The REAL reason Slack became a billion dollar company

And why its business model is evilly brilliant and Twitter is going nowhere

Slack is a dream for every startup founder, investor and VC. It’s the fastest growing startup ever, which is amazing considering it’s a SaaS startup. And all that within its short 2 years of existence. No wonder everyone is analysing how to reproduce this success for their own startup or company.

The first analysis — Slack’s $2.8 Billion Dollar Secret Sauce — that went viral is from Andrew Wilkonson of Metalab, the design agency that helped Slack founder Stewart Butterfield (who was founder of Flickr in case you didn’t know) to turn his code into Slack’s fun product. Being a designer, Andrew’s conclusion is (not that suprising) design focussed: Slack just feels, looks and sounds different than the boring enterprise competition. And that makes it fun to use and turned it into a billion dollar company.

I don’t know about you, but I know tons of startups that look, feel and sound different than their competition and fail miserably.

So pretty soon this answer — Slack’s design is not secret sauce — from Hipchat (competitor) designer Matt Bond popped up. Matt’s conclusions is that the mix of design, product, timing, team, hype and marketing is what made Slack the unicorn company it is today. Sounds like Slack just got lucky then, being there at the right time at the right place with the right people.

And then Slack founder Steward Butterfield posted his (must-read) own experience: From 0 to $1B — Slack’s Founder Shares Their Epic Launch Strategy. His conclusions are that their team focussed on education, feedback, customer happiness and metric analysis drove their success.

I think none of these analysis are spot on. They are definitely not wrong, but they simply count for every successful startup. Now let me tell you the real reason that Slack is a billion dollar company.

The REAL reason

We’ve been using Slack for over a year now at Fileboard (a sales engagement platform) and basically all our internal communication goes via Slack.

So the other day this new team member joined our company, got access to Slack and soon afterwards asked the following question:

“Ok…question…I want to post a comment to a previous specific comment, regarding @satya.vh ‘s posting on May 9, but am not seeing an option to do so…can i not do that, or is this just one long continuous thread???”

Then I remembered when we started using Slack at Fileboard, I really didn’t like it, and it certainly wasn’t fun to use for the exact same reason that my new colleague was pointing out. Slack is just another stream (or streams, if you have more than 1 channel) of continuous information that you have to follow. I remembered I found it distractive and stressful because it was so easy to miss out on stuff.

So then I answered

“Yeah the power of slack is that if you don’t follow it all then time you’ve lost the conversation. So just Slack 24/7 and you’ll be fine :-D”

And then it hit me. This is what drives Slacks’ success. Because if you don’t follow Slack all the time you do not and cannot take part in the conversation with your team members anymore. And that results in:

#1 Social isolation / pressure
Because if you don’t follow Slack all the time, other people reference or know stuff on Slack that you don’t know and you don’t take part in. Within companies it is very important to inform yourself about what’s going on, not only for your job but also for your position within the company and your future ambitions. You start to feel social pressure to follow Slack and post to Slack 24/7.

#2 Addiction
Now you start to follow Slack all the time. It’s addictive, resulting in unconscious stress, because you have the feeling you might miss something. I see colleagues Slack at night, weekends, days off, when their wife is labouring, etc, which basically put Slack on the same level as email, Facebook and Whatsapp.

#3 Single source of information
Everyone is now going all-in on Slack because of reason #1 and #2. More and more information is going into Slack (exactly the reason why Slack has hundreds of integrations and more popping up each week) and the team is heavily invested in it.

And before you know there’s no way out. Slack has become part of your company and your companies’ culture. And then you find out you need to start paying…

Why Slack’s business model is evilly brilliant

Slack’s business model is essentially based on historical messages. The free searchable (and viewable limit) is 10.000 messages before you start paying. If you want to find a historical post on Slack, it only gets you 10k messages back unless you start paying.

Which means, If you didn’t check Slack enough, which put you in social isolation, now there’s no way for you to see a conversation ever again and drag yourself out of social isolation. Unless you pay. And the more people in your company, the faster you reach that 10k messages limit. For example, if each person in your company is posting like 100 messages a day, and your team is 100 people in total, you already need to start paying to avoid social isolation.

Before you know, half the people are missing conversations, find themselves socially isolated, and you start paying licenses because your internal communication and company culture just fails.

