The man who made money from data (and the price we are all now paying)

...every technological advance that promised to make the task easier has been closely and seemingly inevitably accompanied by an increase in the complexity of the figures and systems involved.

Iain Banks - The Business

Herman Hollerith (February 29, 1860 - November 17, 1929) was an American inventor who developed a machine that utilised punch cards to rapidly tabulate millions of pieces of data. He was a founder of one of the companies that later merged to become The International Business Machines Corporation (IBM) and is widely recognized as the 'father of modern automated computing'.

However, the bonanza that he facilitated has only really begun to bear real fruit more recently with 5 of the world's 10 top companies all deriving their fortunes from data.

Less than 10 years ago, 5 of the same top 10 were oil companies, today only ExxonMobil still makes the list. Thus the expression that data has become 'The New Oil', but that only truly covers less than half of the reality as data, unlike oil, can been used over and over again! It truly can be the gift that just keeps giving.

The Man

Herman Hollerith was born to German immigrants in Buffalo, NY in 1860. His parents were immigrants to the United States arriving in 1848 after political disturbances in their home country.

School was not easy for him despite the fact that he was clever. It is noted at that time:

Herman is said to have been a bright and able child at school, but had an inability to learn spelling easily. His determined teacher made his life miserable to the extent that he used to avoid school whenever possible and run away when his teacher showed renewed effort to improve his spelling.

Hollerith was eventually taken out of school and tutored privately by the family's Lutheran minister.

He subsequently enrolled at the City College of New York in 1875 and then graduated from the Columbia University School of Mines in 1879. He had impressed one of his professors, W.P. Trowbridge, to such an extent that he was asked to become that teacher's assistant. It was a fortuitous appointment as Trowbridge went on to work at the US Census Bureau and Hollerith joined him there as a statistician.

In 1884, Hollerith took a position in the U.S. Patent Office in Washington, D.C. This, again, proved to be highly fortuitous as it provided him with the insights required to develop and secure patents for his own inventions. That same year, he applied for the first patent of over 30 that he would be granted by the United States along with many others from overseas.

That said, not all of his inventions were a success, during the 1880s, he had invented an electrically-actuated brake system but lost out to the Westinghouse steam-actuated brake. And it was obvious that he didn't enjoy being rebuffed. So when he tried to interest his family in a machine to process data more quickly than humans could manage and the family wouldn't help him out, he cut them off. His children grew up with no idea they even had relatives on their father's side.

The Mission

Hollerith's invention addressed a very specific issue. Every 10 years, the US government conducted a census. Governments had long needed these to know how many people it had, initially to raise taxes and to locate conscripts for the military.

Hollerith had worked briefly for the Census Office in the lead up to the 1880 count and this had convinced him that they desperately needed a better way to tabulate data than hand counting. Hollerith had invented a device that could do that and more!

But if you're going to go to all the time, effort and expense of running a census then you'll want to get the most out of the exercise. Accordingly, the US census had been expanding. In 1870, it had five different kinds of questionnaire, by 1880, it had 215. It was clear that adding up the answers would take years and that they would barely have finished in time for the next one.

Knowledge is power, as 19th Century bureaucrats understood just as well as any 21st Century platform company. Yet with the 1880 census, the bureaucrats had bitten off way more data than they could chew! And a lucrative government contract would await anyone who could resolve that.

Happily for Hollerith, he was that person and the bureaucrats were more impressed with his idea than his family had been and so they rented his machines to count the 1890 census (to which they had added even more questionnaires).

Compared with the old system, Hollerith's machines proved years quicker and millions of dollars cheaper. The counting was concluded on December 12th 1890 having taken three months to process instead of the projection of two years for hand counting. Fun fact: The total population of the United States in 1890 was found to be 62,622,250.

Businesses were quick to spot the potential of the technology and Hollerith's Tabulating Machine Company grew accordingly. It remained a market leader as punched cards gave way to magnetic storage, and tabulating machines to programmable computers in it's eventual form as IBM.

