SMART Goals

When I went to UC Irvine, one of my first classes introduced SMART goals as a way to begin planning out our educational journey. I originally intended this article to be about making a potential productivity tool more accessible. To some extent, that will still be true.

However, after I graduated, I noticed some recurring patterns in the way number games are played in every institution (go watch The Wire and save yourself the firsthand reckoning with this idea). In school, we have grades and standardized tests. In higher education, rankings from media companies like US News. In research, we have a publish or perish culture and the h-index. In law enforcement, we have arrest and ticket quotas. In corporations, we have quarterly reports and share prices. In politics, we have poll numbers. In marketing and public relations, we have Q scores. In news and online content, we have clicks, likes, shares, and other engagement metrics. In the tech job market, you can go to Team Blind to see absurd competitiveness over compensation packages and levels. In healthcare, we have 30-day readmission rates and surgical report cards.

So, looking back at SMART, the way it emphasizes objective measurements tickled something in my brain. How do we reduce complex educational, career, and research goals into something that demands measurable milestones? It would be strange to hire a researcher and demand they have a cure ready for a disease that works 75% of the time with incremental improvements in the success rate by 5% once a month. It might make sense for a weight loss regimen though; losing 1-2 lbs a week by running a 500-1000 calorie deficit that you calibrate yourself to over the course of a month seems reasonable. Clearly SMART isn’t nonsense, but I wondered how this system arrived in an academic environment.

About Management Literature

It’s always difficult to analyze the leadership of a community from within the jurisdiction of the institutions and networks they lead. Victors tend to write history in a way that justifies their victory and deters future challengers. Because our instruments to perform research are developed, operated, and funded under the very leaders we aim to analyze, appropriate data collection is difficult. Conducting experiments scientifically, even more so.

For instance, how could we take a robust and fair data-driven approach to evaluating the personality profiles of Presidents? We have such a small number of them, and they infamously have been selective about sharing unflattering evidence from their archives with libraries and museums. Some biographers approach Presidents with the goal to knock them off a pedestal while others seek to polish it, and autobiographies do little more than indicate what kind of narrative around their legacy they wanted history to remember. Any President who implemented a review of Presidential personalities might reasonably get the criticism that anyone who wrote that report would be conscious of the current President being their most important reader.

The literature around leadership has to be understood through the dynamic of someone under the leadership who exists in some relation to it. Everyone who writes about the President has an opinion on the President. Even when we do retrospectives, we still exist in some relation to the precedent, history, or legacy left behind. This isn’t unique to topics pertaining to leadership, but it’s especially important to remember it in management literature where leadership is the primary subject.

Scientific Management

From about 1870 to 1910, America was experiencing a Second Industrial Revolution as railroads connected national markets, electricity became more commonplace, and corporations evolved into hyperscaling entities. What seemed like a manufacturing economy built on a network of small elite workshops with educated engineers began transitioning into giant factories that organized thousands of laborers without specialized training who composed individual parts of assembled products. The challenge from the perspective of factory managers and corporate owners was organizing large numbers of workers, which they initially framed as engineering and administrative hurdles requiring technical innovation.

In a standard factory, workers are incentivized to seek more pay and fewer hours, and by forming unions that could strike to shut down all production, they were in a constant state of negotiation with management for resources. The 1877 Great Railroad Strike and Haymarket affair especially put some urgency into stabilizing manager-worker relations. Frederic Winslow Taylor, an engineer coming from the Mayflower pilgrim families which were educated at Harvard and later owned factories, saw the problem primarily in the way workers were incentivized to throttle their own productivity out of fears of layoffs or increased quotas. If workers and management couldn’t measure productivity, they couldn’t come to a stable agreement about production. He put forward The Principles of Scientific Management to argue that informal relations and manager intuition should be replaced with measurable standards that aligned with principles applied through the scientific method.

This also abstracted the concept of a worker into interchangeable mechanistic components of a factory, away from treating interpersonal relations, reputation for intuition, and trust as vital capital for a corporation as a group of workers. “Why did you fire my failson?” a part-owner might ask a manager. “He’s 2 standard deviations below his coworkers” a manager might be able to reply. This sounds very good in theory, maybe even the inklings of a meritocracy and competitive market. This was likely Taylor’s vision at least, as he genuinely seemed to want a way for managers and workers to stop being at odds and thought measurable standards would help.

However, the other side of that coin is asking “why did you fire 15,000 people,” as a manager might reply “the LLM I procured seems to do equivalent work according to our key metrics.” A manager without the reassurance of those metrics might be far more hesitant to cut the institutional experience and interpersonal networks that were slowly built in a company’s culture over decades for experimental technology. Workers, without the reassurance of those metrics to give feedback on their relative competitiveness, might be more willing to act cooperate collectively while appreciating their unique strengths and weaknesses.