It’s brilliantly evil.

Why the Slacks alternatives didn’t work.

Now you probably think, yes but there was Yammer before, and Campfire and Hipchat and Skype. I’ve used them but none of them has been as addictive as Slack.

The reason is that they all either implemented threaded comments or didn’t put a message limit. The result is that you don’t have to check the conversation all the time, because you can always take part in it later. Even years later. There’s no social isolation effect, no addiction effect and no reason to keep checking.

If I’d be in competition with Slack I would implement those message limits immediately on my free plans.

Why Twitter fails

So then I started thinking. Slack is like Twitter for business so why is Twitter failing? Twitter also has this continuous stream of information and if you don’t check Twitter you miss great stuff.
I realised that Twitter misses this social pressure effect because Twitter is focussed on individuals whereas Slack is focussed on teams. If you miss stuff on Twitter, nobody cares. If you miss stuff on Slack, oh man, you’re a bad colleague.

Hence the reason why Facebook is a billion dollar company as well: if you don’t continuously like stuff of your friends, how well a friend are you? Hence, social pressure makes you keep checking and liking the stuff of your friends.

The same counts for Snapchat. You need to keep checking or the messages just disappear.

How to replicate the success

So yes, to replicate Slacks’ success you need a well designed product, an experienced team, perfect timing and amazing execution, as Andrew Wilkonso, Matt Hobb and Slack founder Steward Butterfield rightfully concluded.

But honestly, there are tons of startups out there that do this well and have not gained as much traction as Slack. So what you simply need to take into account when building, marketing and selling a social tool like Slack is psychology. You will become successful if people feel emotionally locked in because of social pressure. And this pressure makes them invest heavily in your product. (Slack actually needs people to invest a mere 2000 message for the effect to take off)

This is the real reason that makes social platforms like Slack, Facebook and Snapchat billion dollar companies. And it is the reason why Twitter is going nowhere except to failure.

Source :

The REAL reason Slack became a billion dollar company

Can Your Genes Make You Happier?


When people think about genes, they tend to apply the words good or bad to them. Someone with talent or unusual beauty « must have good genes, » we say.  And ourselves? Depending on how your life is going, you may thank your genes or feel victimized by them. Most of the time, however, the average person will assume that their genes are a mixture of good and bad elements. Yet this kind of thinking seriously misrepresents how genes work.

Only about 5% of disease-related gene mutations are fully penetrant, the terms geneticists apply when a gene directly causes a disorder. Otherwise, 95% of genes linked to disorders act as an influence. They can sway one way or another, depending on other factors. Your biology doesn’t spell your destiny. You have many choices, because « other factors » include a vast range of influences, including diet, exercise, stress management, and emotional events we take as everyday occurrences.

These influences don’t change the genes you were born with, which remain the same all your life. Instead, what changes is genetic activity, meaning the hundreds of proteins, enzymes, and other chemicals that regulate the cell. As the cell thrives, so does the entire body, and so do you. Imagine that your body is like a social network passing messages back and forth about how your day is going. These messages are received by every cell, so there is total participation even when you assume that a single event–getting a parking ticket, being yelled at, having a bout of indigestion, or being in a bad mood–is separate and isolated.

In reality, no mind-body event is isolated; therefore, genetics is being turned on its head. Instead of being fixed and locked away, our DNA is dynamic and involved. You are speaking to your genes with every thought, word, and action. Experiences are recorded and remembered at the genetic level (such markers are studied in a special field known as epigenetics, which focuses on how sections of DNA are activated or suppressed). Without going into the complexities of genetic activity, a single lesson is emerging: a person’s genes and their lifestyle form a single feedback loop.

An Introduction to Super Genes 

What this means is that you have a choice of what input you want your genes to receive, and secondly, the more positive the input, the more positive the output the genes will send as a response. Tracing lifestyle choices to the genetic level is a recent, extremely exciting development. It helps to make the mind-body connection irrefutable. At the same time, it underlines how important meditation, self-care, and expanded consciousness are, because these are the chief avenues of positive input.

Meditation puts you in direct contact with the source of the mind-body system, which means your thoughts have direct access to beneficial genetic activity. The effects of meditation, such as a sense of being calm, at peace, centered, and unstressed also have crucial importance to how well your cells are functioning, via the genetic activity inside the cells.

