Communication Overhead

A Hidden Diseconomy of Scale

We Have Wi-Fi At Home

In what is probably a relatable struggle for many, I am the Wi-Fi technician for my house. We have a pretty simple set-up from the phone company – a Wi-Fi enabled modem on the first floor. Bad reception on the higher floors is unacceptable, so we connected to the phone wire going upstairs and set up two routers, each with their own Wi-Fi extender for total coverage. Being the one with the best idea of how the Wi-Fi works, I’m naturally the one asked when the internet is slow, or suddenly cut off, or if a password needs to be changed[1].

One of my most common responsibilities is to test the Wi-Fi for speed at the modem downstairs. We’ve had some issues before with the box outside the house connecting to the underground line being eaten through by rats or poorly connected, so the occasional speed test is an important part of maintenance.

Getting Internet Speed

The speed test shows me upload speed, download speed, and ping (in miliseconds). Upload and download speed are how fast information reaches and is sent out from my device, while ping measures latency – how long it takes data from my device to actually reach the server. Downstairs, our connection is rated for 100 MBps – that’s megabytes, not megabits. Upstairs and on good days, I usually get up into the eighties or seventies, as the family all connects to the closest router on the network, with everything coming back down to the single modem that’s actually connected to the Internet.

Speed, however, isn’t everything. Each individual modem and the connections between them has to deal with the data moving back and forth, creating latency in the connection. Latency is how much time it takes for the user’s device to communicate with the destination server. Once the information reaches the modem, it has to contend with bandwidth. Bandwidth is the carrying capacity of the internet connection going between your home modem and the rest of the Internet. Finally, there is the added problem of network speed – the speed at which the local Wi-Fi network transfers data from the last router and over the Wi-Fi to the device. This is based on both the strength of the signal and the Wi-Fi capabilities of the device. Taking all of these together – bandwidth, latency, and network speed, we get the metric of throughput, defined as the amount of information actually delivered during a certain period (after the effects of latency and network speed).

Maximizing throughput is about maximizing network speed and bandwidth and minimizing latency. Maximizing bandwidth depends mostly on your plan and hardware – the speed provided in your Internet plan and the frequency of the routers are the two main determinants of bandwidth. Reducing latency involves simplifying the network, cutting down the steps from device to modem to only what is necessary, as well as shifting internal connections to wired as much as possible. Network speed, by contrast, is recalcitrant – it depends on where and when devices are used relative to the extenders and routers, and while it is possible to change one’s habits, the locations of the routers can make it inconvenient to extend signals to odd corners of the house.

Multiple devices make this even more difficult, particularly mobile devices like tablets and smartphones. Now each individual node and its connections has to be stress-tested in order to prevent overloading, which means that the system has to increase capacity. In practice, most home Wi-Fi routers are made with plenty of devices in mind, so any issues from moving devices are more about what’s being done with them – too many people streaming video on one router can cause slowdown, for example.

Even with off-the-shelf solutions combined to set up a simple home wireless mesh, we can see that setting up Internet connectivity to satisfaction is complicated. Communicating a message can be just as complicated.

Failures to Communicate

Failure to communicate is a ubiquitous experience in modern life. Technical hiccups like accidentally replying all to a group email or dashing out a message to the wrong channel on Discord or Slack are simple. More fundamental miscommunications occur quite often. Code documentation and comments can be ambiguous or incomplete, procedures not fully documented, or an order from the boss can be quickly dashed out without much thought, and the subordinates scramble to understand the meaning of the message.

Sometimes, the results of failure to communicate can be disastrous. Korean Air once had an abysmal safety record, in part thanks to failure to clearly communicate problems due to cultural norms. On the subject of air travel, airports are a complicated dance of airplanes taking off, holding for landing, and landing at the airport, led by the air traffic controllers up in the tower. A combination of bad weather preventing visual checks and unreliable communications combined to produce the Tenerife airport tragedy - a collision between two Boeing 747’s right on the runway.

All this makes my home wireless mesh feel like no trouble at all. But it got me to thinking - how do we decide which way to communicate is best?

The Medium and the Message

There are two parts to every communication – first the message, which is the actual information to be conveyed, and the medium, which is the method by which it is conveyed. This blog, for example, is the message, conveyed to you through the medium of text. If I were to record myself reading it out, as Michael Woudenberg does for his articles at Polymathic Being, that would change it to an audio recording.

