In a Pandemic, We Buy What We Know


This post is by Chelsea Galoni from HBR.org

How fear and disgust drive consumer behavior.

How Jeff Bezos Built One of the World’s Most Valuable Companies


This post is by HBR.org from HBR.org

A conversation with HBS professor Sunil Gupta on what makes Amazon so successful.

When Should Health Systems Invest in New Tech?


This post is by John Glaser from HBR.org

Beware of shiny objects.

Don’t Focus on Digital Transformation; Focus on Quick, Strategic Wins – SPONSOR CONTENT FROM SLALOM


This post is by HBR.org from HBR.org

Sponsor content from Slalom.

4 Ways to Reconfigure Your Sales Strategy During the Pandemic


This post is by Scott Edinger from HBR.org

Approaches that will give your team an edge.

A Practical Guide to Building Ethical AI


This post is by Reid Blackman from HBR.org

AI doesn’t just scale solutions — it also scales risk.

The Dreaded Pivot


This post is by Frederic Filloux from Monday Note - Medium

THE DONE SAGA, EP. 07

Feedback from users and investors has been unambiguous. In the Done app, there is only one killer feature, the calendar system. Over the last two months, we pivoted the company around it — and learned a lot.

by Gilles Raymond and Frederic Filloux

Photo by Ondrej Trnak on Unsplash

This series recounts the creation of the startup Done, from the first idea to the final product, with its best and worst moments (past and coming). No bullshit. Promised.

Episode 1, “The Two Foundations” After a near-death experience, the sale of News Republic to a giant Chinese tech company.
Episode 2, “Three Chinese lessons” What it really means to work for a Chinese company. The need to rethink how we interact digitally.
Episode 3, “Strategic Branding” How Done drew inspiration from the hospitality industry.
Episode 4, “The Design Sprint” A product in five days, from the idea to testing.
Episode 5, “The Hard Thing About Funding” Nine months of toil and rejection.
Episode 6, “Tribe Selling” The mechanism of a product’s mass adoption

I detected the signs early on in the life of the company. Every pitch I made to potential investors brought the same conclusion: “…Your key differentiator is your clever system to quickly arrange meetings, the rest could be found anywhere”.

Done was meant to be “The only app you need for work” thanks to the integration of a chat system, a task manager, a file-sharing feature, and a calendar. That one was the true gem of the application, at least according to the VCs I probed. In short, we had devised a super fast way to set up meetings with a single swipe, where our competitors required multiple interactions, and sometimes even cryptic instructions that looked like a line of code.

Then came the market confirmation with good and not so good returns. On the plus side, we went from 2,000 “real” users in June to 10,000 in September, a natural growth achieved without a shred of marketing expenses. Our core users still spend about 20 minutes each day on the app, not bad for a service that had yet to achieve a network effect. But the churn was worse than our competitors. And COVID didn’t help either: our product was designed to be primarily for mobile usage, with an interface based on quick taps and swipes. But with the pandemic and the explosion of remote working, people went back to the desktop with its multiple tabs often displayed on a large screen.

Time to Pivot

Create, experiment, learn, and adapt are the key components of tech culture. It entails a much higher tolerance for failure. This trait is the biggest impediment to the tech revolution in Europe, where I came from. There, a business failure is prohibitive and the culprit behind one can be barred from rebuilding anything for years.

Being prepared to pivot is part of managing a startup. It is a frequent twist of the tech narrative: Slack was originally an internal communication tool used at Tiny Speck, a gaming startup. YouTube was a video dating site. Twitter a podcast downloading platform. Instagram was originally a FourSquare-like service.

Pivoting rarely involves joy and exhilaration. For the team, it is always a traumatic experience with months of toiling going down the drain. Sometimes it comes along with finger-pointing or criticism for initiating the move too late or too soon.

At Done, it went relatively well. Probably because we are a small team, always aware of the pulse of the business. While the heartbeat was faint early this summer, we knew we had the defibrillator to restart: the meeting setup feature that was pointed out by VCs and customers was the true jewel of our service. We had sixty days to build it.