Peter Drucker

Peter Drucker was an Austrian economist who immigrated to the United States in the midst of the rise of Nazis in Germany, so he was highly motivated by an appreciation for institutions that kept democracy healthy as a bulwark against authoritarianism. He was asking questions about institution resilience and class interests alongside the focus those like Taylor had on the industrial productivity side. When he approached a theory of businesses, he did so as a sociologist with a social philosophy rather than an engineer or economist trying to tailor a technical solution.

Drucker praised Taylor, but also saw that 20th century organizations didn’t just consist of managers and workers haggling it out on the factory floor. We now had accountants, statisticians, educators, lawyers and paralegals, engineers and mechanics, historians, psychologists, designers with art portfolios, and other professions where knowledge was the primary operating capital rather than machinery. Managers in these situations wouldn’t always, or even ever, have the requisite knowledge, skills, experience, judgment, intuition, creativity, cultural alignment, or influence networks to be productive. It wouldn’t be possible for a manager to prescribe a set of protocols to manualize and operationalize for these roles in order to scientifically evaluate them. Instead, Drucker proposed that managers should declare objectives and allow knowledge workers to decide the best path forward. Managers had the role of translating institutional missions and principles into objectives, and workers were more free to tackle problems through their increased autonomy.

I feel like it’s worth commending Drucker’s perspective. He foresaw large organizations as the primary institutions through which people would operate as citizens with authority, responsibility, and community. Rather than Taylor’s emphasis on optimizing operations according to objective measurements, Drucker’s emphasis was in aligning a worker’s autonomy and responsibility with an organization’s broader purpose. He actually coined the phrase management by objectives and self-control. By shortening the phrase, we actually miss the entire point. Management by Objectives alone implies more centralized planning through management supervision, but Drucker’s goal was to reduce the need for top-down supervision by increasing worker autonomy. He encouraged managers of all levels to be involved in setting the organization’s mission, purpose, and objectives, which displays a recognition of the merits of democratic governance in integrating the varied objectives of an organization’s members.

He popularized MBO through The Practice of Management (1954) by outlining a system of management that aligned individual goals with the goals of an organization alongside a structure for continuous evaluation and rewards. Throughout the 1960s and 1970s, Drucker maintained relationships with corporate executives and continued to work as a consultant, so his theory was able to come out of academic circles and into actual implementation at upper leadership levels. The system proved to be very popular among corporations that turned into giants. Hewlett-Packard, Xerox, DuPont, and Intel all credit the system with helping manage the growth of professional networks and organizations around the world.

Predictably, Drucker’s theory was distorted into Management by Objectives with the self-control portion becoming a very distant concern. Managers certainly liked the idea of being able to set objectives that professional workers had to achieve without needing any idea of how the profession or the subject matter actually works. Getting to be hands-off while hiring workers with a higher ceiling on productivity seems like an undeniable win. The problem lies in letting go of power so workers can make those decisions instead of trying to hit the numeric objectives that management set. Without actually doing that part, we just end up at Taylor’s original problem where workers are concerned primarily with doing exactly as much work as they need to keep their paycheck.

William Edwards Denning

Actually, we end up at somewhere possibly worse than square 1 with Taylor.

William Edwards Deming was a contemporary of Drucker who shared many of the same premises: America had unprecedented economic advantages after WWII and enthusiastic workers incentivized to be productive and pursue upward mobility, but poor management was throttling America’s ability to sustain that advantage. Large organizations and their management were likely to become the primary places of decision-making in American life. As Deming said, “I should estimate that in my experience most troubles and most possibilities for improvement add up to the proportions something like this: 94% belongs to the system (responsibility of management), 6% special.”

While Drucker initially pursued the alignment of individual purposes through collective objective-setting, Deming took an approach that was much heavier on systems thinking, perhaps as a result of drawing on some of the work out of Bell Labs. His argument was that managers determine the constraints for the organizations and therefore the constraints for the workers. Equipment, manuals, communication, incentives, trainings, suppliers, ethics, team structure, morale, public branding, and culture are all driven by managers setting parameters for a system that workers navigate. Thus, to Deming it seemed obvious that most variations between organizations are the result of the system designed by management, not the methods of evaluating individual workers for performance and productivity. Drucker did acknowledge later that organizations almost never implemented the self-control part and distorted his theory to continue centralized planning.