In my opinion, different mediums of communication can be rated along two important and mutually exclusive scales – richness and portability. These two scales are mutually exclusive, the former referring to the ability to convey more information in less time, while the latter referring to the ability to convey information repeatedly without adaptation.

The richness of a medium refers to its ability to communicate more information in less time - they are high thoroughput and data-intensive, meaning that they communicate very well, but are costly to transmit and broadcast. This usually occurs through the inclusion of contextual details. These include details such as the time of day, the location, or the speaker’s body language or tone of voice, all of which convey information in ways not directly part of, but integral to, the message. This is best exemplified by face-to-face communication. Both sides can directly observe facial expressions, body language, and tone of voice. In addition, conversations occur at a certain time and place, with particular people and in particular contexts. All of these things convey additional information without increasing the number of words actually said.

Rich media, however, are by nature difficult to replicate and transmit while preserving meaning. Repeating the same or conversation more than once, for example, changes the time, place, and other contextual details, altering the meaning even if the message is the same. Particularly in the time before video cameras, capturing face-to-face conversations or stage acts for repeat viewing was impractical. Old books and old speeches even come with additional modern forewords or appendices to help explain things that would have been obvious at the time of the work. Writing’s use to explain the context of old media shows its true use - as the most portable method of communication possible.

Portability is the direct opposite of richness. Portable mediums are low-thoroughput, but also data-light, allowing for much more to be stored and transmitted than rich mediums. Because of this property, they can explicitly include important contextual details that are difficult to include when rebroadcasting richer media.

All other things equal, maximal richness means a powerful message that cannot go very far, maximal portability means a weak message that can reach the ends of the earth.

Portability is exemplified by text. Writing involves the use of pre-agreed symbols (letters or pictographs) to signify sounds or concepts, agreed combinations of symbols being called words. Because of this pre-arrangement, the same message can be reliably repeated as long as the same words are used. In order to do this, we often make context explicit, setting the scene for future readers and providing information that they may no longer have (think the introductions of old books, or appendices and references going into further detail). The written word has this advantage because it can simply be repeated, word for word. Even in the digital age, this advantage holds, as writing is the least data-intensive form of communication.

All this, however, comes at a cost. Writing often becomes difficult to read without a writer actively cutting down on words, organizing paragraphs, and making sure sentences flow into one another. As the medium has to compromise between length and completeness, the richness of a message often suffers. This makes it inferior to face-to-face conversations for situations where rich and complete communication is required – it is much more suited to simple orders like shopping lists or meeting points. The best use of writing, however, is for records, as they are relatively easy to preserve, can include all pertinent information including the definitions of words as they were used at the time of recording, which can change drastically over time.

Choosing the right medium for your message consists of balancing the needs for richness and portability. Below is a rough list of mediums in descending order of richness.

Richness (descending order)

• Face to face communication

• Stage acts and speeches

• Video presentations or video calls

• Audio-only presentations or calls

• Pictures, pictographs, and diagrams

• Charts, tables, and graphs

• Writing (plain text)

As richness and portability tend to drive one another out, simply reverse the list when looking for the most portable form to the least portable form.

The balance between these two needs often shifts with the size, scope, and dispersion of the audience. While face-to-face communication is clearly the best form of communication in one-on-one situations, writing is the best form of communication for mass distribution and for therefore moving information across time and space. This causes what I call communication overhead in organizations – the progressive march of internal communications from rich to portable, and the accompanying consequences thereof.

Communication Overhead – Institutionally Poor Communication

Communication overhead is the phenomenon where the internal lines of communication in any organization become more strained as the organization becomes larger. This takes many forms, including corporate classics like sales promising more than operations can deliver, accounting double-checking your transaction for some abnormality, or simply that messages from the employees on the front line never reach the managers. I believe that this is fundamentally caused by two things - more nodes of communication requiring a greater quantity of communication and greater differences in context between each node of communication.

First, let us consider the one-man show – a business where one person is the nexus of all efforts. No communication needs to take place here, other than the classic stress phenomenon of someone talking to themselves, so communication overhead cannot be observed. As the person knows exactly what they themselves know, they can naturally communicate seamlessly with themselves.

Let us expand to a business where there are two people, say a boss and a secretary, or a master and an apprentice. There is likely enough time in the day for the two to regularly communicate face to face or via video call, which means that internal communication will be very rich. Communication overhead will arise in the form of misunderstandings or missing contexts - the usual misunderstandings between two people having a conversation. As they can still primarily use face-to-face conversation, these misunderstandings are still relatively easy to rectify.