Pain Point

About 56 million meetings are organized every day in the United States alone, according to a research by LogMeIn. And everyone agrees that it is a universal pain point. We confirmed it through various polls and surveys made on Facebook and Google and the conclusion was that it was a great nut to crack.

Of course, as Apple’s marketing stated, “there is an app for that”. Several actually. The most used and the most powerful are functionalities of platforms like Google Suite, Microsoft Teams, or Slack. They work fine as long as everybody interacts in the same bowl. Once you want to include people from the outside world things get complicated and third-party apps need to be called in. Calendly and Doodle work just fine. I witnessed people from Google using Doodle to arrange a 15-person meeting of the media industry… However — and I say it with all due respect to their vast user base — if these two services were a car, they would be more a Trabant than a Tesla. And in 2020, no one would be satisfied with the iconic car from East Germany.

Hence our idea: building a service that will do the tedious work for you. We called it Meeting’s Monster.

Right now, organizing a meeting goes like this: you or an assistant will contact the participants to see their availability. Again, it’s easy within the company. When schedules aren’t shared it is done by a string of emails, instant messaging or even phone:

“John: we should evaluate the plan, may I suggest a meeting on Tuesday or Thursday? 10 AM
– Alicia: Tuesday is fully booked, can we try Thursday 11 AM
– Eddy: does not work for me, Friday PM will be great..”
– John : I’m sorry, I’m off this Friday PM… I will set up a Doodle for next week
– Eddy : Go for it, I will try to answer quicker than last time…”

And so on… And that is only for a meeting with three people.

The preliminary conversation will grow exponentially based on the number of attendants based on the formula in which the level of pain (P) is equal to Meeting importance (M) to the power of the Attendees (A):

Delving into details shows the innate complexity of what looks in theory so simple and so routine. Setting up a meeting carries a significant number of moving parts and untold rules.

For instance: I want to set up a session involving an external partner and people from marketing (to discuss the strategy), finance (we are about to spend a lot), and legal (only the paranoid survive). Here are the usual constraints: X often has business lunches, meaning he must be free between 12:30 and 2:30 pm. Y’s office is 45 min away (best case), 1h20 (worse case), so if we want her in person, we better take that into account. Z said that Wednesday afternoon should remain meeting-free and Friday is WFH only. Another one doesn’t want meetings back to back, etc.

Then there are more subtle factors: in this meeting, Sarah, Mathilde, and Henry are “Must have people” while six others are “Nice to have.” But what if none of the latter are available? Should we keep the meeting? Even more delicate, if Sarah is in, Jonas must be invited otherwise he will, once again, go nuclear on Sarah. But if Jonas and Lisa are in the same room, sparks will fly (Jonas is decidedly a pain, but that’s another subject).

Any human assistant will factor in these things and act accordingly. They will know that this meeting is an important one and should take precedence over others that week. They will cleverly manage egos and competencies.

Let’s be clear, with Meeting’s Monster, we do not pretend to solve all those delicate issues, but we intend to get as close to it as possible. More importantly, we want to build a system that will allow the user to simply ask: “I need an hour with these two people in the next 48 hours [or] this group by the end of the month”. And the service will take care of it.

On our new app, it looks like this:

Most of the heavy lifting is performed under the hood.

Rules vs. Behavior

In order to address constraints and exceptions, the simplest route is to build a catalog of rules that will govern the way that people organize their meetings. Again, we will need to take into account many usages that could vary from one country to another, based on culture, or working habits. We did the chart. Its complexity rivals the map of the Tokyo subway. Hence the questions: how deep should we go with the rules? Which ones should be suggested or left to the users, depending on their sectors and location?

The set of rules approach has its limitations. They can be overwhelming and profuse and yet not nearing exhaustivity… Other industries have gone through similar hurdles. Ten years ago, Natural Language Processing and text-generation were largely rule-driven. And the pioneers of the self-driving car tried first to teach the software how to drive. In both cases, artificial intelligence, relying on a large dataset of behaviors, settled the situation.

Another option is then to have the software look at the user’s meeting history to detect patterns resulting from individuals habits and general practices of the company. There is no doubt we will find valuable inputs that will make our software better. Then reality set in, with issues like the exploitability of the data and also some obvious privacy issues (though these will be solved through a carefully designed opt-in process).

