In this chapter
01Those who saw ahead of everyone else
Every time you open a program, tap an icon or ask a digital assistant something, you are using ideas someone had before they were obvious. Many of those ideas came from women. Not as a curiosity to remember on March 8: as foundations, on which everything else was built.
I start here because the question I am asked most often, as a woman running a digital business, is always the same: “so how did you end up making software?”. The honest answer is that software has always been made by people with a vision and the method to make it real. And the history of the digital world is full of women who had both.
1843: a machine that didn’t have to stop at numbers
Ada Lovelace translates an article by the Italian engineer Luigi Menabrea on Charles Babbage’s Analytical Engine, and adds notes of her own, longer than the article. In those notes is the first public description of an ordered sequence of steps for a machine to solve a problem: that is why she is often called “the first programmer”. But the most surprising point is another one. Lovelace writes that the machine might act on things other than numbers, and even compose music. It was 1843: she was describing the very idea of the computer as a machine for any kind of symbol.
1952: getting machines to speak in human words
In 1952 Grace Hopper completes the A-0, considered the first compiler: a program that translates instructions written in symbolic form into machine language. From that work came FLOW-MATIC, the language with English words that paved the way for COBOL. The underlying idea was revolutionary and very simple: it shouldn’t be the person who adapts to the machine, but the machine that understands the person.
The Apollo years: software that couldn’t fail
Margaret Hamilton led the software engineering division of the MIT laboratory that developed the onboard software for the Apollo missions. In 2016 she received the Presidential Medal of Freedom, the highest civilian honor in the United States. Of her work, MIT recalls the ideas that became the foundations of reliable software: giving priority to important tasks, anticipating errors, leaving the decision to the human being at critical moments.
None of these women had an easy road. They had one thing in common: they looked at the problem from the side of whoever would use the result. The reader of a machine’s output, the programmer writing code, the astronaut in flight, the person switching on a computer for the first time. It is the same outlook that makes a digital business work today.
02Design you understand without instructions
In 1983 Susan Kare joins Apple’s Macintosh team. She comes from sculpture, not computing, and she has a task nobody had ever tackled in that way: drawing the images for a screen made of dots. The Smithsonian tells the story: a colleague suggests she use the smallest graph paper she can find. She marks out a square of 32 by 32 squares and colors them one by one. Her mother had taught her counted-thread embroidery: the same principle, the same grid.
From those squares came the smiling Mac that greets you at startup, the trash can, the floppy disk for saving, the sheet with a folded corner, the error bomb. And Chicago, the Mac’s first system font, even though she had never designed a typeface before. In 2019 Cooper Hewitt, the Smithsonian’s design museum, gave her its lifetime achievement award.
The brand as a system, not a drawing
Paula Scher has been a partner at the Pentagram studio in New York since 1991. She designed the identity of the Public Theater, which mixes popular and highbrow typography, and those of brands such as Citibank and Tiffany. In 2006 she was the first woman to receive the Type Directors Club medal. Her work shows something every business should know: a brand is not a logo, it is a system of typefaces, colors and rules that is recognizable everywhere, on a poster as on a phone screen.
03Algorithms: those who look at them closely make them better
An algorithm that learns from data is only as good as the data it learns from. This sentence, almost taken for granted today, became taken for granted partly thanks to the work of some researchers who had the courage to measure instead of trusting.
Giving machines something to see
Fei-Fei Li, a professor at Stanford, created ImageNet and the challenge of the same name: a large collection of labeled images and a competition to measure how well programs recognize them. It is one of the works that launched the era of artificial intelligence we see today. She co-founded the Stanford institute for human-centered artificial intelligence, and co-founded AI4ALL, an organization for AI education open to everyone.
Measure before you trust
In 2018 Joy Buolamwini and Timnit Gebru publish a study, “Gender Shades”, that puts three commercial facial recognition systems to the test. They use a set of faces balanced by gender and skin tone. The result: for lighter-skinned men the maximum error was 0.8%; for darker-skinned women it reached 34.7%. The same program, presented as accurate, made very different mistakes depending on who was in front of it.
