — org chart, charter, away-day, the word "team" in the unit's name — is cheap. The reward instrument is not, because it alone determines what happens to a person who behaves as though the opposite were true.
Two kinds of unit are routinely called teams. In one, members contribute independently: each holds a deliverable and the unit's output is the sum of what its members produce. In the other, no member could produce the output alone and the contributions cannot be cleanly separated afterwards. The first is a work group, the second a work team, and accountability, reward and meeting ritual should differ between them.
What the teaching material almost never supplies is a test. The distinction is asserted, illustrated with examples chosen to be unmistakable — a dozen specialists filing independent reports on one side, a handful of engineers integrating a system on the other — and handed to a manager's judgement. Such examples teach nothing about the ambiguous middle, where nearly all real work sits.
The Strategic Context
ERANORTH's contention is that most enterprises get the pairing backwards. Collective reward goes to work groups, because a shared bonus reads as collaborative, costs nothing extra to administer and offends nobody at design. Individual reward goes to work teams, because the performance system already produces individual measures and joint output is hard to attribute.
Backwards is worse than random. A collective bonus over independent contributors prices nothing: each member's effort moves the pool by a fraction, so the rational response is to reduce effort, and the strongest contributor is taxed by the weakest with no way to influence them. Individual reward over a genuine team prices the wrong thing: it rewards the separable part of a contribution, which in joint work is the least valuable part, and penalises the integrating behaviour the team exists to produce.
What Leaders Commonly Misread
The first misreading is that collective reward builds collaboration. It does not build anything; it distributes — along the real structure of the work where output is joint, arbitrarily where it is not. People notice within one cycle.
The second is that the classification describes how people feel about each other. Cohesion and shared identity matter, but the variable is whether the output can be separated. A close, mutually supportive collection of people with independent deliverables is a work group with good morale, and paying it collectively erodes the morale it already had.
The third is that the name is evidence. Units acquire the word "team" because it is complimentary, and once acquired it never leaves, so an enterprise classifying from titles errs in one direction only.
Reframing the Issue
Stop asking whether a unit is a team in the aspirational sense and ask a narrower question with a checkable answer: can its output be attributed to individuals without a joint investigation?
Take a hypothetical commercial nursery and landscape supply business. Propagation, potting and dispatch run as crews with countable individual output, and the business pays a site-wide seasonal bonus on dispatch volume — collective reward over three work groups. Its contract landscaping unit, where a designer, a horticulturalist, an installer and a logistics coordinator jointly produce a garden that either works or does not, is paid on individual utilisation and hours booked to job. Both failures show on the ground: the dispatch crews carry passengers every season, and the landscaping unit argues about hours while nobody owns the result.
The same inversion runs through a hypothetical regional newspaper group, where reporters filing independent stories share a masthead bonus tied to circulation while the investigations desk, whose output is one piece no single person could have produced, is measured on bylines per reporter. Bylines are the separable part, which is the part that does not matter there.
Where the Classification Actually Fails
The middle is not rare; it is the norm
Teaching examples sit at the extremes because extremes are teachable. Operating reality is a unit whose work is independent most of the period and briefly, intensely joint at integration or close. Classified as a team all year it prices coordination that is not happening; classified as a group it prices nothing during the weeks that decide the outcome. Neither instrument is wrong; applying one for twelve months is.
The classification is made by the person with the most reason to get it wrong
The unit's manager usually decides. Classifying the unit as a team brings a larger bonus pool, a stronger claim on headcount and a more impressive charter; classifying it as a group brings clearer attribution and harder conversations. The incentive runs one way, and the enterprise has no independent check on it. What a unit's record then says about the person leading it is a separate confound, and [Related article: Leadership, or Headroom?] owns it — tenure-long performance data cannot distinguish leader quality from inherited headroom.
Decision Framework
The joint-output test. Run it on every unit carrying a collective incentive and every unit with "team" in its name, using the last two periods of output records. It takes under an hour per unit.
Answer four questions from the record, not from impression.
| Question | Answer indicating a team |
|---|---|
| Remove one member for a fortnight: does output fall in proportion to their share, or does a class of output stop? | A class of output stops |
| Could you have written, before the period, a sentence assigning each unit of output to one named person that would still be true at the end? | No |
| When the output is defective, can a single owner be identified without a joint investigation? | No |
| Does any member routinely need another member's unfinished work as an input? | Yes |
Count the team-indicating answers. Zero or one is a work group: individual reward, individual accountability, ceremonies for information-sharing rather than joint decision. A collective bonus here is a transfer, not an incentive, and should be withdrawn. Three or four is a work team: collective reward on the joint output, explicitly mutual accountability, review of the whole rather than the parts. Two is not a classification. Do not split the difference — the work design is the problem, and the unit should be split into a group and a team, or have its interdependencies deliberately increased until it scores clearly.
Two governance rules make it hold. The test is run by someone who does not manage the unit; and the classification, not the title, determines the reward instrument — a unit scoring zero or one loses the word "team" in the same cycle.
From Strategy to Execution
Immediate. Run the test on the three units with the largest collective incentive pools. Most executives assume changing a reward instrument needs a board paper; it usually does not, and [Related article: What Are You Already Permitted to Do?] owns that question — executives read a mandate for what it forbids and miss what it permits, and the delivery strategy almost always lives in the permissions.
Medium term. Make the classification a required field wherever an incentive is designed, with score and date recorded, and re-run it whenever a unit's work changes materially — classification is a property of the work, not of the people.
Long term. Design the work first and the reward second. Where the enterprise wants joint output it must create genuine interdependence; where it does not, it should stop asking for teamwork it has not built the conditions for. Whether a reclassification effort survives its own launch depends on something else again, and [Related article: Which Half of Your Capability Programme Has an Internal Buyer?] owns it — you can predict which parts of a capability programme survive by asking which have an internal buyer with a commercial reason to keep delivering them.
Signals to Monitor
Watch for members of a collectively rewarded unit privately keeping their own records of individual contribution: the clearest evidence a group is being paid as a team. Watch for defect investigations that always run long, which usually means attribution is joint and the reward is not. Watch for high performers requesting transfers out of shared-pool units. And watch the first genuinely bad period, since that is when a mismatched instrument produces its bill.
Questions for the Leadership Team
- For each unit carrying a collective incentive, what evidence in the output record shows the work is joint?
- Which of our units named "team" would score zero or one on the joint-output test, and who benefits from that name remaining?
- Where have we asked for collaboration in a charter while paying entirely for separable individual output, and for how long?
- Who currently decides a unit's classification, and what do they gain from the answer they give?
- In our last three defect investigations, how long did attribution take, and what does that say about the units involved?
- Which units sit in the ambiguous middle, and is that because the work is genuinely mixed or because nobody has designed it?
Closing Perspective
There is no neutral option available here. Every unit is already being paid one way or the other, and that instrument is teaching the people inside it which of the enterprise's two accounts of teamwork is the operative one.
The responsibility sits with whoever signs the incentive design, not whoever writes the charter. Signing a collective instrument over separable work, or an individual one over joint work, is a decision about what the enterprise will get — and it will get exactly what it has priced, on schedule, regardless of what it has been saying.
About EraNorth Insights
EraNorth Insights publishes practical analysis on strategy, projects, operations, transformation and decision intelligence for professional and organisational use. About EraNorth.