The upheaval we see in everyone’s working habits, especially with the rise of online interactions, will help us a lot. They will be way easier to use and to interpret for companies like ours. Then we will need to rely on the law of diminishing returns: past a certain point, it will be useless to work on features that will be rarely needed by the customer. The question is where to set the cursor.

For the rest, there is both a vast addressable market and a painful friction to overcome. Sounds like a great business opportunity and we intend to grab it.

Gilles Raymond & Frederic Filloux


The Dreaded Pivot was originally published in Monday Note on Medium, where people are continuing the conversation by highlighting and responding to this story.

The Dreaded Pivot


This post is by Frederic Filloux from Monday Note - Medium

THE DONE SAGA, EP. 07

Feedback from users and investors has been unambiguous. In the Done app, there is only one killer feature, the calendar system. Over the last two months, we pivoted the company around it — and learned a lot.

by Gilles Raymond and Frederic Filloux

Photo by Ondrej Trnak on Unsplash

This series recounts the creation of the startup Done, from the first idea to the final product, with its best and worst moments (past and coming). No bullshit. Promised.

Episode 1, “The Two Foundations” After a near-death experience, the sale of News Republic to a giant Chinese tech company.
Episode 2, “Three Chinese lessons” What it really means to work for a Chinese company. The need to rethink how we interact digitally.
Episode 3, “Strategic Branding” How Done drew inspiration from the hospitality industry.
Episode 4, “The Design Sprint” A product in five days, from the idea to testing.
Episode 5, “The Hard Thing About Funding” Nine months of toil and rejection.
Episode 6, “Tribe Selling” The mechanism of a product’s mass adoption

I detected the signs early on in the life of the company. Every pitch I made to potential investors brought the same conclusion: “…Your key differentiator is your clever system to quickly arrange meetings, the rest could be found anywhere”.

Done was meant to be “The only app you need for work” thanks to the integration of a chat system, a task manager, a file-sharing feature, and a calendar. That one was the true gem of the application, at least according to the VCs I probed. In short, we had devised a super fast way to set up meetings with a single swipe, where our competitors required multiple interactions, and sometimes even cryptic instructions that looked like a line of code.

Then came the market confirmation with good and not so good returns. On the plus side, we went from 2,000 “real” users in June to 10,000 in September, a natural growth achieved without a shred of marketing expenses. Our core users still spend about 20 minutes each day on the app, not bad for a service that had yet to achieve a network effect. But the churn was worse than our competitors. And COVID didn’t help either: our product was designed to be primarily for mobile usage, with an interface based on quick taps and swipes. But with the pandemic and the explosion of remote working, people went back to the desktop with its multiple tabs often displayed on a large screen.

Time to Pivot

Create, experiment, learn, and adapt are the key components of tech culture. It entails a much higher tolerance for failure. This trait is the biggest impediment to the tech revolution in Europe, where I came from. There, a business failure is prohibitive and the culprit behind one can be barred from rebuilding anything for years.

Being prepared to pivot is part of managing a startup. It is a frequent twist of the tech narrative: Slack was originally an internal communication tool used at Tiny Speck, a gaming startup. YouTube was a video dating site. Twitter a podcast downloading platform. Instagram was originally a FourSquare-like service.

Pivoting rarely involves joy and exhilaration. For the team, it is always a traumatic experience with months of toiling going down the drain. Sometimes it comes along with finger-pointing or criticism for initiating the move too late or too soon.

At Done, it went relatively well. Probably because we are a small team, always aware of the pulse of the business. While the heartbeat was faint early this summer, we knew we had the defibrillator to restart: the meeting setup feature that was pointed out by VCs and customers was the true jewel of our service. We had sixty days to build it.

Pain Point

About 56 million meetings are organized every day in the United States alone, according to a research by LogMeIn. And everyone agrees that it is a universal pain point. We confirmed it through various polls and surveys made on Facebook and Google and the conclusion was that it was a great nut to crack.

