Loss Aversion Isn't What You Think

Loss aversion is a widely known psychological concept.

The mainstream understanding? People tend to avoid situations that could result in a loss. “No sh*t”… I hear you thinking.

A slightly more nuanced understanding?

In a choice-making situation with risk, people bias towards avoiding a loss rather than making a gain of equal value.

If I were to pitch a game of flip a coin whereby heads loses $50, tails wins $50, most would decline.

A significant reason for that, according to legend-status psychologists Daniel Kahneman and Amos Tversky, is people feel losses more intensely than equivalent gains.

This biases them to avoid situations of loss even if the mathematical or rational risk is equal — or favourable.

That’s right, people are loss averse in situations where the perceived objective gain is more than the loss, even though the likelihood of both happening is equal. For example, if I changed the coin toss to heads loses $50, tails wins $60, most would still decline.

This theory was put to the test in experimental studies. From the results, Kahneman and Tversky suggested the emotional intensity experienced from loss could be 2X more powerful.

Here’s a classy visualisation:

In this example, the intensity of feeling sad is 2X that of feeling happy.

But, beware: this “2X” figure is flaunted around as if it’s some-kind of universal constant. Which, it is not. Better to think of it as an approximated value for the specific conditions of their study.

The circumstances massively impact this multiple, particularly the individual presented with the choice. Everyone has their own risk-reward comfort zone, which itself is context-dependent.

The main takeaway here is not the 2X figure, but the premise that losses are felt more intensely than gains — whether it’s 0.5X, 2X, or 5X.

Gains and losses do not have to be financial, either. It’s anything of perceived value.

For example, if an employee champions buying an expensive SaaS product and it doesn’t deliver against expectations, that person could feel like they are losing credibility with their colleagues (“Who bought this heap of crap?”). Therefore, they could be loss averse against this perceived outcome, and will opt out of buying the SaaS product in the first place.

The amount of the equivalent value of the loss or gain is also a key variable. The more it is, up to a certain point, the more intensely emotion is felt. Feeling bad escalates disproportionately too feeling good.

This is intuitive from our own lived experience. Losing $50 sucks compared to losing $5. Losing $1,000 is a mind stab compared to $50. Winning equivalent amounts doesn’t provide an equal amount of happiness value, in the inverse.

Main takeaway from this?

People bias towards buying psychological value not rational value.

Said differently, your proposition may “make sense” expressed as a mathematical or functional value that takes into consideration the risk of buying the product and the likely utility value it delivers. The facts. However, this is not how people evaluate propositions.

They are motivated by psychological value, taking into consideration how bad something will feel if it doesn’t work out versus how good it will feel if it does.

Much of the evaluation that determines this happens at an unconscious level.

Commercial application

There’s numerous ways loss aversion can be utilised to influence choice-making to meet a desired business outcome.

For example, framing a situation to place focus on loss triggers a strong emotional response.

Old school parenting tactics are fabulous at this… “brush your teeth or they’ll fall out”.

The reverse — “brush your teeth to keep them in” — focusses on the gain and is much less motivating, yet they are the same outcome.

Toothpaste brands have been using the same trick for decades.

It’s not just teeth, of course.

This concept is easily transferred to a commercial context of practically any nature, whether it’s a $1 consumer packaged good or a $10,000,000 enterprise product.

Just think of the old saying “Nobody got fired for buying IBM”. Most B2B buyers (e.g. non-early adopters) are acutely loss averse.

Their professional reputation, which means their livelihoods and ability to provide for themselves and their families, is on the line. The psychological value of buying IBM, the safe option, trumps that of lesser-known direct competitors.

Prospects across all B2B categories implicitly perceive this. The phrase really means: “No one gets fired for buying the category leader” — whether that’s Salesforce, Oracle, or Microsoft.

All these behemoths have to do, to activate loss aversion, is to flex their category leadership. A huge chunk of their market cap is attributable to this.

This is a key reason why it’s imperative for startups to find early adopters, i.e. prospects that are mobilised by the psychological value of:

1. The loss of inaction / continuing with legacy solutions

2. The gain from adopting a new solution

Let’s explore that second one.

