Interfaces Between Departments: Where Information Gets Lost

Why information gets lost at the interfaces between departments

During the sales conversation, the customer mentions that the machine must not be taller than 2.80 meters because of a hall ceiling. The salesperson notes it. In the quote it appears as a footnote, in the order confirmation it is gone, and the design team plans the standard height. Nobody did anything wrong, and yet the information disappeared on its way through three departments.

Such cases are not a sign of poor employees. They arise because each department has its own goals, its own tools, and its own idea of what is important. The more a company grows, the more such transitions there are. The obvious answer is more information: more fields in the system, more distribution lists, more coordination meetings.

This article asks whether that is right: what prevents information loss between departments better, more exchange or more shared responsibility? To answer it, we first clarify what an interface is, then present a thinking model from research on cooperation between departments, examine the findings, and derive what an owner can do.

Information does not get lost because it is missing, but because nobody on the other side feels responsible for it.

What is an interface, and why does it work like a game of telephone?

An interface is any point at which work or information passes from one organizational unit to another. Typical interfaces in a small or midsize company lie between sales and order processing, between order processing and production or assembly, between production and accounting, and between customer service and sales.

The children's game of telephone describes well what happens there. Each station passes on what it understood, not what was said. Each leaves out what seems unimportant to it, and in the end a message arrives that nobody sent in that form. In a company, each station is also under time pressure and transfers the information into a different tool.

The chart shows a pattern many companies know from complaints. The largest loss rarely occurs at a single point. It is spread over several transitions, at each of which only a little is lost.

The thinking model: interaction and collaboration

In the 1990s, the marketing researcher Kenneth Kahn proposed breaking down cooperation between departments, called integration in the literature, into two components. Interaction covers the formal, tangible contacts: meetings, reports, phone calls, exchanging documents. It can be mandated and counted. Collaboration covers working together: sharing goals, exchanging resources and ideas, jointly taking responsibility for a result. It is voluntary and cannot be decreed.

The model is useful because it reveals a confusion that happens in many companies. Anyone who responds to coordination problems with more meetings and distribution lists increases interaction. Whether collaboration improves as a result is another question.

What studies show about cooperation between departments

Ellinger, Daugherty, and Keller (2000) tested the model in a survey of logistics managers in U.S. manufacturing companies. Of 360 managers who had agreed in advance, 309 responded, a rate of 60.4 percent. They were asked about three forms of cooperation between logistics and marketing: information exchange, consultation, and collaboration. The result contradicts the obvious assumption. More information exchange had no measurable relationship with the perceived quality of cooperation. More frequent consultation was even slightly negatively related to it. Collaboration, by contrast, had by far the strongest positive relationship.

The authors conclude that simply obliging departments with different outlooks to exchange more information achieves little. Reports and documents that the other side does not consider useful change nothing. A good relationship between the departments was in turn associated with better delivery performance, and that with better company performance. A direct relationship between a good relationship and company performance did not appear. The effect therefore runs through performance for the customer.

A side finding is remarkable. The researchers asked the logistics managers to have a colleague from marketing fill out a short questionnaire. 30 pairs came back. On collaboration, the two sides agreed on only one of seven items. The authors read this as a sign of missing integration, especially since the 30 pairs are likely to come from companies where cooperation is better than average. The two sides of an interface therefore often experience it very differently.

Kahn and McDonough (1997) surveyed 500 managers from marketing, manufacturing, and research and development in companies from the United States, Europe, and the Far East. Their results point to a relationship that holds across the regions and in which collaboration plays a central role in success. And Guillén and Kunze (2019) found in data on 305 project managers that collaboration across department boundaries went along with more innovative behavior and better appraisals.

Two readings: form or attitude?

Two conclusions can be drawn from the findings that seem to contradict each other. The first is: formal handoffs, fields, and meetings achieve little, what matters is attitude. Whoever wants to cooperate finds a way, and whoever does not ignores every form.

The second is: precisely because collaboration is voluntary, it needs a form that makes it likely. A handoff at which both sides sit at one table is interaction, but it creates the occasion for collaboration. A metric for which both departments are jointly responsible is a formal rule, but it makes the shared goal visible.

I consider the second reading the practically correct one. The studies do not show that form does harm, but that the sheer quantity of information is not enough. The consequence is to design interaction so that it triggers collaboration: fewer reports that get sent, more short handoffs where questions get asked. The return rule from the article on how to optimize order processing is a good example. An incomplete handoff sheet goes back the same day. That is a formal rule, and it forces both sides to talk about the content.

How to check the interfaces in your company

1. List the interfaces

   Draw the path of a typical order from inquiry to invoice and mark every point where another department takes over. In small companies there are usually four to six.

2. Ask both sides separately

   At each interface, ask the giving and the receiving side separately: what has to come along? What is missing most often? How well does it work on a scale of one to five? Different answers show where the interface is experienced differently.

3. Define the handoff

   For each interface, define which information has to come along, in what form, and who is responsible. Where responsibility is unclear, the RACI matrix helps.

4. Choose a shared number

   For the most important interface, set a metric for which both sides are responsible, such as the share of orders without follow-up questions after the handoff.

5. Clarify mistakes together

   When something gets lost at an interface, both sides clarify together how it happened. Not who, but at which station and why.

For the handoff from sales to project teams, the article on the handoff from sales to the project team describes a handoff sheet with ten points. For regular meetings between departments, what the article on fewer meetings and better decisions describes applies: a meeting without a decision is usually interaction without collaboration.

What follows beyond the individual case

As a company grows, the number of interfaces rises faster than the number of employees, because every new specialization creates new transitions. A company with ten people in which everyone knows everything hardly needs handoffs. A company with 60 people in six areas cannot deliver reliably without them.

This results in a task for the owner that is rarely named. The owner is often the only one who oversees all interfaces and therefore the one who can set shared goals between departments. Collaboration cannot be ordered, but you can create the conditions for it: a shared number, a fixed occasion, a clarification without blame. Especially in distributed teams, where hallway conversations are missing, these conditions become more important, as the article on how to organize distributed teams shows.

• Every interface along the order path is known and has a responsible person on both sides.
• For every interface it is defined which information has to come along.
• At least one metric is jointly owned by two departments.
• After mistakes at an interface, both sides clarify the cause together.

Conclusion: more information or more shared responsibility?

The question at the start was
better prevents information loss between departments. The findings favor shared responsibility. In the survey by Ellinger and colleagues, more information exchange had no measurable effect, collaboration the strongest. That does not mean form is superfluous. A handoff with fixed information and a return rule is what creates the occasion at which both sides talk to each other. More reports, distribution lists, and mandatory fields, by contrast, increase the quantity of information, not its effect. At the interface, what counts is not what is sent but what the other side understands and considers part of its own job.

The next step

1. Take the last complaint or rework and trace back at which interface the crucial information was lost.
2. Ask both sides of this interface separately how well the handoff works, and compare the answers.
3. Agree together on which three pieces of information must always come along with this handoff.
4. Agree that for larger orders the receiving side summarizes the handoff in its own words.
5. Choose a metric for which both sides will be jointly responsible from next month.

A single well-designed interface often changes more than a new system. It shows both sides that the other department is not a recipient of documents but a partner for the same result.