Of course, as Apple’s marketing stated, “there is an app for that”. Several actually. The most used and the most powerful are functionalities of platforms like Google Suite, Microsoft Teams, or Slack. They work fine as long as everybody interacts in the same bowl. Once you want to include people from the outside world things get complicated and third-party apps need to be called in. Calendly and Doodle work just fine. I witnessed people from Google using Doodle to arrange a 15-person meeting of the media industry… However — and I say it with all due respect to their vast user base — if these two services were a car, they would be more a Trabant than a Tesla. And in 2020, no one would be satisfied with the iconic car from East Germany.

Hence our idea: building a service that will do the tedious work for you. We called it Meeting’s Monster.

Right now, organizing a meeting goes like this: you or an assistant will contact the participants to see their availability. Again, it’s easy within the company. When schedules aren’t shared it is done by a string of emails, instant messaging or even phone:

“John: we should evaluate the plan, may I suggest a meeting on Tuesday or Thursday? 10 AM
– Alicia: Tuesday is fully booked, can we try Thursday 11 AM
– Eddy: does not work for me, Friday PM will be great..”
– John : I’m sorry, I’m off this Friday PM… I will set up a Doodle for next week
– Eddy : Go for it, I will try to answer quicker than last time…”

And so on… And that is only for a meeting with three people.

The preliminary conversation will grow exponentially based on the number of attendants based on the formula in which the level of pain (P) is equal to Meeting importance (M) to the power of the Attendees (A):

Delving into details shows the innate complexity of what looks in theory so simple and so routine. Setting up a meeting carries a significant number of moving parts and untold rules.

For instance: I want to set up a session involving an external partner and people from marketing (to discuss the strategy), finance (we are about to spend a lot), and legal (only the paranoid survive). Here are the usual constraints: X often has business lunches, meaning he must be free between 12:30 and 2:30 pm. Y’s office is 45 min away (best case), 1h20 (worse case), so if we want her in person, we better take that into account. Z said that Wednesday afternoon should remain meeting-free and Friday is WFH only. Another one doesn’t want meetings back to back, etc.

Then there are more subtle factors: in this meeting, Sarah, Mathilde, and Henry are “Must have people” while six others are “Nice to have.” But what if none of the latter are available? Should we keep the meeting? Even more delicate, if Sarah is in, Jonas must be invited otherwise he will, once again, go nuclear on Sarah. But if Jonas and Lisa are in the same room, sparks will fly (Jonas is decidedly a pain, but that’s another subject).

Any human assistant will factor in these things and act accordingly. They will know that this meeting is an important one and should take precedence over others that week. They will cleverly manage egos and competencies.

Let’s be clear, with Meeting’s Monster, we do not pretend to solve all those delicate issues, but we intend to get as close to it as possible. More importantly, we want to build a system that will allow the user to simply ask: “I need an hour with these two people in the next 48 hours [or] this group by the end of the month”. And the service will take care of it.

On our new app, it looks like this:

Most of the heavy lifting is performed under the hood.

Rules vs. Behavior

In order to address constraints and exceptions, the simplest route is to build a catalog of rules that will govern the way that people organize their meetings. Again, we will need to take into account many usages that could vary from one country to another, based on culture, or working habits. We did the chart. Its complexity rivals the map of the Tokyo subway. Hence the questions: how deep should we go with the rules? Which ones should be suggested or left to the users, depending on their sectors and location?

The set of rules approach has its limitations. They can be overwhelming and profuse and yet not nearing exhaustivity… Other industries have gone through similar hurdles. Ten years ago, Natural Language Processing and text-generation were largely rule-driven. And the pioneers of the self-driving car tried first to teach the software how to drive. In both cases, artificial intelligence, relying on a large dataset of behaviors, settled the situation.

Another option is then to have the software look at the user’s meeting history to detect patterns resulting from individuals habits and general practices of the company. There is no doubt we will find valuable inputs that will make our software better. Then reality set in, with issues like the exploitability of the data and also some obvious privacy issues (though these will be solved through a carefully designed opt-in process).

The upheaval we see in everyone’s working habits, especially with the rise of online interactions, will help us a lot. They will be way easier to use and to interpret for companies like ours. Then we will need to rely on the law of diminishing returns: past a certain point, it will be useless to work on features that will be rarely needed by the customer. The question is where to set the cursor.

For the rest, there is both a vast addressable market and a painful friction to overcome. Sounds like a great business opportunity and we intend to grab it.