One of Deming’s most popular quotes: “a bad system will beat a good person every time.” When you reward someone for hitting a target, they will optimize hitting that target. If you run a door-to-door canvassing operation and you tell each canvasser they have a specified number of doors to hit, they will speed through them with low-quality pitches. If you demand a certain number of deals to be closed, salespeople will seek low-quality deals that hit the assigned quota. If you set quarterly production numbers, factory managers will sacrifice long-term maintenance and worker health; we’ve all heard the horror stories of people dropping dead in the middle of a warehouse while their coworkers are told to keep going so they can hit the magic numbers. Hospital administrators and insurance companies alike might be incentivized to shorten the length of a patient’s stay at a medical facility to report that they’re better; readmission due to premature release means they can report that they handled two visits instead of one longer visit. Teachers end up teaching to the standardized tests when it starts to reflect both on their job performance and their students’ performance. Just watch The Wire.

Drucker and Deming arrived at similar ideas from different positions: management consistently fails at systems thinking and attempts to compensate by holding individuals up to metrics instead of attempting systemic reforms. This shouldn’t be too much of a surprise if we consider the historical context of their lives. Both grew up to witness three leadership paradigms: fascism, Soviet-style communism, and American corporatism. All three pursued centralized planning and totalitarian monopolization, and all three seemed riddled with preventable patterns of failures from management. Demings called out a phenomenon of managerial tampering, where managers have a tendency to interpret normal variation in a process as signals instead of noise, especially since they often lack the subject matter expertise of their professional workers. Not only does this lead to fast and expensive pivots in policy, but it makes it almost impossible to have useful data about the state of the organization and efficacy of its policies. He thought anyone could juice short-term profits without any strategic consideration for organizational health.

Cheating in a Rigged System

I can’t really blame students for using LLMs when “writing” is something you get graded on from a 1 to 100 scale (or a 1 to 9 for AP testing, which is even dumber), and researchers are happy to haphazardly throw out new indexes that make LLMs sound like they have the general intelligence to pass tests. We’re actually training our students to calculate the expected value of cheating more efficiently than ever before: weigh the promise of a passing grade from an LLM against the probability of getting caught for slightly changing the answers, which is how traditional cheating on homework worked anyways. You can’t prove anything came from an LLM. Any professor or teacher that’s using an AI-checker is almost certainly buying snake oil from the other direction and unfairly damning students over it. As time goes on and LLM-generated content floods the Internet, people will naturally adopt that style of communication. Non-native speakers will especially see it as an example of standard English.

If you tell the student they’re depriving themselves of an actual education that would develop critical thinking skills and produce an enlightened active citizen, they’ll just think you’re out-of-touch with a changing world in which LLMs are being praised by the leaders of our wealthiest and largest bureaucracies as an innovation equivalent to discovering fire. If they’ve really spent some time rationalizing it, they’ll tell you that rich families can afford private tutors who do the homework assignments for their kids and private schools that look the other way, so LLMs are leveling an already-rigged game. We can chastise them for their solution, but they’re not wrong about the problem, so we better have better alternatives to give them. It’s a different version of the problem of STEM bros denigrating the value of social sciences and humanities; students can clearly see that even if tech is moving fast and breaking everything, STEM has a much higher chance of giving students an escape from the financial struggles of other fields that are clearly not being compensated for their education level. They’re just adopting the values inculcated into major institutions by leaders, same as it’s always been.

MBO and SMART

The basic outline of MBO is simple:

• Review organizational goal
• Set worker objective
• Monitor progress
• Evaluation
• Reward

In 1981, George T. Doran published *There’s a SMART way to write Management’s Goals and Objectives,” which was attempting to address a very narrow part of Drucker’s idea of an MBO since managers struggled to articulate individual objectives under a larger agreed-upon purpose. They could say “improve communication” as an objective, but they were often stumped about how to break that down. Having something like “reduce the number of miscommunication incidents by 50% over the next 6 months” helped people understand the scope and proxies for the problem. Doran was proposing a starting point that was not intended to be a formula and was still intended to operate within Drucker’s larger context of determining an organization’s mission.

The actual MBO process is very simple. The hard part is actually getting enough people in a room to agree on an endeavor they all consider worthwhile, where everyone’s objectives could be negotiated and reconciled. Obviously, there was overall very little interest from upper management to part with centralized power and hand autonomy over to subordinates. Frameworks like SMART became desirable as ways to state objectives but retain the ability to produce and monitor metrics. From management by objectives and self-control, we arrived at what Demings termed management by numbers. If you’ve been around the Internet for a hot minute, you’ve likely bumped into someone quoting (paraphrasing, technically) Goodhart’s law: “When a measure becomes a target, it ceases to be a good measure.” Demings argued we should eliminate MBO altogether and aim for something more systems-oriented. Drucker remarked “Management by objectives works if you first think through your objectives. Ninety percent of the time you haven’t.” He didn’t abandon MBO but he certainly thought almost all cases of it in practice failed to honor the original ideas.