Self-care includes all areas of tending to your own well-being. Once you know that a simple action like standing up from your desk every hour sends positive input to your cells, even the smallest input becomes more important. Starting with organic foods, pure water, and clean air, then proceeding to stress management all the way to everyday causes of negative emotions like anger and anxiety, the field of self-care is hugely optimistic, because you get to be the controller of the feedback you receive from your body. Instead of seeing yourself as the victim of your body’s flaws and problems, you form a working alliance that mutually benefits you as a person and every cell in your body.


Expanded awareness opens the door to higher states of consciousness. Through meditation and self-care you unburden yourself of habits, conditioning, negative beliefs, stressful memories, and other kinds of debt to the past. As you unburden yourself, you begin to see that you are alive here and now. Everything that is truly you is expressed only in the present moment. This is the natural perspective of your cells and your genes–they don’t regret the past or anticipate the future. With this awareness, you can fully focus on being here in the now, and as this attitude matures, all the elements of a happy life are nourished, including love, intelligence, creativity, and personal growth.

In some ways what I’ve been discussing is simply a new format for age-old wisdom. The program for achieving happiness through one’s state of awareness isn’t new. What’s new is the reassurance offered by knowing that no effort, however small, is wasted. Your entire lifestyle is recognized at the genetic level, and the more you improve it, the more beneficial the response you will receive.

By Deepak Chopra, MD and Rudolph E. Tanzi, PhD

Can Your Genes Make You Happier?

A quoi correspondent first, second et third party data ?

Dans le monde de la donnée, la segmentation first / second / third reflète avant tout une notion de propriété. La donnée first-party est ainsi celle qui appartient à l’annonceur. La donnée second-party est la propriété d’un partenaire business qui a accepté de l’échanger avec l’annonceur. Enfin la donnée third-party est la propriété de fournisseurs de données qui la revendent aux annonceurs. Augusta vous aide à comprendre les différences entre ces données en terme de disponibilité, de collecte ou d’utilisation.

First-party data

La first-party data est la donnée collectée directement par l’annonceur sur ses actifs. Elle lui appartient. Elle est souvent gratuite. Elle peut-être :

  • Personnelle : données déclaratives (email dans un formulaire) ou comportementales (historique de navigation sur un site ou une appli loggée…)
  • Anonyme : bannière publicitaire vue ou cliquée, historique de navigation non loggé…


L’annonceur dispose de nombreux outils pour collecter de la donnée first-party. Les suites marketing 360 pour les données personnelles (Selligent, Splio, Adobe, Oracle…) via leurs outils d’emailing, de web analytics, etc. La donnée est alors stockée en base directement dans l’outil ou mieux dans un RCU, un Référentiel Client Unique (mis en place par des sociétés comme ETO ou Camps de Base). En ce qui concerne les données non-déclaratives, les annonceurs disposent de la DMP (Data Management Platform) : Adobe, Oracle, Weborama, Ysance, Krux, Turn… La DMP fait office d’outil de collecte (via notamment un tag) mais aussi de stockage.


La donnée first est stratégique pour l’annonceur : elle est de très bonne qualité, gratuite (si l’on exclut le coût des outils) et facilement accessible. Elle sert aussi de base à toutes les actions média, notamment sur les “walled-garden” comme Facebook et Google où les annonceurs peuvent injecter leurs données first-party pour cibler leurs clients ou des profils qui leur sont similaires.


Tous les annonceurs ne sont pas forcément égaux devant la first-party data : certains croulent sous la donnée(opérateurs de téléphonie) alors que d’autres (Produits de Grande Consommation) en ont très peu car ils n’ont pas de lien direct avec le consommateur. La second-party data peut-être une solution pour eux.



Second-party data

La second-party data est la fist-party data d’un partenaire business qui a accepté de la partager avec l’annonceur dans le cadre d’un partenariat second-party. Des annonceurs comme Danone, Nestlé ou Mondelez qui ont peu accès au consommateur et dont le trafic sur site est limité peuvent nouer des partenariats avec des distributeurs (Carrefour, Auchan…) ou des sites de contenu (par exemple des sites de recette de cuisine). Ils récupéreraient ainsi des données d’achat (transaction sur les sites de drive) et de comportement (recettes consultées) qui leur permettraient de mieux cibler leurs publicités.