If you focus on promoting the positive outcome of the proposition — rather than the negative — the perceived gain needs to be some order of a multitude higher than the perceived loss in order to be compelling.

The perceived loss, in this scenario, is what the prospect feels they are risking and can therefore be lost as a result of buying into the proposition.

It doesn’t have to be plain old money (the cash investment). It can also be things like credibility, opportunity cost, and notions of self-image.

Let’s revisit my coin toss. To make it appealing, I would need to keep heads loses at $50 and increase tails wins. If I kept increasing the winning outcome by $5 increments, eventually I would hit a number where the pleasure from winning would outweigh the displeasure from losing the $50.

Maybe at $50 heads and $100 tails, whoever I’m pitching would be game for it. For this thought experiment, let’s say winning $100 feels 1.1X that of losing $50.

At that point their perceived happiness from the winning outcome would exceed the perceived intensity of sadness from losing. Now, it’s a deal worth making. The incentive to avert loss has been neutralised.

Here’s what that would look like in the mind:

I could make this more appealing by increasing the won amount even further: $150, $250, $500, etc.

This would increase the conversion rate (folks agreeing to play the game) but come at the cost of margin (I wold lose more money).

Somewhere along the scale there’s an optimal ratio where I convert the highest amount of people and lose the least amount of money. The intersection of the two.

However, applying this logic at a cohort-level would be a ‘dumb’ approach. Since each person has their own risk aversion profile, I would be offering way too much of a reward for those that are more risk tolerant and not providing enough for this that are less. The optimum approach would be to find the minimum dollar amount to induce a sale for each prospect.

Winding back a little, this strategy of increasing the perceived psychological value of happiness (if the proposition delivers) relative to perceived sadness (if it doesn’t) is used widely commercially.

For example, with free trials and money-back guarantees. These pricing mechanisms neutralise loss aversion by reducing the perceived psychological strength of a negative outcome to less than that of a positive. A company could also achieve the same outcome by doing the opposite; improving the benefits but keep pricing the same. Or, some mix of the two.

Perceived probability

Another major variable at play is perceived probability. The coin toss scenario I gave has a mathematical 50/50 outcome, which most people perceive to be the case (unless they think its rigged). Perception matches reality.

However, in practically every value proposition scenario, this is not the situation. Whether a prospect is buying a Big Mac or Hubspot, rarely will they consider that a positive or negative outcome of that purchase is 50/50. For example, maybe they perceive the Big Mac has a 95% positive outcome success rate (way to go Speedee Service System) and Hubspot is 45% (because “no one f*ckin updates it properly”).

With a higher perceived positive outcome rate, more weight is given to the positive. Result? More psychological value. The Big Mac gets bought. Nom nom.

For example, imagine I offered you a coin toss with a rigged coin at $100 win or lose. Tails has a 95% chance of coming in and you are free to pick whichever side you like — heads or tails. Walking away from this proposition may feel like you are losing $100. The perceived likelihood of winning $100 feels so tangible that a crisp Benjamin is practically weighing down your pocket.

Increasing the perceived probability of success, like this, is a critical lever to overcoming loss aversion. If it ‘s high enough and the positive outcome is desirable enough, it can create its own sense of loss aversion that mobilises prospects to buy — “I can’t lose this opportunity”.

Here’s a few principles to remember:

• Framing effect — People’s decisions can be influenced by how a situation or choice is framed (e.g. loss vs. gain). Equivalent loss is more powerful than gain.

• Variability — People aren’t all the same. They have different risk-reward profiles, which affects their decision-making when facing potential loss.

• Probability misperception — People tend to overestimate the likelihood of low-probability events and underestimate high-probability ones.

• Risk aversion — In scenarios involving potential gains, folks tend to be risk-averse. They’d rather take a 100% chance of getting $90 than a 90% chance of getting $100.

• Risk seeking — In scenarios involving potential losses, folks tend to become risk-seeking. They’d rather take a 50% chance of losing $100 than a 100% chance of losing $50.