Buolamwini then founded the Algorithmic Justice League, an organization that works to make the effects of artificial intelligence known to everyone and to prevent its harms. The lesson for business owners is practical: an average figure is not enough. Before handing a decision over to a program, you look at how it behaves with everyone who will use it.
Quiz
Who did what
Choose one answer for each question: the explanation appears right after.
Who wrote, in 1843, that a machine might also work on things other than numbers, even on music?
Ada Lovelace, in the notes to her translation of Menabrea’s article on Babbage’s Analytical Engine.
What does a compiler, like Grace Hopper’s A-0, do?
The compiler translates what the person writes into instructions the machine executes. The A-0 dates from 1952.
On what grid did Susan Kare draw the first Macintosh icons?
A square of 32 by 32 squares on graph paper, colored one by one like the dots on the screen.
What did the 2018 “Gender Shades” study show?
A maximum error of 0.8% for lighter-skinned men, up to 34.7% for darker-skinned women.
04Today’s numbers, and the direction they are heading
Stories matter, but people who run businesses also want numbers. Here they are, taken from official sources and with no convenient rounding.
Europe: fewer than one in five, but growing
According to Eurostat, in 2025 10.45 million people worked as ICT specialists in the European Union, 5% of the employed. Women were 19.5%, three percentage points more than in 2015. The European Commission sets this figure next to another: women are 51% of the European population, and among graduates in science and technology subjects one in three is a woman. The room to grow is huge, and every year the Union tracks progress with a dedicated index, the Women in digital index.
Romania: the country where women carry the most weight
The figure closest to home for me, working between Italy and Romania: in 2025 Romania is the EU country with the highest share of women among ICT specialists, 27.8%. Latvia and Bulgaria follow. It is a figure little known in Italy, and it reflects a solid tradition of technical education for women.
Italy: over one million three hundred thousand women-led businesses
At the end of 2023, Unioncamere (the Italian union of chambers of commerce) counts 1,325,270 women-led businesses in Italy, 22.2% of all businesses. Of these, 27,298 work in information and communication services, the digital sector, and they grew in 2023. And there is a forward-looking sign: among women-led businesses, 10.6% are run by women under 35, against 7.9% of other businesses.
The numbers don’t say that digital is a world for the few. They say it is a world where there is still plenty of room.
05The strength it takes: vision, method, a critical eye
I have told the stories of famous pioneers, but the strength I am talking about doesn’t belong in a museum. It is the strength I see every day in the women entrepreneurs who work with me: the owner of a restaurant who wants to update her menu from her phone, the professional who decides to have her own website instead of relying only on social media, the bride preparing the invitation for her guests in two languages. It can be broken down into three abilities, and all three can be learned.
1
Vision
Seeing the result before it exists and being able to describe it: who will use it, what problem it solves, how you will know it works. It is what Lovelace did with a machine nobody had built yet.
2
Method
Turning the vision into verifiable steps. Hamilton anticipated errors before they happened; Kare worked inside a grid. Clear rules don’t take away creativity: they make it reliable.
3
A critical eye
Not trusting a number just because a computer wrote it. Buolamwini measured instead of believing. In business it means always asking: on what data, tested with whom?
4
Ownership
Keeping the tools of your work in your own hands: the domain, the website, your customers’ data. A vision that lives on someone else’s platform can be switched off by others.
In everyday digital life
Translated into a business’s daily practice, these abilities become concrete choices:
- A website of your own, with your own domain, instead of a shop window on social media only: I talk about it in the chapter on website or social media.
- Getting found by people searching nearby, with a well-kept business profile: that is the subject of local SEO.
- Speaking your customers’ language, even when there are two: the chapter on the website in two languages explains how.
- Using artificial intelligence with judgment, knowing where it helps and where you need a person who knows the trade: I talk about it in the chapter on professionals, AI and page builders.
