Tradeoffs - Convenience vs Control

I can’t drive a stick-shift car.

Now, it’s not like it would be physically impossible for me to get the car in drive. I’ve just never tried; every car I’ve ever had to drive has been automatic. Driving a manual was completely unnecessary.

And yet the most advanced versions of cars, in formula and other forms of racing, exclusively use manual transmissions, as do plenty of other engine-driven vehicles like tractors.

The choice between a manual transmission and an automatic one is centered around what I’ll call the “control vs. convenience” tradeoff.

The core of this tradeoff is automation. In every example, there’s a “manual” side and an “automated” side. In the car example, it’s a choice between manually shifting gears and having your car do it for you. In this case, it’s a lot more convenient to have an automatic transmission. But shifting the gears manually is still essential for other kinds of driving (like racing). This exposes the first element of the tradeoff: your decision should depend on your use case. You often don’t need a lot of control over the process (like driving in traffic), and in those cases, doing things manually is unnecessarily taxing. But where you could use a manual transmission where an automatic would be easier, the reverse isn’t true; if you have to drive a manual, an automatic transmission just won’t cut it.

In other words, a natural consequence of convenience is restrictiveness. This complicates the process of choosing between convenience and control.

Let’s unpack this with another example. In Fall 2018, I was taking a class on probability. As part of a homework assignment, we had to work out an extension of the Secretary Problem, with much of it requiring some computational tools. This left me a choice between using Excel (convenience) and using Python (control).

Excel offers a lot of pre-built functions and a fully designed structure: information is stored and called in cells. In this example, it’s the “convenient” option. On the other hand, I’d have to do some more manual work in Python, deciding how to store information (like in a dictionary or a list), how to organize the data, and which tools to use for the computation, while also offering more functionality. It will serve as our “control” option. Eventually, I opted to use Python, even if just to get some more experience under my belt.

I spent a few hours working on the assignment, doing some probabilistic analysis by hand and computation in the program. Because of the initial time investment, Excel started to look like the better choice.

But as I neared the end of the assignment, I got some wonky results: my graphs were wrong.

While I knew I’d made a mistake, I didn’t immediately know where. After some thought, I suspected I had made a mistake with my hand calculations (taking the derivative of a nasty equation). To verify this, I wanted to calculate and plot some numerical derivatives.

Now this was a fairly easy operation in Python. With the way I set up the data, I could easily use a library (shoutout SciPy) to find the derivatives. When I did that, it only took a few minutes of googling to figure out how to set up a sort of diagnostic test to see where I’d fudged my calculations.

But in Excel, where the tools are more limited, I’d be flying blind. Instead of using a simple and convenient test to see where I’d made a mistake, I’d have had to walk through the computations start-to-finish, deciphering long and convoluted formulas to figure out where I’d made my mistake. And if you’ve ever done something “quick and dirty” in Excel, you know that debugging is often more difficult than just starting over.

This case illustrated an aspect of the convenience/control tradeoff: future needs. As noted before, the convenient option has less functionality than the control option. By choosing the tool with a wider functionality, you give yourself more freedom in the future. Choosing a tool requires an analysis of your future needs, not just your current ones.

And if you find yourself trying to make adjustments, the convenient option can stop you in your tracks. While the Excel vs. Python example is compelling to someone like me, you’re really applying this principle whenever you choose between a disposable or reusable product (think razors, grocery bags, and water bottles). In these cases, the “convenient,” disposable option is cheaper and easier to manage (just buy a new one) while the “control” option is usually better quality, but also more expensive and more to keep up with.

We’ve already covered a lot of ground, seeing how this tradeoff affects the cars we drive, computational tools, and even the stuff we buy. And yet, it takes root deeper and more broadly in our communication.

Let’s use baseball baseball for this example. To someone who watches baseball, a “strike” is a simple concept.

