Optimizing for the Middle vs Optimizing for the Tail

Hi everyone, 

Most of my Marketing BS essays connect to current news and events. Today’s piece looks at an evergreen idea: the value of “doing the right thing.”

This week’s podcast features Charbel Zreik, the managing principal at Manifestation Capital. In 2012, Charbel founded a search fund, purchased DCI (a hospitality-focused telecom company), successfully turned it around, scaled it to a national footprint and then sold it. He now advises entrepreneurs attempting to scale similar-sized businesses. 

—Edward


Quake Books 

When I interview guests for the Marketing BS podcast, I always ask about their “quake books” — the ones that fundamentally changed the way they look at the world. The idea of “quake thinking” isn’t strictly limited to books; your perspectives could be shifted by blog posts, news articles, lectures, podcasts, or documentaries. I’ve yet to hear about a “quake painting,” but it’s not inconceivable.

Here are two of my quake books that I’ve held in my head for a long time: 

  • Guns, Germs, and Steel: The Fates of Human Societies — evolutionary biologist Jared Diamond’s transdisciplinary perspectives about human history

  • How Brands Grow: What Marketers Don't Know — Byron Sharp’s evidence-based look at advertising, price promotions, loyalty programs and more

Last October, I stumbled upon a new piece of writing to add to my quake list: Brian Lui’s essay on a concept he calls “upside decay.” (Even if you don’t normally follow links from this newsletter, I highly recommend you check out this one — it’s a very quick read). 

Lui’s piece did not fundamentally change the way I see the world, so much as it provided a helpful vocabulary and mental structure for an idea I have long believed — the risk of “optimizing for the middle instead of the tails.”

Lui defines upside decay as the outcome that happens when “an organization doesn’t get any lucky breaks.”

Here’s the essence of his theory: 

[P]rogress happens in bursts, when everything goes right and a new invention or discovery happens. For example, we might need six things to go right for a successful discovery. 

Imagine six 50/50 events all turning out in your favor, around the same time. As an analogy, that’s like flipping a coin six times and landing on heads with every single toss. The chance of getting six straight heads is only 1.56% — that’s very unlikely (hence the low probability that amazing things happen in life).

But as low as those odds are, they are not THAT low — 1.5% is rare but not unheard of. 

That coin-flip example is based on events with 50/50 odds. Suppose an organization takes certain actions that reduce the chances of those good things happening from 50% down to 30%. In that case, the odds that all six things happen in a row drops down to 0.07%. For most organizations, a 0.07% chance of something occurring might as well be zero. 

Lui notes that you can’t see upside decay with any of your normal feedback mechanisms. The fact that nothing amazingly good has happened doesn’t tell you much about the odds of it happening in the future — 1.56% is still pretty unlikely. As he writes, “the absence of rare positive events is unexceptional.” No one is surprised that amazingly good things don’t regularly happen to them. 

So how can you know if you have (or haven’t) taken the right steps to set yourself up for amazingly good things?

Lui argues that upside decay is preceded by a “lack of virtue.” In other words, a lack of virtue is a leading indicator that upside decay will follow. He reasons that a lack of virtue is not just “a feel-good luxury” — it has a significant impact on “weak ties.” (The concept of weak tie relationships — casual contacts that people only see occasionally — was popularized in a 1973 paper by sociologist Mark Granovetter). 

A lot has been written about the value of weak ties in contexts like job searches. You are far more likely to find your next job through your acquaintances than your close friends. Why? Because acquaintances are, by definition, less connected to your social and professional networks. As a result, they are more likely to be aware of job opportunities that you would not have otherwise known about. 

Before an acquaintance can be helpful, two things need to happen: 

  1. You need to make contact with the acquaintance 

  2. The acquaintance needs to want to help

Close friends will usually help you, regardless of your moral character; they have decided you are worth being friends with “as you are,” rather than as you might be. But acquaintances possess far more flexibility on their degree of helpfulness. And this is where the power of “virtue” comes into play. Here’s what Lui wrote about this idea:

Strong ties are conspicuous. Weak ties are inconspicuous but numerous, and help in unexpected ways. When weak ties are activated, they can be more helpful in aggregate than strong ties.

But weak ties will not help an unvirtuous organization! Weak tie assistance is voluntary and altruistic. This means that they only help those they think are virtuous.