But, here’s the thing, we are missing a key element of loss aversion theory.

Reference-dependence

The meaning I describe above is what the mainstream understanding of loss aversion is. However, there’s an additional dimension to loss aversion that often goes overlooked and is therefore not incorporated into business use cases effectively.

That is? “reference dependence” — as Kahneman and Tversky call it.

To utilise loss aversion bias optimally to achieve business goals, context matters.

The anticipated pain or pleasure of a loss or gain is dependent on a person's point of reference. It’s relative.

People aren’t just averse to loss in an abstract sense — they are averse to loss relative to where they believe they should be.

This notion of “should be” is a powerful mental anchor that people refer to when evaluating a choice. It’s not so much about what someone has, but what they think they should have.

In a very simple hypothetical sense, if someone’s networth is $10,000 this could be their point of reference when evaluating a choice to participate in the $50 coin toss. The emotional intensity from losing the $50 would be more pronounced than if their networth was $1m.

How so? In either situation the person feels they should have $10,000 or $1m, but the relative difference between the two amplifies the perception of the loss.

Even though $50 is of equal value in both scenarios, the implicit perception $50 represents a material portion (0.5%) of their networth for $10,000 versus an immaterial portion (0.005%) for $1m amplifies the pain.

At $10,000 they’re like “Oh sh*t their goes my Doordash budget this month.”

At $1m, they’re like “meh”.

However, this is an oversimplification. Loss aversion is more meaningful when we consider that people's reference points aren’t static. They are situation specific and a lot less objective than we’d to think.

There’s plenty of people that feel rich at a $10,000 networth and poor at $1m. In this case, the $10,000 person would feel relatively better than the $1m person after losing $50 in a coin toss.

Why? Their reference point is not their objective networth (as an accountant would determine), but what they feel it should be.

For example, the $10,000 person may have just won $9,000 before the coin toss. Their networth point-of-reference could be $1,000. So, through their eyes, they’re “still up $9,950” (895%) after just losing $50.

Meanwhile, the $1m person may have just been through a divorce in which they lost $1.5m to their ex-spouse. Their reference point is $2.5m so they feel pretty bummed out. The notion of losing a further $50 would antagonise this feeling of loss further.

Brewster’s Millions

There’s a classic movie called Brewster’s Millions in which the main character, Montgomery Brewster, is faced with a wild choice.

At the start Montgomery is a broke minor league baseball player with seemingly no prosperous future. Suddenly, a distant (and mega wealthy) relative dies and leaves him an inheritance upon which he is forced to pick one of two options:

• Option 1: The chance to inherit $300m. To receive this, he must first spend $30m in 30 days under a strict set of conditions (much easier than it sounds). If he doesn’t pull it off, he walks away with zero — “just the shirt on his back”

• Option 2: Walkaway with $1m straight away, no strings attached

He picks Option 1.

Despite the opportunity to go from broke to an instant millionaire, he opts for risking it. Why?

Under loss aversion theory, when presented with this choice, Montgomery’s reference point shifted from the $0 he had in his bank account as a broke baseball player to the $300m he ‘should have’ for his inheritance. By accepting the $1m through Option 2, he would have felt he’d lost $299m rather than gained $1m.

This example is still an oversimplification of how loss aversion works in practice. It implies more rationality and rigidity than is often the case. The mind is a dynamic self-rationalising machine that’s trying to balance competing priorities.

Identifying reference-dependence

As Brewster’s Millions alluded too, a reference is often a moving target.

Let’s take slot games as an example. For a period of time I ran a startup that operated in this space.

What I discovered? Players have multiple reference points that are tiered and dynamic, referencing them in different ratios in different moments to inform choices regarding the value proposition placed before them (the slot game).

What the hell do I mean by that? I’ll address tiers first.

Each tier is basically a degree of zoomed in (or out) from the specific context of the situation. There’s 1X Zoom, 5X Zoom, and 10X Zoom.