Expanding to a small hypothetical team of the boss, an accountant/finance officer, an engineer, an operations manager, and the shop floor foreman, we begin to see communication overhead rear its ugly head. The team needs to develop a product, and each member of the team has their input to give, and the boss has to sign off on everything. Frequent, whole-team meetings to get things in order become impractical, because each person needs to see to their own function. First, far more communications now need to occur between the functions to ensure things run smoothly. The engineer and the accountant may need to meet on the feasibility of an item, the foreman and the operations manager to figure out the new plant layout, and evverything has to go through the boss - all of these approvals require separate instances of communication. Second, each person now needs to explain their domain of expertise to all of the others and how it relates to the product - the boss has to set the priorities, the accountant the finances, the engineer what can be done, the operations manager what they will need, and the foreman how they will do it. This adds even more to the message burden while increasing the probability of misunderstandings and miscommunications, as each field of expertise runs on different principles and uses language differently.

Meetings begin to run long and everyone suffers from information overload. Productivity suffers. Something must be done. The boss decides to have each individual department be the ones doing the meeting, and then reports will be written and due on the third week of every month. A meeting between all the department heads and the boss will occur during the last week of the month to discuss the submitted reports. This structure appears to be the best of both worlds, both creating a paper trail of reports for people to follow, and a meeting to voice concerns. This should solve the problem.

Most of the next meeting is then taken up by clarifications and questions on the reports made available via e-mail a week ago. Questions are lobbed in every direction, misunderstandings painstakingly resolved, and by the time everyone is on the same page, the meeting has run an our over schedule. The reports are universally panned as e-mail clutter - though they can be referred to, they became obsolete the day after the meeting.

How could this happen?

Writing reports and preparing graphs and charts is a skill of its own, and knowing what the relevant context to someone outside your field is another - neither of which was likely a skill prioritized when selecting the employee. This creates situations where people may write things nobody outside their team can truly understand – the more portable methods of communication require context to be explicitly stated, which may simply not be included since it wasn’t deemed worth mentioning. This impairs communication and ends up requiring duplication of effort as face-to-face clarifications become necessary.

If you try to imagine this at a multinational corporation, where you may have to write for people half-way across the world, educated and raised far differently from you, the level of context needed is staggering. It is more a miracle that anything gets done at all, rather than it being the expected outcome.

We can summarize Communication Overhead as follows:

As any organization increases in size, the frequency of internal communications and variance of its members follow suit, reducing the amount of shared context. Therefore, a greater need to use portable, but context-poor methods of communication arises to bridge the context gap. These more portable methods of communication, however, require a specific set of skills to relay context with. Combined with the greater volumes of communication and the greater variance of interpretations, an impairment in the quality, and consequently, the reliability, of internal communications within an organization must result.

This appears to be quite a limited statement, but I believe there is much to gain from it. As an example, explaining the great importance of communication and “soft skills” as a criterion for promotion. From the perspective of the line worker, it seems like a dodge - a simple way to pick the good old boys who know how to work the watercooler and go out drinking out of work - rather than an actual qualification. Certainly, the guy who bosses us around should be the one who’s the best at our job, right? We keep this company running, we deserve that much!

Consider, however, what a manager’s tasks. The manager of the team is its leader at work, setting schedules, calling others in when some call in sick, making sure to log everyone’s hours, but it is also its representative in all things. The manager’s unique role separating them from the line worker is in preparation of reports for and receiving orders from upper management, disseminating necessary instructions and information to their team, and providing feedback on new policies from their front line. In other words, their job is making sure everyone - their team, their management, and they themselves, know the score.

Line managers are human Wi-Fi extenders for the different departments.

The metaphor holds as we advance up the hierarchy to the departments. Finance and accounting ascertain and communicate the financial state of the company. Operations remembers, brainstorms, and communicates operational practices that have worked for the company, past and present. Marketing and Sales listen to find out what people like about you and repeats that to them, along with new reasons to like you. Human Resources’ job is to understand and respond to the needs of the employees - salaries, benefits, satisfaction and all. Information Technology’s role is to understand the problems of the other functions and communicate technological solutions.

Departments are the Wi-Fi routers that recieve orders from above and distribute them to the line managers.