It wasn’t until I had to explain what a strike was to one of my Lebanese cousins that I realized how complicated it truly is. If the player swings at a pitch and misses, it’s a strike. Unless they hit the ball outside an acceptable area, in which case it’s a foul ball, and therefore a strike. But only if it’s not the third strike; foul balls can’t result in strikeouts. And if the player doesn’t swing, but the ball is thrown in the acceptable zone, it’s a strike.

Complicated, right? As the game went on, I had adjust my communication to explain what was happening. “The player did not swing, even though the ball was in the acceptable pitch zone.” “The player hit it outside that white line, so it counts as a strike.” Just saying “it’s a strike” would’ve been a disservice to my unfamiliar company.

This particular manifestation of convenience vs. control is jargon. In cases where both speaker and listener have some baseline understanding, jargon makes the conversation more efficient and lively. But in cases where one doesn’t have that understanding, jargon fails to communicate anything at all!

Like other concepts related to our communication, this use of jargon mirrors a process that happens mentally.

When I see the result of a pitch, my brain makes an instant connection to the definition of a strike without walking through a mental flowchart I outlined above. Over time, I’ve gotten an intuitive, convenient understanding of what a strike is. It gave me plenty of time to know why a pitch was(n’t) a strike and look over to my cousin to see her mental gears turning, trying to explain the result. In Thinking, Fast and Slow, Daniel Kahneman explains this as the difference between System 1 (my) thinking and System 2 (my cousin’s) thinking. Without going too deep into the nuances of the book (which you should definitely read for yourself), System 1 is quick, intuitive, and accessible where System 2 is both far more powerful and far more taxing. In case it isn’t clear yet, System 1 is the “convenient” option and System 2 is the “control” option. Now in the baseball example, the tradeoff is purely dictated by previous experience. In a way, my cousin was restricted to the “control” option because of her lack of previous experience with baseball.

That’s not how it usually goes.

In most examples, we rely on the convenient, intuitive type of thinking to answer questions that need a little more rigor. Next time you tell someone how long it takes to get somewhere, note how much you’re really thinking about it. Are you just spitting out a convenient number? Or are you deliberately making justifiable assumptions and using rigorous methods? Whatever you do, you’re balancing the convenience vs control tradeoff.

Like other tradeoffs, this one extends into organizational behavior. And with that, it brings some new challenges to consider.

The first of these deals with outsourcing tasks. To illustrate what I mean, I’ll use two organizations on opposite sides of this tradeoff.

The first is the “convenient” method (outsourcing everything). One of my friends is about to wrap up an internship for a small company (heretofore known as “the bug spray company”) that makes pest control products. Including him, the company has 4 people. Now they don’t actually make the sprays; they figure out the formulas and outsource manufacturing to other companies (who we will call “specialists”).

When you think about the company’s size, it makes sense. It’s only a few years old, and with so few employees, there’s just no room for manufacturing management. In their position, independently manufacturing is neither feasible nor worth it.

On the other extreme, we can take a company like Google. From what I’ve gathered, Google makes most of its software tools in-house. This means they have their own programming languages, their own version of Stack Overflow (a site where people can ask questions about programming), and programming environments. A sizable portion of Google’s employees focus solely on internal tools. This helps Google ensure they have customized tools for all of their needs. In many cases, these custom tools are better designed to fit Google’s particular culture and structure.

So how should organizations find their balance of convenience and control?

On the surface, the “control” option seems to be the best way to run your organization. For companies, this cuts out the specialists and can help lower costs. By outsourcing manufacturing, the bug spray company is paying both the specialists’ cost and their profit. But if they could manufacture for the same cost, that profit would stay in house.

The tradeoff manifests here in an initial cost. Buying your own equipment or creating your own tools requires a large up-front cost. For companies that don’t have a lot of cash stored up, this can be a really big risk, especially when the company’s members have vested stakes in the venture. But beyond the money, it costs a lot of time. In the quick, growing phase of a company, this can take the focus away from the mission, vision, and products of the company. Then there’s the assumption that a company can run as efficiently as the corporations it outsources to. This assumption is typically only true for very large corporations, who benefit from the same economy of scale as their specialists.