What that maps too in the context of slots:

1X Zoom. This is the most zoomed out. Where referencing networth (or a bank balance) may come into play.

• ➡️ Example: someone with a higher networth maybe more likely to choose and play a slot with $100 spins than $1 spins, because there’s little psychological pain (or gain) from the outcome of $1 spins. Another example: the reference point could be the budget for their trip — “I came to Vegas with $100,000”

5X Zoom. With this level of zoomed in, the reference point becomes the slot play session itself.

• ➡️ Example: someone’s index finger-deep in a Buffalo slots session and they’re $200 down. They’ve put $500 into the machine so far, so their reference point to where they think they should be is $500. This can be dynamic, so if the player has a winning streak and increases their credit balance to $1,000, this could become the new reference point.

10X Zoom. With this level of zoomed in, the reference point becomes the moment of play. What is happening transiently, right in front of them.

• ➡️ Example: someone triggered the bonus game where they were eligible to win $1,000. They didn’t win but feel like they just missed it — that they should’ve had it because they got 7/8 symbols right. Their reference point increases by $1,000 for a transient period of time (maybe seconds, minutes, or longer).

In any given moment, players will reference these zoomed in tiers to different degrees of significance and intensity.

For example:

• 1X Zoom. When deciding what slot game to play in the first place, how much budget they want to allocate to it and how much they want to bet per spin

• 5X Zoom. When evaluating how much longer they will play a slot game, how many more spins and value of bets they want to make — “Let’s put another $50 in so I can win my way back to my starting $500 balance”.

• 10X Zoom. When presented with stimulating information, adjusting their evaluation of the situation — “I should’ve won $1,000 in the bonus, let’s keep playing until this game pays it out.”

At any given moment these three tiers are having an influence on reference-dependence based decision making. Depending on the context, some more than others.

Slots may seem a far stretch from most other value propositions. But, the same zoomed-in tiers of reference-dependence apply there too.

The key is to understand what this looks like in the mind of your prospects. What are they anchoring against? Uncover the nuances.

I’ll share a personal example.

B2B — Yavli

I previously co-founded a B2B adtech company — Yavli — that generated revenue for digital publishers from ad blocking visitors.

There were two primary ways we utilised loss aversion to mobilise publishers to adopt our product:

1. Create a new reference point. “You’re currently making $100k per month in ad revenue, but you should be making $130k”

2. Frame loss against an existing reference point. “Ad blocking is growing 50% year on year, over the next year you’ll lose 10-15% of existing revenue”

However, this was merely the 1X Zoom reference point. This got us in the door and fueled meaningful conversations — up to the point of being in a deal against other directly competing vendors — but it’s not enough on its own.

After that, the sales cycle went deeper and it exposed more reference points in the minds of publisher executives (the stakeholders involved in making the purchasing evaluation).

At 5X Zoom, publisher execs were evaluating the choice to use us versus broader competing priorities. This plays out in many ways. Including, an assessment of “What if this goes wrong?” compared to a positive outcome.

Reference-dependence at this level often becomes personal-level, not just business. For example, the publisher exec maybe thinking: “I’m in the running for SVP of Revenue. Is my promotion at risk if I do this?”

To tackle this we focussed on framing the gain from a position of loss: “You’ll regain $7,500-10,000 per month of lost revenue” — while emphasising the high likelihood of success by anchoring against proven solutions that publishers were already familiar with in a different setting.

Later, we focussed more on mitigating the perceived potential loss: “You can now integrate us via your CDN, dramatically reducing the integration costs and the cybersecurity risk compared to before”.

At 10X Zoom, the reference point became terms and contract negotiation. For example, publishers would sometimes say “I get a 70% revenue share from my other vendors, so I want 70% with you”.

When they said this they were focussing on the 30% loss and not the 70% gain. To counter this, we reframed it by introducing a new reference point: “without us you are losing 100% per month”.

That’s it for today. I’ll be back in your inbox soon. 🤘

Martin

Did someone forward you this email? Happens allll the time.

Subscribe to receive more content like this, weekly(ish). 👇