Stochastic Strata
WritingAboutGet in touch

About

Writing that tells you when it can't answershould tell you who wrote it.

One person writes everything here, and putting a real name and a checkable track record on the page is the point — you should be able to work out whether I have any business making these claims. So here is the person, and how this started.

In my own words

I studied business administration and management at the Universidad de La Rioja —Grado en Administración y Dirección de Empresas — with my focus in strategy, and I've spent the last seven years in media buying: starting junior, working up to senior, with stretches in between as a freelancer, inside tech startups, and at digital-marketing agencies. Budgets from €500 a month to €200,000.

Somewhere in those years the same thing kept bothering me. Decisions that move real budgets get made on instinct, when the data to answer them properly is sitting right there. Is this channel driving sales, or just showing up at the end? Would these conversions have happened anyway? Those questions have answers — if you are willing to apply statistical methods to the data instead of guessing. That is what pulled me toward data science.

So I started writing them down. One method at a time, in plain words, with the maths unpacked and a notebook you can run yourself — the rigour a scientist would insist on, put in front of a marketer who has budgets to move. I am curious by default, I like taking things apart to see how they work, and I would rather hand you an honest range than a confident number I can't stand behind.

— Andrei Popescu

How everything here is written

Three rules, applied to every post. They are also, more or less, why I started writing.

Every number carries its range

A figure without its uncertainty is an opinion wearing a decimal point. Nothing here is quoted as a single certain value when it is not one.

The working is shown, not asserted

The maths gets unpacked, the sources get named, and a notebook comes with each post so you can run it on your own data and disagree with me.

When the data cannot answer, say so

And say what would settle it. A stated limit is more useful than a confident number nobody can stand behind.

The facts

Written by
Andrei Popescu.
Education
Grado en Administración y Dirección de Empresas (business administration and management), Universidad de La Rioja — focus in strategy.
Experience
Seven years in media buying, junior to senior — across agencies, tech startups, and freelance. Budgets from €500 to €200,000 a month.
Based
Bratislava, Slovakia. Available for freelance work remotely across the EU.
Get in touch
andrei@stochasticstrata.com, or on LinkedIn.

Don't take my word for it

Every post comes with the notebook that produced its numbers. Run it on your own data and see whether the method holds up — that is a better test of any of this than a page about me.

Read the writing

Or email me: andrei@stochasticstrata.com