Gilles Raymond & Frederic Filloux


The Dreaded Pivot was originally published in Monday Note on Medium, where people are continuing the conversation by highlighting and responding to this story.

When Efficiency Goes Too Far


This post is by HBR.org from HBR.org

A conversation with Rotman professor emeritus Roger Martin on why leaders should stop treating companies like machines.

You Don’t Have to Pivot in a Crisis


This post is by Daniel Isenberg from HBR.org

How to stay the course.

How Amazon Automated Work and Put Its People to Better Use


This post is by Alex Kantrowitz from HBR.org

Its “Hands Off the Wheel” initiative proves that AI doesn’t have to replace employees.

10 Unwritten Rules for Venture Savvy Founders


This post is by Parul Singh from Startup Grind - Medium

Because there’s no user manual for venture capital.

Why Hasn’t Apple Pay Replicated Alipay’s Success?


This post is by Ian Gross from HBR.org

Two key drivers for successful platform adoption.

An Agile Approach to Budgeting for Uncertain Times


This post is by Darrell K. Rigby from HBR.org

The pandemic is making the annual process even more challenging.

The Secret to AI Is People


This post is by Nada R. Sanders from HBR.org

It’s not just a plug-and-play technology.

SaaS Generations 1, 2, and 3


This post is by David Cummings from David Cummings on Startups

Last week I listened to an interview of Benchmark partner Eric Vishria where he talked about the three generations of Software-as-a-Service (SaaS). I like how he framed the idea and wanted to capture the concept here. I’ve thought deeply about the first and second generations of SaaS, but haven’t spent enough time thinking about the third. Let’s dive in.

1st Generation SaaS

The 1st generation of SaaS is defined as moving legacy enterprise applications into the cloud and delivered as a service. Sometimes the technology is literally the legacy solution (previously called Application Service Provider) but more often it’s a new product that’s natively SaaS. Like traditional enterprise software, it’s sold by salespeople, typically with contracts and the usual enterprise process. The huge innovation is the ability of the software vendor to abstract away most of the customer headaches that come with managing software and deliver it as a monthly or annual fee. This, combined with more efficient product development, due to customers always being on the same version, makes for an excellent business and customer model.

Examples: Pardot, Salesforce

2nd Generation SaaS

The 2nd generation of SaaS reimagines business software as more of a consumer product where you can sign up for a free trial or a free, limited version without talking to a salesperson and become productive immediately. This category is often called bottoms-up or freemium as it starts with functional business users getting a job done and spreads in organizations without I.T. involvement. 2nd generation SaaS is bought differently than 1st generation SaaS and feels more natural as both a buyer and user of the product.

Examples: Calendly, Slack

3rd Generation SaaS

The 3rd generation of SaaS is SaaS delivered exclusively as APIs — a way for software programs to talk to other software programs without a user interface. APIs are the building blocks of modern software development representing re-usable components that make programming faster and more productive. APIs can now power many aspects of software that 5-10 years ago would have been custom including user authentication/oAuth, sending/receiving email, video conferencing, payment processing, recurring billing, testing, visualizations, reporting, analytics, and much more.

Examples: Twilio, Sendgrid

All three generations of SaaS are growing fast and have tremendous runway ahead. Look for the 2nd and 3rd generations to grow even faster and become more commonplace.

SaaS is an incredible segment of the software industry and understanding the three generations helps frame the thinking about product types.

Future-Proofing Your Strategy with Scenario Planning


This post is by HBR.org from HBR.org

A conversation with consultant Peter Scoblic on scenario planning lessons from the U.S. Coast Guard.

How to Analyse a Competitive Landscape: A Framework for VCs and Founders


This post is by Omar Ismail, CFA from Startup Grind - Medium

‘Competition is for losers. If you want to create and capture lasting value, look to build a monopoly’

Every Business Can Be a Subscription Business


This post is by HBR.org from HBR.org

A conversation with consultant Robbie Kellman Baxter on how your company can adopt a subscription business model.

How to Market Your Product Without Spending Money


This post is by MRINMAY from Startup Grind - Medium

It’s definitely possible if you know the trick — this is where the magic happens.