Education and SMART

In the 1990s and 2000s, goal-setting and self-regulated learning methods grew as trends in the education space that encouraged students to take more initiative over their own learning process. Learning to learn is a good premise. I was around to remember the No Child Left Behind ads on PBS and the grand speeches about teacher accountability, though I didn’t see how much it was shifting blame away from a poorly funded system that elite families are exempt from and onto underpaid individuals burning out quickly.

SMART rode the coattails of those trends in the late 2010s and early 2020s, which is the range in which I encountered this class at UC Irvine. Education had its own Outcomes-Based Education (OBE) framework for designing content around demonstrable learning objectives rather than delivery itself. SMART is one possible way to structure this, though it’s not a dominant form by any means. In fact, it seems to have gotten its footing in education through a desire to treat public organizations more like corporations with measurements for individual administrators and teachers. It trickled down to students after educators became familiar with it off the professional development circuit, so it doesn’t really come out of a tradition of pedagogy or education research as much as a practical corporate tool that float across other institutions. I imagine at some point, the idea that it’s good for kids to get used to industry processes was a rationale for someone.

The rest of this page is lifted from the cheerier and shallower listicle-esque version I drafted. Looking back, I can see that I was running out of enthusiasm for it as I went down the letters. Oh well. Here it is.

S: Specific

The 5Ws and How is a good starting point. Some examples:

Who is working on the goal? Do you have partners or stakeholders outside a team? Who are you accountable to for the goal? Whose help or cooperation might be needed? Who is impacted by your actions?

What are you setting out to achieve? What actions will you take? What do you need in the way of help or resources? What resources do you already have? What constraints do you have?

Where are you going to be working on this goal? This one may seem less important than the others, but visualizing yourself working can help you brainstorm about the other questions. It can also help keep you committed to your goal by associating somewhere you’ll be with the work you have to do. Our brains are built for pattern recognition and you can preemptively prime yourself with cues in the future.

When are you going to work? When do you need the work done by? When do you have to issue progress reports? When do you have any other time-sensitive constraints kick in?

Why are you working towards this goal? Why do you benefit from this goal? Why do others benefit from this goal?

How are your actions going to further your progress towards this goal?

M: Measurable

How will you track progress towards the goal and if the goal is met? Having clear criteria for your goals lets you check the pace of your work, design actionable plans, and stay motivated as you see your goal get closer.

If your goal is simple, like a certain amount of money to earn or weight to lose or paintings to make, then you can monitor your progress in a straightforward comparison across time. If your goal is long-term or complex with multiple stages, you might consider breaking it up into milestones that each have their own SMART goals.

Measurement can be tricky. For example, if your goal is a coding project, you might be tempted to use the number of lines of code you’ve written as a way to track progress. A lot of companies used to, and some still do, track lines of code or number of changes to a codebase as proxies for productivity. However, good code is generally concise and easily understood (sometimes to an extent we would call self-documenting), and lines of code don’t correlate with complexity or impact. You could add tons of unnecessary variables and methods to pad out the line count, or you might expand a loop into a longer version of itself. We have to be careful and challenge our assumptions about our metrics to make sure we’re measuring something meaningful. Consider a checklist of features that can be tested with users (not your mom) where you set a goal for bugs reported or survey feedback.

A: Achievable

Is the goal doable? Do you have the necessary skills and resources?

Imposter syndrome will mess with you here, especially because in education, you don’t know what you don’t know until you try to know it. It’s a normal, healthy, and expected outcome of learning that questions will outpace answers. Just remember that you haven’t failed just because you have to readjust your plans after checking in with yourself.

R: Relevant

How does the goal align with broader goals? Why is the result important?

Contextualize this goal within the broader scope of your life. Consider how your values and commitments align with this goal. As a general rule of thumb, I think you tend to attract more of the things you do. Leadership opportunities lead to more leadership opportunities in an organization, networking with someone leads to more networking with people like them, certain resume items will resonate with certain hiring managers, so on and so forth. If you’re in school, you’re in some of the most optimal years of your life to make a pivot in a new direction with minimal friction. Don’t let the hypercompetitive nature of our rat race dissuade you from starting fresh so long as you’re still planning a sustainable life for yourself.

T: Time-bound

What is the time frame for accomplishing the goal?

AWhether you’re motivated by a hard deadline like a project due date or keeping a reasonably challenging pace, you probably want to have some rough idea of a schedule for completing parts of the goal. If you have a year-long goal, you should check-in with yourself month-to-month or so to make sure you’re further along enough than you were before. It can also help to share your goals with other people as well for some accountability and validation. Social commitments and a sense of urgency are natural sources of motivation for people.