La récupération de cette donnée est encore assez complexe. Plusieurs possibilités existent. Si l’annonceur est équipé d’une DMP, il pourra poser le tag de sa DMP sur le site de son partenaire pour faire remonter la donnée du partenaire directement dans sa DMP. Si le partenaire a un système de Tag Management, cela peut être effectué très rapidement. Certaines DMP (Krux, Adobe) ont mis en place des fonctionnalités d’échange automatisée de données, à condition que les deux partenaires utilisent la même technologie. Sans DMP, la chose se complique et il faut passer par des solutions du type Weborama.


Le modèle économique de ces partenariats n’est pas encore figé. Plus que l’aspect technique, cet aspect économique peut s’avérer bloquant. Plusieurs configurations sont possibles, du simple échange gratuit à des échanges avec des contreparties financières ou de l’achat média vers le site qui fournit la donnée.


La second-party data constitue donc une donnée extrêmement pertinente et souvent très complémentaire de la donnée first-party d’un annonceur. C’est un formidable accélérateur pour les annonceurs ayant peu de données first-party.



Third-party data

La third-party data est la donnée achetée à des fournisseurs de données. Les plus connus sont Acxiom, Exelate, Datalogix, Weborama… Des acteurs comme Facebook ou Twitter peuvent aussi être considérés comme des fournisseurs de données dans le sens où ils permettent à l’annonceur d’affiner la personnalisation de son marketing. Il s’agit souvent de données socio-démographiques (âge, sexe, revenus…) même si des données plus spécifiques existent (intentionistes de voitures par exemple chez Acxiom via son partenariat avec AAA).


La donnée 3rd party est utilisée essentiellement dans l’achat programmatique pour enrichir les données d’un annonceur lors d’une enchère. Elle est très facilement accessible à partir du moment où vous payez : il suffit de cocher les catégories qui vous intéressent sur votre DSP ou votre DMP. La facturation s’effectue ensuite au CPM (à chaque fois que vous faites de l’achat média sur un segment enrichi par de la third-party on vous facture). Les tarifs pratiqués sont de l’ordre de quelques euros au CPM.


Très en vogue durant les dernières années, la third-party subit un retour de hype. Très liée aux cookies, assez chère, de qualité souvent disparate, la réalité diffère parfois de la théorie. Les fournisseurs l’ont compris et se spécialisent sur des types de donnée très précis (Datalogix avec les sorties de caisse aux États-Unis, Temelio avec l’onboarding…) ou des cas d’usage particulièrement efficaces (exploration d’audience avec l’outil Cyclone de Weborama). Enfin ces fournisseurs subissent une concurrence de plus en plus forte de la part d’acteurs comme Facebook qui ont une donnée plus pérenne (car basée sur l’email), plus globale (présence mondiale) et souvent très qualifiée.



Les annonceurs prennent conscience de l’aspect stratégique de la first-party. Il s’agit de la donnée la plus riche et la plus fiable. Mais elle permet aussi de limiter la dépendance à des fournisseurs de données qui pourraient décider du jour au lendemain de doubler leurs tarifs. La donnée third apparaît néanmoins comme un bon complément à la first, par exemple pour élargir son audience. Enfin, la second-party est pleine de promesses notamment pour des annonceurs ayant peu de first party mais le modèle économique reste encore à définir.

A quoi correspondent first, second et third party data ?

13 Stats That Should Terrify CMOs

I find numbers terrifying. In college, I skipped more than half of the classes for my required math course—not because I was some sort of badass, but because learning the ins and outs of concepts like “correlation coefficient” and “factorial” stirred a fear deep within me.

So when marketers talk about how much they work with numbers now that we’ve entered the sci-fi1 age of Internet data collection, I cringe. Let’s just say I’m glad I’m covering marketing for a living rather than actually doing it.

But just because numbers may be marketers’s new best friend, that doesn’t mean they’re all good. Some, in fact, are a bit terrifying.

And because I’m a sick bastard who wants you all to feel my mathematical pain, I’ve gathered 13 stats that should send chills down any marketer’s spine. Just to be clear, all of these statistics have pros and cons, and some stats may not have an impact on your marketing strategy. But hopefully at least one of them will get you to jump. Enjoy.

