Cross-functional decision rights: who decides when two functions disagree?
Product wants to ship in October. Sales has already promised two accounts it lands in August. Both positions are defensible, both leaders are competent, and the disagreement has now consumed four meetings. Eventually it reaches you, and you make the call in eleven minutes.
I get asked to fix this often, and the request almost always arrives as a relationship problem. Product and Sales do not trust each other. The teams need to align.
A workshop will not hold, because the disagreement is not the failure. The failure is that your cross-functional decision rights were never assigned. Most of what looks like interdepartmental conflict is an unowned decision sitting in the seam between two functions, and it travels upward because escalation is the only route anyone ever built.
- Vertical authority is usually designed and horizontal authority almost never is, so a decision that belongs to two functions has no owner and defaults to the only mechanism available, which is escalation to the nearest common boss.
- The Seam Contract: for each recurring shared decision, name one decider, one consulted function, a fixed consult window, and a pre-named tiebreaker with the standard they will apply. Four lines, written once, reused indefinitely.
- Input right is not veto right. Most cross-functional deadlock comes from both functions holding an implicit veto, which makes agreement the only legal outcome and turns every genuine disagreement into a stall.
- Blenko, Mankins and Rogers surveyed nearly 800 companies globally and found top-quintile companies averaged 71 out of 100 on decision effectiveness while everyone else averaged 28, a gap they attribute to the compounding of decision quality, speed and yield.
- Sy and D'Annunzio studied 294 managers across seven multinational corporations and ranked unclear roles and responsibilities second and ambiguous authority third among the top five challenges of matrix structures, ahead of anything about personalities.
- Escalation is a legitimate safety valve and a terrible default: every seam you leave undefined converts into a recurring appointment on the calendar of whoever sits above both functions.
Why do cross-functional decisions stall?
Cross-functional decisions stall because no single function holds the authority to close them, and no rule says what happens when the two functions want different things. The decision is not hard. It is unowned.
Look at how authority actually gets assigned in most companies. The org chart runs vertically, and so does everything built on top of it: reporting lines, budget approval, performance review, escalation. Ask who a person answers to and the answer is instant. Ask who decides a pricing exception that Sales wants and Finance resists, and you get a pause, then a name that turns out to be a guess.
That pause is the whole problem. Both leaders believe they have a legitimate claim, because both do. Product owns the roadmap. Sales owns the commitment to the customer. Neither is overstepping. The structure simply never said which claim wins, so the two of them negotiate, and negotiation without a decision rule produces either a stall or a trade that has nothing to do with the merits.
The research on matrix structures found the same ordering. Thomas Sy and Laura Sue D'Annunzio surveyed, interviewed and ran workshops with 294 top-level and mid-level managers across seven major multinational corporations in six industries. The top five challenges they identified were misaligned goals, unclear roles and responsibilities, ambiguous authority, lack of a matrix guardian, and silo-focused employees (Sy & D'Annunzio, 2005). Four of those five are structural. None of them is about whether people get along. Worth noting the scope: these were large multinationals, not small local firms, though the mechanism does not appear to require scale.
What are cross-functional decision rights?
Cross-functional decision rights are the allocation of authority over decisions whose inputs, costs or consequences land in more than one function: who decides, who must be consulted, what happens when the consulted function objects, and who resolves a genuine impasse.
They are a specific subset of decision rights and the one most likely to be missing. A company that has never written down who decides anything at least fails consistently. A company that has assigned authority carefully within each function, and left the seams between them blank, fails in a way that looks personal, because the only visible artifact is two capable people arguing.
In Leadership Architecture this shows up as two things at once. On the demand side, Decision Coupling Density measures how many roles a typical decision touches and how far its consequences travel across departments. On the capability side, Decision Clarity measures whether authority is defined and applied consistently. Growth raises the first without touching the second. The seams multiply, the rules do not, and a structure that worked at forty people starts producing standoffs at ninety with nobody behaving any differently.
The Seam Contract: four fields that close a shared decision.
The fix is smaller than the problem suggests. For each recurring decision that sits between two functions, write four lines. I call it the Seam Contract, because it is an agreement about the boundary rather than about the topic.
Not both functions, not the pair, not a committee. If you cannot name a single decider, you have not found a decision that genuinely needs two owners. You have found a decision nobody wants to be wrong about.
Write who, and write what they are entitled to: the data, the reasoning, a response. This is the field that makes the contract acceptable to the function that does not hold the pen.
A week, three days, one meeting. Without a clock, "we are still consulting" becomes indistinguishable from a veto, and the decision dies quietly rather than visibly.
Naming the standard matters more than naming the person: "whichever choice protects the annual revenue commitment" is a rule, and "the COO will decide" is just a queue.
Decider, consulted, clock, tiebreak. If that reads like a role-assignment framework, it is close to one, and the RAPID, RACI and DARE family all cover similar ground on a single decision. The difference is the clock and the named standard, which is what turns a role chart into something that resolves rather than describes. The work takes about ten minutes per decision and you do it once, outside the heat of any particular argument. That last condition is not optional. A Seam Contract written during a live dispute becomes a negotiation about that dispute, and you will get a settlement instead of a rule.
The field that does the most work is the second one, and specifically the distinction it enforces. An input right is not a veto right. In most stalled organizations both functions hold an unstated veto, which means agreement is the only outcome the structure permits. Every genuine disagreement then has exactly one exit, upward, and the leaders learn to route around the deadlock through relationships and favours rather than through the decision system.
Why does escalating to the common boss make it worse?
Escalation deserves defending before it gets criticized. A structure with no upward path is worse than one that overuses it, and some decisions genuinely belong to whoever sits above both functions.