Words of the chapter
The pioneers’ vocabulary
Tap a card to flip it.
Test yourself
Your business in the digital world: where do you stand?
06Zenith, as a digital entrepreneur
I write this chapter in the first person because it concerns me. Zenith Develop is my business: I design and write software, websites and apps, between Venice and Romania. I don’t tell the pioneers’ stories to place myself beside them, but because their ideas are the tools I use every day.
True or false?
Five things people say
Decide whether each statement is true or false.
The first published algorithm designed for a machine bears a woman’s signature.
Ada Lovelace’s 1843 notes contain the first public description of a sequence of steps for the Analytical Engine.
In Europe the share of women among ICT specialists has been stuck for ten years.
It rose by three points from 2015 to 2025, to 19.5%, according to Eurostat.
Among EU countries, Romania has the highest share of women among ICT specialists.
27.8% in 2025, the highest value in the Union, ahead of Latvia and Bulgaria.
An algorithm that is accurate on average is accurate with everyone.
The Gender Shades study showed maximum errors from 0.8% to 34.7% depending on the group of people.
To design famous icons you necessarily need a computer science degree.
Susan Kare came from sculpture and had learned the grid from counted-thread embroidery.
Frequently asked questions
How many women work in digital in Europe?
According to Eurostat, in 2025 women were 19.5% of ICT specialists in the European Union, three points more than in 2015. The highest share is Romania’s, at 27.8%.
How many women-led businesses are there in Italy?
Unioncamere counted 1,325,270 at the end of 2023, 22.2% of all Italian businesses. Of these, 27,298 work in information and communication services.
Who is considered the first programmer in history?
Ada Lovelace, who in 1843, in her notes to an article on Babbage’s Analytical Engine, published the first public description of a sequence of steps for a machine to solve a problem.
Who designed the icons of the first Macintosh?
Susan Kare, who joined Apple in 1983. She drew the smiling Mac, the trash can, the floppy disk and other icons on a grid of 32 by 32 squares, as well as the Chicago typeface.
What is algorithmic bias?
It is a systematic distortion: the algorithm works well with some groups of people and worse with others, often because the data it learned from were unbalanced. The 2018 Gender Shades study measured an example of it in facial recognition.
Where do I start to take my business digital?
With a clear description of your customers and with tools that remain yours: a domain registered to you, a website of your own, customer data in a place you control. Then you add one step at a time.
Sources
- Eurostat, “Number of ICT specialists in the EU continues to grow” (May 2026): 10.45 million ICT specialists in 2025, share of women 19.5%, +3 points since 2015, Romania at 27.8%.
- European Commission, “Women in digital”: women 51% of the population, one in three STEM graduates is a woman, the Women in digital index.
- Unioncamere, press release on women-led businesses 2023 (March 2024, in Italian): 1,325,270 businesses, 22.2% of the total, 27,298 in information and communication services, 10.6% run by under-35s.
- Computer History Museum, “Ada Lovelace”: the translation of Menabrea’s article, the 1843 notes, the machine that acts on things other than numbers.
- Centre for Computing History, “Grace Hopper completes the A-0 Compiler” and Computer History Museum, interview with Grace Hopper: the 1952 A-0, FLOW-MATIC and COBOL.
- MIT News, Margaret Hamilton and the Presidential Medal of Freedom (November 2016).
- Smithsonian Magazine, how Susan Kare designed the icons for the first Macintosh and Computer History Museum, profile of Susan Kare.
- Pentagram, “Paula Scher”: partner since 1991, Public Theater, Citibank, Tiffany, Type Directors Club medal 2006.
- Stanford University, profile of Fei-Fei Li: ImageNet, Stanford HAI, AI4ALL.
- Buolamwini and Gebru, “Gender Shades”, Proceedings of Machine Learning Research, 2018: errors up to 34.7% against 0.8%; Algorithmic Justice League.
Chapter updated October 2026. All numbers come from the sources above; where the source gives a year, the year is shown next to the figure.