His success in 1890 led to contracts with foreign governments and his devices were used in 1891 for censuses of Canada, Norway, Austria and much later for the 1911 UK census. He provided machines for the 1900 census count in the US but used his monopoly to raise his lease pricing excessively. An act that came back to haunt him when Census Bureau employees were able to create their own, more advanced, machine (borrowing heavily from Hollerith's patents) in time to introduce them for the 1910 census.

Hollerith's company, which by then had changed its name to the Computer Tabulating Recording Company was close to ruin and in 1918, Thomas J. Watson, an accomplished salesman, joined as an executive. He revolutionized the company and transformed it back into a successful enterprise again.

Hollerith retired to his farm in Maryland and spent the rest of his days raising cows.

But if the power of data was apparent to Hollerith's customers, why did the data economy take another century to arrive?

Let's return to the oil analogy which in fairness only really talks to present day and perceived value. Yes, data can be used many times and oil only once. But like oil, data in it's raw form does not produce or provide real value...that will only occur once it is refined. Oil is refined to produce a range of by-products. Likewise, once data is refined or analysed it provides the insights that inform decisions.

And in today's era of Big Data, Big Tech, Big Government and Big Bother those decisions are now being made increasingly for us based on what we say, think, read, watch, do and buy. These business models need information and today power comes not just from data alone but from the interplay of data and algorithm.

The Machine

Hollerith had worked on the 1880 census, so he understood the issues but it was a train journey that helped him to solve the central issue.

Rail tickets were often stolen. So railway companies devised a way to link them to the person who had purchased them, the so-called 'punch photograph'. Conductors used a hole-punch on the ticket to select from a range of physical descriptors, hair or eye colour, build, ethnicity to tag the purchaser.

Having seen this system Hollerith realised that people's answers to questions could be represented as holes in cards. This was a method that had been used to control machines for some years at that time such as the Jacquard loom that wove patterned fabric based on them.

That idea had been suggested to him by Dr John Shaw Billings with whom Hollerith had come into contact whilst working for the US Census Bureau. In 1882 Hollerith joined the Massachusetts Institute of Technology where he taught mechanical engineering and he used this time to investigate the suggestion, examining the way that the Jacquard loom worked with a view to seeing if it could be used in data tabulation.

He developed the early work he had done at MIT on converting the information on punched cards into electrical impulses. These, in turn, would activate mechanical counters. Initially he used at railway ticket punch to make the holes in the cards, but whilst this proved that the system worked the limitation that the punch could only make holes near the edge of the card meant that it's full potential wasn't being realised.

Picture something that looks a bit like an upright piano but instead of keys, it has a slot for cards, about the size of a dollar bill, with holes punched in them. Facing you are 40 dials, which may or may not tick upwards after you insert each new card. In that piano-like contraption, a set of spring-loaded pins descended on the card; where they found a hole, they completed an electrical circuit, which moved the appropriate dial up by one.

Hollerith designed special punches which were manufactured for him by Pratt & Whitney, he improved the machines which read the cards and had them produced by the Western Electric Company. These allowed the electrical current to activate a mechanical counter and the amount of information which could be handled on each card increased enormously.
Speed was not the only benefit of Hollerith's system. The amount and complexity of the data collected could also be increased with questions about the number of children born in a family, the number of children still alive in a family, and the number of people who spoke English all being featured in the 1890 census. A sign of things to come..?

In 1896 Hollerith founded the Tabulating Machine Company to exploit his inventions. By this time he had added a mechanism to feed the cards automatically and other automatic sorting procedures which added sophistication to the original simple mechanical counting process.

Herman Hollerith passed away in 1929.

Ironically, having been the child of German immigrants who had fled their home, his invention was to go on and be used in Nazi Germany to assist with a renewed wave of persecutions.

See how punched cards notoriously helped organise the Holocaust.

So the next time you see an ad for something appear on your PC, or a service suggests what your next purchase or viewing preference should be or you receive a piece of unsolicited junk mail from some random political group, remember to thank (or curse) Herman Hollerith and his infernal computation engine!

Amazon, Alphabet, Alibaba, Facebook, Tencent - five of the world's 10 most valuable companies, all less than 25 years old - and all got rich, in their own ways, on data.