Without weak ties, organizations resort to strong ties and hard assets. This leads them to adopt a mercantilist approach. Their zero-sum mindset alienates others and makes them even less virtuous, because their positive-sum actions are now viewed suspiciously by others. Left with no choice but to double down on their zero-sum approach, they’ll antagonize all their weak ties and enter upside decay.

This also explains why their good luck disappears but they don’t suffer much additional bad luck. Weak ties mostly aren’t motivated enough to hinder an unvirtuous organization, but they’ll gladly refuse to help.


Organizational Virtue 

Organizations can be virtuous. Most founders and employees WANT to work for a virtuous organization. And virtuous companies can enjoy a significant competitive advantage — improved recruitment efforts. (I’ve articulated this concept of “marketing to employees” in a number of previous Marketing BS essays, like herehere, and here.)

But Lui isn’t writing about virtuous organizations in the context of specific corporate strategies. He’s arguing that virtuous organizations are more likely to build weak ties — which increases the odds that unlikely positive events will happen in the future. 

For companies, the decision to act virtuously is often balanced against profits. In the best-case scenario, being virtuous aligns with profits. Problems emerge when there is a trade-off between virtuousness and clearly articulated profits. For business leaders, it can be very hard to “do the right thing” and very easy to rationalize doing a “slightly” unethical thing when under pressure to achieve certain results (e.g., “We need to increase shareholder value”).

You can find lots of examples of the tension between long-term virtuousness and short-term results. In 2019, I wrote about email optimization and the question of “how many emails per week is TOO many?” Research on the subject is clear: increasing the frequency of emails will usually drive more sales. 

Within many companies, there are discussions between people who want to send MORE emails (to boost sales) and people who want to LIMIT the number of messages (because it’s “the right thing to do”). But in these debates, the hard data always favors the side of more, more, more. It’s no wonder that organizations, over time, tend to drift towards short-term optimization at the expense of virtuous behavior.

Here’s a recent example: Apple originally charged a commission in the App Store to “cover their costs.” Today, the App Store generates more than $70 billion in revenue — which is almost all pure margin. When COVID hit, Apple agreed to reduce their charges for smaller developers. But rather than charge zero commission for the first $1 million in revenue, they instead reduced fees for apps that generate less than $1 million in revenue — and ONLY IF those developers apply for the discount. Apple’s structure not only forced companies through some extra hoops, but it also cornered some companies into an awkward decision. If you have an app that brings in $950,000 per year, then you need to be careful about generating an extra $50,000, because it would cost you far more than $50,000 in additional Apple fees. In this case, Apple chose the route of maximizing short-term revenue, even if it was not the “right” thing to do. (In contrast, Google automatically opted developers into a new fee structure that discounted the commission — for everyone — on the first $1 million.)

The tendency to “optimize for the obvious in the short term” doesn’t just exist at the organizational level. Individual people can fall into the same trap. There are many books (like David D'Alessandro’s Career Warfare) that describe tactics for using office politics to claim the corner office. The techniques outlined in these books are often effective. You CAN play office politics to get ahead. However, this approach is (definitely) NOT virtuous. 

Playing the office politics game can help in the short term and increase your chance of promotion, but it hurts your “weak ties” — thereby decreasing the probability of an “unlikely positive event” happening in your future. 


Personal Perspective

In my career, the most important “highest-value, lowest-probability” event happened when I left Expedia to join A Place for Mom. That decision eventually boosted me to a CMO role and also introduced me to a prestigious private equity firm that I still work with today (and was responsible for a significant portion of my net worth). 

After reading Lui’s essay, I thought about this chapter of my life — specifically, all of the things that had to occur for the “unlikely event” to take place.

  • In business school, I became close friends with a guy who went on to lead the acquisition of A Place for Mom. Every time he came to Seattle, we met up for dinner and drinks — based solely on our friendship (not business). 

  • When he told me his PE firm was looking for a new CEO to run APfM, I immediately thought of my current manager at Expedia. I loved my manager — he recruited me to the travel company and supported me there. I held a senior enough role that if he left, I would find myself without a “sponsor,” and my career at Expedia would likely go sidewise. From a personal career perspective, the “right” thing to do was clear: I should not tell my Expedia manager anything about the opportunity. But instead of “optimizing” my career, I did what I thought was best for him — I introduced him to the private equity firm.