A convenient way of visualizing this is with trees. Very few trees can support large branches when they’re saplings. As they grow, some species (like oak trees) may grow branches large enough to be new trees.

Oak Tree

And yet, others (like palm trees ) grow branches small enough to be considered part of the leaves.

Palm Tree

The choice to outsource to specialists also makes long term impacts on an organization’s inertia, or tendency to stay the same. In a general sense, the “control” option has less volatility; by building your equipment/software/method yourself, you have the ability to control its changes. The “convenient” options are more subject to volatility. For outsourcing, that means that your specialists’ financial health has an enormous impact on your own. And when it comes to using off-the-shelf tools, changes in those tools can require painful and inefficient changes for the organization. In college, I saw this in SolidWorks. As SolidWorks releases a new version each year, they don’t preserve forward compatibility. In other words, a 2019 file cannot be run on a 2018 application.

And if SolidWorks ever goes out of style, storing data in the SolidWorks file format can bring a company’s compatibility to its knees. While it’s not likely to happen, this kind of reliance on a “convenient” tool adds another layer of volatility, and with it a risk for catastrophic consequences.

Of course, this volatility comes with some organizational flexibility. By outsourcing to specialists, an organization can stay on top of the changes happening in a field. For example, if the bug spray company discovered a new manufacturing technique with a more efficient or sustainable method of manufacturing, it would take a lot less effort to switch specialists than it would to retool an entire factory. So while the “control” option (as in the Excel vs Python example) can build in more flexibility for an individual project, tying resources to it can severely slow down an organization.

Finally, organizations have to think about accessibility. Because the convenient option is easier, it usually requires less specialized skills to use. For a project touched by many people, this can be a big advantage over the control option.

While this tradeoff takes many forms, identifying it is typically the most difficult part. The method for responding is both simple and uniform.

The first step is what I’ll call “needs and capabilities.”

As the name suggests, this analysis starts with “needs.” This should include the current needs of the project and potential future needs, as well as when those needs may materialize. As the case extends to a longer term, this analysis shifts to estimating the likelihood of certain needs. Further, there’s always the option to start over in the future. In other words, this needs analysis needs to identify what this case may require with reasonable likelihood.

After you’ve figured out what your needs are, you need to determine the capabilities of each option. In the simplest sense, you’re figuring out what each option can and cannot do, and ranking them from most convenient to most powerful. This should include some analysis of accessibility and the various tradeoffs outlined before.

Once you’ve done these, the decision is formulaic: pick the most convenient option that has enough capabilities to satisfy your needs. While the technique is the same for every instance of the tradeoff, you should obviously devote different amounts of time to needs and capabilities depending on the decision. For example, a decision you make often (like choosing between using a reusable grocery bag and a disposable one), you may just establish a default choice. On the other hand, larger scale and less common decisions (like choosing to work for a company or start your own) may require a lot more thought.

While I didn’t want this post to be so long, I think the sheer number of examples illustrates how widespread this tradeoff is. And with each version of it comes a set of challenges. To summarize everything, I made a little bulleted list with the key points.

For Individuals:

• Convenience option just won’t work in some circumstances, where the control option may just be a little more difficult (can’t do donuts in an automatic transmission)

• Sometimes you have unexpected needs that make the convenient option intractable later (like using Excel for that probability assignment)

• This tradeoff underpins everything from using reusable grocery bags to jargon to intuitive vs deliberate thinking

For Organizations:

• Clearest example is outsourcing (think Bug Spray Co vs Google)

• Outsourcing requires less initial investment but often results in a suboptimal fit

• Building in-house insulates you volatility, but also ties up resources in practices that require more effort and take longer to keep up with the state of the art than a specialist