1. Standard banner ads have a 0.12 percent CTR

Source: eMarketer

Banner ads have been around since the Internet’s earliest days, and by all reports, they’ve never been all that effective—that shouldn’t come as news to anyone.

What’s incredible is just how ineffective they really are. I could make a joke that the only time users click on a banner ad is when they pass out and bounce their head off the mouse. According to the data, that’s actually pretty much accurate.

But it’s worth pointing out that CTR isn’t everything. Plus, rich banner ads (basically anything with interactivity, animation, and/or video elements) have a 0.44 percent CTR, which is 267 percent increase. Which still isn’t great, but hey, 267 percent!

Despite these numbers, banner ads are predicted by eMarketer to explode on mobile in the coming years (mostly as a result of increased mobile ad spend in general). On desktop, however, ad spend on banner ads is set to stagnate, along with desktop ad spend in general.

Which brings me to our next stat…

2. Mobile banner ads have a 0.14 percent CTR

Even as money shifts from desktop to mobile, the efficacy of banner ads isn’t going to improve. That 0.02 percent bump is probably due more to how easy it is to accidentally click on a mobile banner ad than anything.

Still, it’s important to note that, on average, banner ads aren’t a huge part of the media-buying war chest. Estimates predict that rich banner ads, the more effective of the two formats, will make up only about nine percent of total ad spend in 2015.

3. Adblocking grew by 41 percent YoY in the past 12 months


This is a highly contested number, as the companies behind the research, Adobe and PageFair, have financial incentive to make adblocking seem more dire than it is. But even if these numbers are exaggerated—and arecent study from the Reuters Institute for the Study of Journalism at the University of Oxford indicates that they are not—the adblocking problem is serious.

As David Kint, CEO of media trade organization Digital Content Next, which represents publishers for the advertising community, recently told me: “It’s real, and it’s tied to a tech arms race happening because the consumer has been ignored for too long. Current adblocking revenue projections dwarf the entirety of digital advertising revenues across our entire membership.”

While adblocking affects the bottom line of media and Internet companies more than ad buyers, the way the public has taken to anti-ad technology suggests users aren’t happy with Internet advertising. And if ads are annoying potential customers more than enticing them to buy your product, you’re spending a lot of counterproductive money.

4. Safari makes up 55 percent of the mobile browsing share

On the surface, this stat doesn’t mean much. But with the news that Apple is planning to launch the iOS 9 version of Safari with adblocking capabilities, the mobile ad market could take a serious hit.

While money will likely just flow elsewhere—social media apps, for example—ad buyers reliant on cheap mobile ads and Google search ads won’t be pleased that a large segment of the mobile audience (and affluent ones, at that) will be out of reach. Apple’s move could also set a trend of mobile adblocking in motion, though Google will likely push back at any attempt to popularize adblocking on Android given how crucial ad revenue is to its business model.

5. 28 percent of users said they tried to hide their activities from advertisers, second only to criminals

Source: Pew Research

As online privacy becomes more important to Americans post-Snowden, marketers and advertisers have turned into potential villains.

Unfortunately, this survey was taken in 20132, and there is no updated version, but it’s unlikely that users have warmed up to being tracked by ad and marketing technology.

Making users resent your ads is not a productive use of resources, and it’s an issue all Internet companies (social media companies, in particular) will have to think about as data collection becomes more and more rampant.

6. By 2018, chat apps will make up 75 percent of mobile messaging traffic share—but only 2 percent of revenue

Source: eMarketer

Chat apps like WhatsApp, Facebook Messenger, and Kik may be growing exponentially in the U.S., but that doesn’t mean marketers will be getting in on the action anytime soon.

As more and more people—particularly teens—choose to communicate over more private micro-networks rather than SMS or social media, advertisers, at least at first, may lose out. As this stat demonstrates, chat apps are often built on a business model of audience growth first, monetization second—which means marketers will have to wait for apps to reel their audience in with strong user experiences before they can get involved.

But there is some good news—Facebook has made clear that it plans to eventually monetize WhatsApp and (potentially) Messenger, two of the most popular chat apps, once they reach 1 billion users.

7. Pay-TV households are now shrinking at an annual rate of 0.7 percent

This stat isn’t nearly as clear cut as others since the move away from traditional cable has upsides: Digital ads tend to be cheaper and come with benefits such as better data collection.