Upper management receives reports from these departments. Their particular role is not merely to issue orders and assume responsibility, but to predict and prepare contingencies. Said predictions and contingencies usually take the form of plans - written communication. Observing trends in the market, in the economy, and in the world as a whole, classifying them into opportunities and threats, coming up with responses, and conveying them to the team - this is deciding what is important and what is not, writing a message, and deciding how to deliver it. This is communication.

Upper management is an incredibly well-paid Wi-Fi modem, facilitating exchange of ideas between the company and the rest of the world and distributing them to their individual departments, or, in the case of the Chief Executive Officer (CEO), to the company as a whole.

Communication skills are not a qualification for management. They are arguably the qualification for management. It may help to think of Management itself as a different department in its own right, whose task is to understand the situation of the company and coordinate responses.

This is your life in management.

Despite the length of this section, I am not done. Communication overhead offers insight into some intractable modern problems.

Why does everything large, even simple things, seem to take so long? The responsibility for them has been split across nodes of communication, meaning it has to suffer from communication overhead.

Why do new stars get promotions and raises rather than old reliables? The new stars effectively communicated their excellence to their new employers, and management responded to the new kid on the block speaking their language. Communication in action.

Why do our leaders and governments always seem to be out of touch? Any information they receive is filtered through the morass of communication overhead, easily third-hand or fourth-hand information.

Why do psychopaths always seem to end up in charge? Because they effortlessly tell us what we want to hear, and because of information overload on the individual level, we shut down and fail to question. This is covered more thoroughly in next week’s article, along with the one from the week after that.

I could continue rattling off answers forever, but I hope to save those for future essays, where I can treat the questions with the attention they deserve.

Fighting Communication Overhead

The steps taken to mitigate Communication Overhead are too many to number, and depend on the scale at which they are implemented. At the individual level are the techniques I lay out for dealing with the Information Superhighway. Here, I propose methods for mitigating Communication Overhead at the organizational level. These methods tend to fall under three general thrusts - devolving authority to reduce the number of people in the loop, institutionalizing matching mediums to messages, and agreeing on predetermined words and principles as shorthand. These three threads run through the suggested methods I give below:

Decentralization

Decentralization refers to the devolution of authority in order for decisions to be taken at a more local level. This allows for more problems to be dealt with closer to where they arise, allowing the smaller pool of stakeholders to share more context and therefore use richer methods of communication, leading to an improvement in the quality of communication. Large institutions having local or regional subsidiaries or the principle of subsidiarity in government are excellent examples. In extreme cases, an entirely new institution may be formed to deal with emerging difficulties that deviate from the original institution’s purpose. This approach, however, makes it difficult to standardize things across the whole organization, making it unsuitable in situations where standardization is necessary.

Historically, before communication became so frequent and accessible, this was the preferred method. Authority and discretion were devolved downwards as much as possible, because letters took far too long. Even large bodies such as the Catholic Church had more local bishops and priests with their own authority, and Islam to this day has imams - individual religious leaders each with their own take. Decentralization is alive and well today.

Mixed Methods

Mixed methods of communication involve splitting a message up into parts and delivering the parts via suitable methods. A key presentation, for example, could have three parts; a memorandum with facts and figures to reference during the presentation, a deck of slides to display the relevant ideas, and a simultaneous face-to-face presentation of findings allows for the strengths of all three forms of communication to be leveraged. The amount of work to deliver the report, however, at least triples, making this mitigation best-suited to the most important or delicate communications, where a misunderstanding could be fatal.

Message Relay

Message relay involves communicating a message to an individual different from the final recipient, and trusting that person to get the information to the other side. Taking inspiration from messengers or TCP/IP, both of which use relay systems to move messages long distance, message relaying involves each link in the chain deciding how best to present this information to the intended recipient[3]. For example, an e-mail report passed from an employee to their superior for comment, before the superior passes it on in turn to their superior. The weakness of such a model is that each communicator in the relay can modify the message, meaning that the original message may never reach the final recipient.

Jargon

Jargon refers to technical terms or meanings within a profession meant to improve communication within that profession. By assigning more precise meanings to words and creating new words for phenomena in the profession, jargon allows for concise and clear communication among members of that profession. An excellent example of this is Black’s Law Dictionary, which provides precise legal definitions for words - even words you would find in normal conversation. This solution is best used for documents meant to be written and read by members of the same profession only, as jargon makes interpretation by those outside the profession more difficult.

