Coordination Cost at the Marketing–Sales Interface: A Dependency Analysis

The marketing–sales handoff is the most examined interface in revenue operations and the least improved. The standard interventions — a shared definition of a qualified lead, a service-level agreement, a joint meeting — are correct in outline and are adopted almost universally, which is a strong hint that they are not sufficient.

There is a body of research on coordination as a general problem that offers a sharper framing, and a body of research on the marketing–sales interface specifically that constrains how much alignment is achievable at all.

1. Coordination as dependency management

Malone and Crowston [1] define coordination as the management of dependencies between activities, and set out a taxonomy: different dependency types call for different coordination mechanisms, and a mechanism appropriate to one type will not resolve another.

This is a more useful frame than "alignment" because it is diagnostic. The question stops being whether two teams get along and becomes which dependencies exist between their activities, and whether each has a mechanism.

Most organisations have a mechanism for the first dependency and nothing for the other four. That is a reasonable explanation for why an agreed definition of a qualified lead does not fix the interface: it addresses one dependency out of five.

2. What is structural rather than fixable

Homburg and Jensen [2] examined differences in orientation between marketing and sales — in time horizon, in customer versus product focus — and found that the two functions differ systematically. Importantly, their findings do not support the conclusion that all such differences are harmful; the effect depends on which difference, and some divergence is functional.

There is a further layer. Plouffe and Barclay [3] describe salesperson navigation — the intraorganisational effort a seller expends working across internal functions to serve a customer. That effort is real, is largely unmeasured, and is a direct cost of unmanaged dependencies. Where coordination mechanisms are absent, individuals substitute for them personally, which makes the cost invisible in every system that reports on the interface.

3. Instrument: measuring coordination cost

Four measures, all computable from timestamps that already exist. Run them over one complete quarter.

1. Acceptance rate. Of records passed as qualified, the proportion accepted without dispute. A low rate indicates a definitional problem; a rate near 100% with poor downstream conversion indicates the acceptance step is ceremonial.

2. Time to first touch, distribution rather than mean. The mean conceals the tail, and the tail is where the loss occurs. Report the median alongside the ninetieth percentile.

3. Re-verification rate. The proportion of received records where the receiving function re-checks or re-enters information already present. This is the purest measure of coordination cost — pure duplicated effort caused by absent trust in the upstream output.

4. Round-trip rework. Records returned upstream and re-passed. Each round trip is a prerequisite dependency without a sequencing mechanism, and the count identifies which field is missing.

4. Interventions matched to dependency

• Producer–consumer: make acceptance a real decision with a stated criterion and a rejection path that returns information upstream. An acceptance step with no possible rejection is not a mechanism.

• Shared resource: write an explicit allocation rule for accounts both functions may act on, including the tie-break. Absent a rule, the resolution is whoever acts first, which is a mechanism, just not a chosen one.

• Simultaneity: check downstream capacity before campaign launch. This is the least practised and among the cheapest.

• Task–subtask: preserve contribution visibility after handoff, so the upstream function can see outcomes. Without it there is no feedback and no learning.

• Prerequisite: enforce completeness before routing rather than after. Routing on incomplete records generates the rework measured above.

5. A caution on measuring the interface

Each of the four measures above is a candidate for the distortion Ridgway [4] described. Acceptance rate in particular is trivially improvable by accepting everything. Measure them, and pair each with a downstream outcome measured a quarter later — acceptance rate against subsequent conversion, time to first touch against subsequent win rate. Interface measures used without a lagged pair reliably become theatre.

6. Limits of this argument

Coordination theory [1] is a general framework rather than an empirical finding about revenue organisations; the mapping in section 1 is an application argued here, not a result. The marketing–sales orientation research [2] establishes that differences exist and that their performance effects vary, and this paper does not claim more than that. The navigation construct [3] establishes that intraorganisational effort is a real and substantial part of the sales role.

What is defended: the interface is better modelled as a set of dependencies than as a relationship, the dependency types are identifiable, and coordination cost is measurable on data already held. Whether it is large in any given organisation is what section 3 determines.