May Monthly Business Aha's

Cups, balls, and lemons

Epistemic status: field notes from one month of experience

1 June 2026
1300 words - 6 min read
Other posts

TOC

TL;DR

A heuristic for two-way decisions; Never trust a good marketer!

■ Decision-making: two cups, one ball

Two cups, one ball: you can follow the ball or the empty cup.

Being self-employed means there are always decisions to make and never enough data to make them.
A different approach is needed: one based not on the data, but on the structure of the choice itself.
As a framework lover, I built a heuristic:
Given two complementary decisions A and B, a bad reason for A is a good reason for B.
It's not really a "heuristic for deciding": it forces the decision into a specific shape.

I've been using it for a while with good results, but only now did it come to me why it works: it's math!
Formally, "complementary" means mutually exclusive and collectively exhaustive (the MECE principle).
When two events are MECE, then P(A)+P(B)=1, or put plainly: "the probability lives in A or in B".
So a bad reason for A increases the probability of B, because it decreases the probability that A is the right choice.

A recent example.
We were deciding how to pitch a consulting service to companies, with two alternatives:

  1. Do it "like everyone else": with the reason "so companies know what to expect"
  2. Do it in a new and our own way: with the reason "so we can charge more, differentiate ourselves..."

It looks like a 1:1 tie on motivations, but using the heuristic it's 0:2, because "doing it like everyone else" is not a valid motivation in this case.
To be clear: telling companies what to expect matters, and it makes sense to use language they're familiar with.
But the service itself didn't need to be the same as other consultancies: partly because "then why hire you instead of (established consultancy)?", and partly because differentiation helped us with visibility and let us charge more for a personalized, premium service.

It worked here because there were only two options, they were exclusive (you do one or the other), and we could only do one (we had already ruled out other options).
But it's not a perfect method: as soon as the events aren't exclusive, aren't exhaustive, or are more than two, it stops working mathematically.
One of the imprints math left on me is exactly this: thinking in terms of exhaustivity and exclusivity of choices, which is why I tend to use the heuristic automatically in the right contexts.
But when I noticed that not everyone thinks this way (and why should they?), I realized that the most important part of a decision is its choice architecture (in the self-optimizing, non-manipulative sense):

In other terms: building the ball and the cups is harder than winning the shell game.


■ No one cares about qualyti

Yes, the misspell is on purpose.

A mistake I am making as a self-employed person is looking for ideas by asking: "What can I do WELL?"
This forgets the most important question: "What problem am I solving, and for whom?"

The last project I worked on was in marketing, with a friend.
The idea came from noticing the abysmal quality of advertising (in Italy): shitty TV commercials made with AI, billboards that make no sense, newsletters AI-slop...
We figured we could easily do better with my technical skills (thinking about marketing automations) and my friend's storytelling expertise (after all: People buy stories).

We studied ads and commercials, worked out how to combine storytelling with AI and automations, and realized that we really could produce higher-quality content.
Amazing!

Yet: nobody cares.

We wrote to ~70 companies pitching workshops and presentations (top of funnel), with few that actually moved forward.
(Response rate was around 20-30%, almost all replies positive, but almost none of them converted... The classic "Very interesting, we'll let you know," or "Let's set up a call," and after the call "We'll get back to you about the presentation," and we never heard from them again...)
To get into the circle, we attended local events and met companies and agencies.
And here's what I figured out: companies don't pay agencies to make great ads, they pay them because they don't want to do marketing themselves.
They turn a problem of skill and time into a money problem.

"Obvious, duh."
It's not.

Companies don't care about the quality the agency can produce: a company that runs well would track marketing conversion rates, brand positioning, sales trends...
We didn't meet a single company doing any of that.
Translation: (local, Italian, and for the small sample we met) companies don't pick agencies based on the ROI their marketing delivers.

At this point our failure is obvious.
We walked up to companies as two strangers, asked for their trust, and promised higher-quality, more effective marketing.
On top of that, we were proposing the opposite of what agencies do: sitting down with the company to think through the goals of the spot/campaign/newsletter (Sell more? Position the brand a certain way? Beat the competition?) so we could sharpen the communication.
The exact opposite of what the company wants!

NOTE

This is the dynamic described in The Market for Lemons. Companies can't assess the quality of marketing, so they offer an average market price. Agencies don't compete on quality (which isn't recognized), so they compete on network size, volume, seniority... A possible solution would be performance-based marketing, which isn't common nor welcomed in Italy.

Don't believe my words

I live in Bolzano, a region packed with apples.
Here's a selected work (!!!) from a big and notorious local agency.
Now I want you to open this link, look at ALL the content, and then answer the following questions:

In my opinion this project completely fails to give me reasons to care about this apple. It's all buzzword filler ("Digital presence optimization"???) and animated movements for no reason.
At the same time, I'm sure the commissioner is thrilled with the thousand shots, the hot-air balloon, the website (super-heavy, css-loaded...) because at first glance it makes an impression... but how are the conversions doing?
If you still don't trust me: put a follow-up to this post on your calendar one month from now. Come back to this section and answer the questions WITHOUT REVISITING THE LINK. Good luck on that.

EXHIBIT A

The lesson here is: "people want problems to be solved".
If you're not solving a problem, quality counts for nothing.
Or in other words: quality != value.

And more: quality should be in how you're solving the problem, not in the byproduct.
In marketing's case, the "quality" is reliability: the company has to trust the agency.
That's it: the problem it solves is "delegate the marketing".

In marketing, the consequence is that agencies aren't incentivized to raise the quality, quite the opposite: once they've cleared the reliability bar, the only economic lever left is the volume of projects they can handle.
Quantity over quality, as quality counts for nothing.

NOTE

We ran into these dynamics in Italy (a country very different from the USA in terms of entrepreneurial culture) and in Bolzano specifically (a special province with special funding laws and a special trust for German speaking entrepreneurs). The USA-EU gap is well documented: in the USA, companies are used to taking a chance on small entrepreneurs with good ideas.

I see a connection to the startup world, where VCs say "don't make something that exists already but better, make something new and different".
From my experience this requirement makes sense because "Wanting to do X better" is a sign that (maybe) you haven't understood why X exists in its current form (even if in low quality).
If agencies survive while producing mediocre marketing, there's a reason, and challenging them on quality means fighting them on the wrong battlefield.

In other words: Don't ever take a fence down until you know the reason it was put up.