Routines and Procedures

Routines and procedures are deterministic chains of actions that form a process, and are bundled into a single entity. They are also the true human superpower. Rather than constraining the discretion of the listener by using loaded words, routine removes it entirely, by dictating which steps should follow based on the outcome of previous steps - like a factory assembly line or a recipe. This has the advantage of minimizing the context required to perform a task, but the disadvantage of becoming either insensitive to context, or so complex as to make the task impossible to perform. Striking the balance between these two needs is the central issue of designing a routine or procedure.

The best example is the use of standardized flowcharts for call center employees, and their cousin, the automated helpline. Many companies have transactions that are routine to such an extent that they can be summarized in a flowchart, neatly separating all incoming issues into categories of issue, each then with their own procedure or list of steps to be taken. This allows them to outsource these processes to call centers - large, centralized phone-answering operations that use those procedures to answer queries on behalf of the corporation that employes them. In theory, they should be able to provide good-quality customer service en masse, since the procedure covers all possible problems.

In practice, the procedure is neither perfect nor regularly updated, leading to problems with no solutions in the flowcharts exist. Linguistic ambiguity exists between the categories of transactions assigned by the flowchart, opening up the possibility of misunderstandings by the caller, which slow things down. In some cases the caller, out of frustration or impatience, attempts to short-circuit the process by repeating their issue, but louder this time, things get even worse. The call center agent, not being an employee of the company, does not have the domain knowledge an employee would. This renders them incapable of understanding and resolving issues not in the flowchart, forcing them to look around for the answer in the book.

When you get on the line with a customer service agent, you are talking through them to their flowchart. They have differing levels of drive to serve, knowledge of the flowchart, and experience at managing customers.

There are clear limits to the power of routine.

Organizing Principles

An organizing principle is a heuristic or rule that allows individuals to make decisions in a predictable manner. Organizing principles can be as simple as the Golden Rule. More commonly, you’ll find them in the form of corporate mission-vision statements that provide a shared lens through which to evaluate incoming information - if it helps achieve the mission and support the vision, it is good, if it doesn’t, it is bad. Organizing principles are different from jargon because rather than making communication easier, they make interpretation easier. It is much easier to read a coded message when you have a decoder ring.

The best example of these are widely-held assumptions, such as the infamous “rational man” of economics. The rational man is an organizing principle that holds that people will always try to maximize their gain or utility for every transaction, meaning that anything with a higher expeted value or a greater profit is automatically taken without question.

The fruit of this is that one can simply call something “rational” if they think it is something to be considered because it maximizes utility, and “irrational” if it is to be rejected because it does not. One word replaces an entire phrase and series of thoughts.

This ability makes organizing principles both the most powerful and the most abusable of the mitigations of communication overhead. I will cover this in the future article, when my thoughts on the matter are fully formed.

Conclusions

Mitigating communication overhead will make you a better communicator. Better matching the method to the message, and knowing which parts of the message need to be delivered in particular ways, is the communicator’s most fundamental skill. Developing this skill rewards you with smoother and clearer communication, knowing when, where, and how to deliver your message for maximum clarity and impact.

Good communication is always coherent.

Second is knowing to set the apopropriate context for communicating things. A superior of mine once told me that good feedback should be given in public, while bad feedback should be given in private. Private negative feedback allows the recipient to correct in private before it becomes a problem, while public positive feedback allows the recipient to bask in their hard-earned glory. In both cases, I would prefer to give that feedback face-to-face, so that the employee can read my body and facial language, can ask questions, and thereby know exactly where they stand.

Good communication is always considerate.

The idea of communication overhead implies that smaller institutions always have a potential advantage against larger institutions. Understanding trends in their specific space and communicationally nimble enough to quickly adapt to them, these smaller institutions are likely to be the true source of growth, being gobbled up by larger corporations to incorporate their unique niche. This will be a key theme for the next two Friday articles, so look forward to that.

Good communication is always adaptive.

Pretty hard for a “soft” skill.

---

[1] So far, the only things I’ve had to deal with are old and faulty cables, accidentally unplugged cables, or just shrugging and saying “someone’s downloading/watching a stream/on a call”.

[2] Even reading a newspaper on the toilet has been disrupted by technology.

[3] If you’ve ever asked your mom for permission so she could ask your dad for you, you’ve done this.