  • My manager took the CEO job. He also recruited a handful of Expedia people to join him at APfM, but he had to be careful. Taking too many people would probably result in a difficult conversation with Expedia’s CEO (and lawyers). Plus, poaching talent would hurt his own “weak ties.” He had a very strong team at Expedia — many of whom would have followed him to APfM — but he chose me. I do not think I was selected because I was the best or the smartest. At that point at Expedia, most of the VPs were — at least to some extent — playing politics within the organization. When a CEO is building a new team, the last thing they want is politics. My “lack of politic playing” was a disadvantage at Expedia, but it was highly valued for recruitment into the new team.

  • At APfM, we needed a CMO. Rather than fight for that role myself, I went out and recruited someone more experienced. I did the “right” thing for the company (and the person I recruited was fantastic; he got us into television — something we never could have done without him). When the CMO left 18 months later, I DID express interest in the role. The fact that I did not fight for the position earlier showed that I cared more about the business than my own career. In doing so, I earned the social capital to be believed when I said, “I truly think that me filling the CMO role would be the best thing for the business at this time.” I was given the head of marketing position — even though the owners were uncertain about putting someone without previous CMO experience in the spot. 

Most of the things I’ve done in my career focused on “doing the right thing” instead of trying to optimize for my own personal benefit. This behavior probably impacted my career trajectory along the way (case in point: I launched my career at P&G, and I was the LAST person in my hiring cohort to be promoted). Still, doing the right thing increased the chances that I would experience an “unlikely positive career event.” And that unlikely event was worth FAR more than a half-dozen promotions at earlier points in my career.


Invest for the Long Term 

Suppose you agree with Lui’s argument that some actions can destroy value in the short term but then provide bigger benefits in the long term. If you want to try and implement this type of strategy, there’s one formidable obstacle: it takes a long time to receive any feedback on the long-term-value-creating choices. In many cases — especially when the impact is slow enough — it’s impossible to accurately link cause and effect. You will find that it’s far easier to optimize last-touch paid search clicks than figuring out if your television brand campaign will decrease price sensitivity over the next four years.

Even when you CAN point to examples, it can be very difficult to quantify them. I saw this tweet the other day (from Mercury CEO Immad Akhund):

When COVID pushed the world to remote meetings, Zoom took off with greater velocity than Google Meet or Microsoft Teams. In response, a product manager for Google Meet worked with Google’s calendar team to automatically add a Google Meet link to every calendar invitation. 

I’m positive that the company’s tests showed a dramatic increase in Google Meet usage. BUT — the new feature was also confusing for many users. When people created an invitation and added a Zoom link, the calendar entry showed TWO links: the Zoom one that was manually inputted by the sender and a Google Meet one that was automatically generated. How on earth would people know which link to use?

During the past year, many of us experienced the result of this mess: two people try to connect for a meeting, but one person joins the Zoom room while the other person opens Google Meet. As a result of this annoying feature, many people have trained themselves to ignore the Google Meet links entirely — even when the meeting is INTENDED for Google Meet. This is a classic case of chasing short-term market share without considering the long-term damage being inflicted upon the brand. 

Here is one more perspective on virtue: The Guardian ran a May 23 story that asked, “why are we so uncharitable to those that do good deeds?” (The author expanded on the question in this Twitter thread.) 

Basically, people CAN improve their reputation by doing good deeds. However, this reputational boost only occurs if other people believe the good deed was motivated by selfless reasons. If, on the other hand, people suspect the good deed was performed for self-interest (or self-aggrandizement), then they might react with apathy or even hostility — just look at what happens whenever Mark Zuckerberg makes a donation. 

To reduce the risk of public blowback, some strategic “good deed actors” will downplay their actions. The most common approach is the use of “anonymous donations” — which should result in no one receiving any credit. If the anonymous donation is somehow revealed, then the benefactor receives the largest possible amount of credit. And it’s easy to arrange the “accidental” release of the information to a selected number of people the benefactor wants to impress or influence. (Of course, this cat-and-mouse game never ends; if people think the information was leaked “by accident but actually on purpose,” then the entire scheme backfires). 

Here’s the takeaway point: we value people who do good things without any desire for personal enrichment. Quite often, that admiration is NOT helpful in the short term, but it increases the probability of future unlikely positive events. And those “long tail events” have a disproportionate influence on the success of individuals, organizations, and countries.

Consider this point when you want to rationalize your organization’s commitment to “doing the right thing” — even when the data doesn’t back you up. 

Keep it simple,

Edward

Edward Nevraumont is a Senior Advisor with Warburg Pincus. The former CMO of General Assembly and A Place for Mom, Edward previously worked at Expedia and McKinsey & Company. For more information, including details about his latest book, check out Marketing BS.