Still, the gradual migration to over-the-top (OTT) TV services could end up affecting ad penetration. Anyone using an adblocker on video sites like YouTube isn’t going to see your ad, whereas on cable TV, they have no choice (except for getting up and leaving the room or watching shows on DVR).

There’s also the fact that the most popular digital video services—Netflix,YouTube, and Hulu—will, or already do, offer ad-free options. Considering that people spend more time with digital video than any other digital content, that’s a huge amount of time consumers are going to spend without an ad in sight. Great for users—bad for advertisers.

8. Only 12 percent of marketers believe they have “high-performance” content marketing engines

Even worse? Of 200 senior marketing leaders, only two percent “consider their existing approach to demand generation highly effective,” per the CMO Council.

As Tessa Wegert discussed in a recent article, the stats don’t look good, but it’s important to remember that content marketing is still in the fetal stage, so check out her article if you’re part of the 88 percent. And if you’re part of the 12 percent, well, you should give yourself a big pat on the back.

9. For every 100,000 followers on Facebook, only 130 people will click on an organic post

In other words, if you’re still trying to live on organic social media traffic, it’s probably time to give up.

Facebook, Twitter, and LinkedIn have all moved to a paid social media model, and spending for reach and clicks on specific posts is almost always more effective than than spamming as many posts as possible, as Shane Snow demonstrates in this fascinating breakdown of paid versus organic social media sharing.

10. “45 percent of marketers still either don’t formally evaluate their analytics for quality and accuracy or, even worse, don’t know if they do or not.”

It appears that even though analytics have taken center stage in the marketing world, some practitioners aren’t checking their work.

That’s a gigantic problem since blind trust of numbers can be just as damaging as trusting whatever the latest self-proclaimed “marketing guru” posted on LinkedIn.

Check out VentureBeat’s “State of Marketing Analytics” for more.

11. 56 percent of display ads are never seen by humans

If it seems like I’m picking on Internet ads here, you’d be right. Unfortunately for advertisers, there are just a lot of problems facing display ads, including this fun stat from Google that proves that a majority of ad spend is going to waste.

I guess John Wanamaker was right.

The major issue here is viewability—some publishers run ads that fall outside of the IAB’s relatively lax standards of what constitutes a “viewable” ad. Another issue is the incredible amount of fraudulent bots, which have exploded in the past few years.

Speaking of which…

12. Fraudulent bots wasted $11.6 billion of ad spend in 2014

Traffic bots have become a big industry as the focus on clicks and impressions in digital advertising has allowed criminals to send fake traffic and drive programmatic ad revenue.

Jordan Teicher discusses the issue in more detail here, but make no mistake it’s one that has flown under the radar compared to more visible problems like adblocking and viewability standards.

13. Two-thirds of readers have felt deceived upon realizing that an article or video was sponsored by a brand

Deception isn’t something you want customers associating with your brand, but it’s one of the many downsides of native advertising and sponsored posts we discovered last year in our extensive study.

While it’s true that native advertising does have its benefits—and that certain practitioners like The New York Times‘s T Brand Studio have created proven, effective native ads—the thin line between editorial and advertorial can be damaging for both publisher and advertiser. And when ad buyers are spending fortunes for native ads, the leeway for negative results like mistrust can be hard to swallow.

Luckily, the solution is pretty obvious: better labeling. Brands should feel proud to put their names next to a great piece of native advertising, and publishers should want to better separate their journalism from their ads. Only then will the numbers start to work in their favor.

  1.  or dystopian, depending on your perspective
  2. I’m sure government has since shot up the list

source :

13 Stats That Should Terrify CMOs

Digital Taylorism

A modern version of “scientific management” threatens to dehumanise the workplace

FREDERICK TAYLOR was the most influential management guru of the early 20th century. His “Principles of Scientific Management” was the first management blockbuster. His fans included Henry Ford, who applied many of his ideas in his giant River Rouge car plant, and Vladimir Lenin, who regarded scientific management as one of the building blocks of socialism. Taylor’s appeal lay in his promise that management could be made into a science, and workers into cogs in an industrial machine. The best way to boost productivity, he argued, was to embrace three rules: break complex jobs down into simple ones; measure everything that workers do; and link pay to performance, giving bonuses to high-achievers and sacking sluggards.