The trouble is escalation as a default. When the common boss becomes the standing resolution mechanism for peer disagreements, three things follow. Their calendar fills with decisions two levels below their pay grade. The two functions stop investing in resolving anything themselves, because the referee is free and reliable. And the escalated decision gets made with less context than either function had, by the one person in the building who was not in the details.
This is the same machinery behind chronic escalation, viewed from the side rather than from above. Vertical escalation is the symptom. The undefined horizontal seam is the cause.
The cost is not abstract. Marcia Blenko, Michael Mankins and Paul Rogers surveyed nearly 800 companies worldwide on decision effectiveness and organizational health. Decision effectiveness and financial results were correlated at a 95 percent confidence level or higher for every country, industry and company size in the sample, and on a zero to one hundred scale, top-quintile companies averaged 71 while all other companies averaged 28 (Blenko, Mankins & Rogers, 2010). Their own wording is careful, and worth keeping careful:
"Decision effectiveness and financial results correlate at a 95 percent confidence level or higher for every country, industry and company size we studied."
Read that as a correlation rather than a cause, which is what it is. The useful part is the size of the gap and where the authors locate it: not in the quality of individual judgments, but in speed, yield and the sheer effort a decision costs to make. Seams are where effort accumulates.
Here is the pattern I run into most often, and I am describing a recurring shape across engagements rather than one company. A leadership team asks for help with two functions that cannot get along. We stop talking about the relationship and list every decision from the last quarter that involved both. Then we ask one question per item: who was authorized to close this alone? For most of the list the answer exists and is obvious the moment somebody says it aloud. The items were contested not because the two leaders disagreed about substance, but because neither could tell whether closing it alone would be read as a decision or as an overreach.
How do you find your seams in one hour?
You do not need a redesign to start. You need a list and an hour.
- List every decision from the last quarter that consumed more than one meeting and involved two or more functions. Pull it from agendas, not from memory.
- For each one, write the two functions whose interests actually collided. Be specific. "The business" is not a function.
- Write the name of the person who could have closed it alone. If a name comes easily, the seam is defined and something else went wrong. If it does not, you have found an undefined seam.
- Count the undefined ones, and sort them by how often the decision recurs. Frequency beats importance here, because a recurring seam charges you every cycle.
- Write a Seam Contract for the top three. Decider, consulted, clock, tiebreak. Circulate it, and apply it the next time the decision arrives rather than announcing it as a policy.
Three contracts will not fix your operating model. They will tell you something more useful, which is whether the conflict you have been managing was ever about the people. If the list comes back long, the seams are a symptom rather than the whole condition, and the wider decision rights audit is the better next hour.
The thing I would ask any leadership team to notice is what happens to a disagreement when the rule already exists. It gets shorter. Not friendlier, and not less pointed, just shorter, because the argument is now about the merits and both people know how it ends. That is the tell. If your cross-functional conflicts feel emotional, look at whether anyone ever told those two functions who wins. Teams do not fight because they lack goodwill. They fight because the structure left the question open and then asked them to settle it privately.
Frequently asked questions
What are cross-functional decision rights?
Cross-functional decision rights are the allocation of authority over decisions that affect more than one function: who decides, who must be consulted, what happens when the consulted function objects, and who resolves a genuine impasse. They differ from ordinary decision rights because the authority has to be assigned across a boundary rather than down a reporting line, which is why most organizations define the second and leave the first blank.
Who should decide when two departments disagree?
One named person in one function, chosen in advance rather than during the dispute. The other function gets a defined input right, a window in which to make its case, and access to a pre-named tiebreaker if it believes proceeding would cause real damage. Assigning the decision to both functions jointly is the most common version of this mistake, because it gives each an implicit veto and makes agreement the only permitted outcome.
What is the Seam Contract?
The Seam Contract is a four-line rule for a recurring decision that sits between two functions: the decider (one name), the consulted function and what it is entitled to, the clock (how long the consult window runs before the decider proceeds), and the tiebreak (who resolves an impasse and the standard they apply). It is written outside any live dispute and reused every time the decision recurs.
Is cross-functional conflict a people problem or a structural problem?
Usually structural, and it is worth testing before investing in the relationship. If two functions clash on the same category of decision repeatedly, and each clash ends with someone above them making the call, the variable is the missing rule rather than the personalities. Sy and D'Annunzio's study of 294 managers across seven multinationals put unclear roles and responsibilities and ambiguous authority second and third among the top five matrix challenges, ahead of anything about how people relate.
How do I know whether my organization has undefined seams?
Take every decision from the last quarter that took more than one meeting and involved two or more functions, then write next to each the name of the person who could have closed it alone. Every item where no name comes easily is an undefined seam. Sort those by how frequently the decision recurs, because a recurring seam costs you every cycle rather than once.
Sources
Thomas Sy & Laura Sue D'Annunzio, Challenges and Strategies of Matrix Organizations: Top-Level and Mid-Level Managers' Perspectives, Human Resource Planning, 2005, 28(1), 39–48. Surveys, interviews and workshops with 294 top-level and mid-level managers across seven major multinational corporations in six industries; source of the top five matrix challenges and their ordering.
Marcia W. Blenko, Michael C. Mankins & Paul Rogers, Decision Insights: Score Your Organization, Bain & Company, 2010. Source of the survey of nearly 800 companies, the 95 percent confidence statement quoted above, and the 71 versus 28 decision-effectiveness scores. See also Blenko, Mankins & Rogers, The Decision-Driven Organization, Harvard Business Review, 2010.
Client examples are anonymized composites; no figures are invented.
How many of your seams are undefined?
Five minutes. No account. A structural read on whether your decision system, not your two department heads, is the real constraint.