Scientific management provoked a backlash. Aldous Huxley satirised it in “Brave New World” (1932), as did Charlie Chaplin in “Modern Times” (1936). A rival school of managers argued that workers are more productive if you treat them as human beings. But a recent article about Amazon in the New York Times suggests that Taylorism is thriving. The article claimed that the internet retailer uses classic Taylorist techniques to achieve efficiency: workers are constantly measured and those who fail to hit the numbers are ruthlessly eliminated, personal tragedies notwithstanding. Amazon’s boss, Jeff Bezos, insisted that he did not recognise the company portrayed in the piece. Nevertheless, it provoked quite a reaction: the article attracted more than 5,800 online comments, a record for a Times article, and a remarkable number of commenters claimed that their employers had adopted similar policies. Far from being an outlier, it would seem that Amazon is the embodiment of a new trend, digital Taylorism.

This new version of Taylor’s theory starts with his three basic principles of good management but supercharges them with digital technology and applies them to a much wider range of employees—not just Taylor’s industrial workers but also service workers, knowledge workers and managers themselves. In Taylor’s world, managers were the lords of creation. In the digital world they are mere widgets in the giant corporate computer.

Technology allows the division of labour to be applied to a much wider range of jobs: companies such as Upwork (formerly oDesk) are making a business out of slicing clerical work into routine tasks and then outsourcing them to freelances. Technology also allows time-and-motion studies to be carried to new levels. Several firms, including Workday and Salesforce, produce peer-review software that turns performance assessments from an annual ritual into a never-ending trial. Alex Pentland of the Massachusetts Institute of Technology has invented a “sociometric” badge, worn around the neck, that measures such things as your tone of voice, gestures and propensity to talk or listen. Turner Construction is using drones to monitor progress on a sports stadium it is building in California. Motorola makes terminals that strap to warehouse workers’ arms to help them do their jobs more efficiently—but could also be used to keep tabs on them.

As stopwatch management continues to conquer new territory, so too does pay for performance. The more firms depend on the brainpower of their employees, the more they are seeking to reward their finest minds with high salaries and stock options. “A great lathe operator commands several times the wage of an average lathe operator,” Bill Gates points out, “but a great writer of software code is worth 10,000 times the price of an average software writer.” Many firms, including Amazon, apply the same Darwinian logic to their worst performers as well, in a process known as “rank and yank”: workers are regularly ranked by productivity and the weakest are culled.

The reaction to the Times piece shows that digital Taylorism is just as unpopular as its stopwatch-based predecessor. Critics make some powerful points. “Gobbetising” knowledge jobs limits a worker’s ability to use his expertise creatively, they argue. Measuring everything robs jobs of their pleasure. Pushing people to their limits institutionalises “burn and churn”. Constant peer-reviews encourage back-stabbing. Indeed, some firms that graded their staff, including Microsoft, General Electric and Accenture, concluded that it is counter-productive, and dropped it.

The meatware fights back

The march of technology can cut both ways. The rise of smart machines may make Taylorism irrelevant in the long term: why turn workers into machines when machines can do ever more? The proliferation of websites such as Glassdoor, which let employees review their workplaces, may mean that firms which treat their workers as mere “meatware” lose the war for the sort of talent that cannot be mechanised. And Mr Pentland’s sociometric badges have produced some counter-intuitive results: for example, in a study of 80 employees in a Bank of America call centre, he found that the most successful teams were the ones that spent more time doing what their managers presumably didn’t want them to do: chatting with each other.

Even so, digital Taylorism looks set to be a more powerful force than its analogue predecessor. The prominent technology firms that set the tone for much of the business world are embracing it. Google, which hires a few thousand people a year from up to 3m applicants, constantly ranks its employees on a five-point scale. Investors seem to like Taylorism: Amazon’s share price ticked upwards after the Times’s exposé. The onward march of technology is producing ever more sophisticated ways of measuring and monitoring human resources. And Taylorist managers are mixing the sweet with the bitter: Amazon’s “Amabots”, as they call themselves, seem happy to put up with micromanagement if they get a nice bonus at the end of the year. The most basic axiom of management is “what gets measured gets managed”. So the more the technology of measurement advances, the more we hand power to Frederick Taylor’s successors.

Digital Taylorism