Learning lessons from Google and Bytedance - what if meetings need a reason to exist?

Author: zmzlois


Verify the problem, form a common agreement, follow up with execution.

Some background notes on recent changes with remote work:

After almost 3 years of remote working, most companies start calling the return-to-work policies. Ironically, it includes Zoom, the company provides the foundational tool for remote work. That spans out to Google, Salesforce and Amazon, just to name a few, and some of them are seeing fierce resistance to back-to-office. Work in the age of post-pandemic has been splitted in four ways: entirely in-person, fully remote, flexible hybrid and structured hybrid. Zoom picked structured hybrid with a more flexible approach: if you live within 50 miles of our office, you can come in the office for two days a week, otherwise you don’t have to. Kastle reported their access control system data over the last 3 years: till this date, most offices in major cities in the US are still not fully occupied.

What if meetings had to have reasons to exist?

Continue with my previous rant about excessive meetings in consultancy, I found that a completely different breed exists: they’d talk about how many meetings they have to drive organisational growth and how they manage a “high growth product”. 🤯

In 2020, Google banned meetings for a week and accidentally reinvented remote work, and what they found was:

Cut costs: Hire people from anywhere, as well as reduce wasted hours in meetings. 

Increase productivity: Less time in meetings means more time to be productive. 

Retain staff: Increasing workplace satisfaction will help reduce employee turnover.

Looks familiar? Some parts of them really resonate with what I wrote from the last article: the harm of low-effective meetings. As I started to get involved in Bytedance’ infra tools modern.js and Rspack, I learnt ByteDance's practice in how meetings were conducted and how one of the biggest companies in the world carried out meetings in order to keep the highly iterative start-up culture.

The notion being meetings weren't required unless they were absolutely necessary. If we are to abstract three parts out of it:


The ethos of meetings was "Form a common agreement by focusing on problems, and preparing for execution." While executing, people will find more problems, have more meetings and execute them. This action loop will help an organisation achieve its goal step by step.

Hence a highly effective meeting has three criteria:

  1. The problems are real problems

  2. Meeting participants need to construct a common agreement

  3. Problems are solved via common agreement

1. The problems are real problems

The most difficult one to achieve on top of three of them is "The problems are real problems". If we discuss a problem that doesn't deserve to be discussed, then all the investment and effort before and during the meeting are wasted. The only value-add is "Discovering whether this problem is worth discussing." before the meeting happens.

The results after a meeting are much more important than the process.

Some common meeting topics like "How do we achieve 30% growth this year?", "Why did this project fail?" or "Why did this parts of the code fail to do what's supposed to do?" imply a proposition that these problems are worth discussing. But the critical thought process of "Is 30% growth an achievable goal?", "Did this project actually count as a failure?" or "Did this part of the code work at all?" needed to be done before the meeting and given the conclusion that "Yes. They are real problems. They needed to be discussed."

You might have experienced this: You spent an hour discussing a topic in a meeting and found out that this problem doesn't exist, then why was it discussed? Because your boss said so? He said this was a problem, so everyone joined the meeting to discuss it? You were treating the meeting as if it was about the process but not the result. The actual effective way before the meeting starts might be: If the leader said this is a problem, someone needs to think, challenge everyone: Does this problem deserve a meeting?

Someone told me this approach would be bad for organisation’s culture as it discourage people to speak up for problems. Think about it in this way: if someone’s goal isn’t align with the organisation enough to act upon for the organisation, do you need them? (This is some harsh thoughts, might not apply)

2. Construct a common agreement

Constructing a common agreement is pretty much relying on the meeting host’s control and management skill — pretty much an art of its own.

At the end of the day, participants are human. They all come with their own opinions and emotions. It is indeed an art to handle argument, conflict, silence and unexpected scenario in limited time and even reach an agreement.

In my experience, the best way to reach agreement is repetition, recognition and confirmation.

Repeat the conclusion you have received, confirm with everyone, and make sure everyone understand that you particularly care about everyone reaches a common agreement in a certain case. You more likely to reach the result when you show everyone that you care, a lot.

To deal with the risk of divergence: the amount of time spent in the meeting to discuss about the “set topic” would pretty much determine whether you can reach a common agreement or not.

We often derail in a meeting, be it discussing too much on unnecessary problem, or discuss something about slightly relevant to the meeting theme but not quite. Meeting host needs to make sure the time spent in the meeting is used for the meeting theme but nothing irrelevant—keeping the string tight.

3. Agreement can solve problem…yes?

When you look at this condition you might be thinking: isn’t this something being determined in the meeting? Because the common agreements are laid out, same as the solution to the problem.

But in reality, “Common agreement can solve the problem” doesn’t determine on “can”, but “solve”.

Meetings don’t solve any problems, it is only the execution part solves the problem, alas someone push the execution after the meeting. And if I take a wild guess—this is why most of meetings fell short.

A lot of organisations make a todo list during the meeting, convert them into task and assign them to assignee. However, during an execution, we can easily find that once you involve more people in the meeting, the execution might diminish simply because there are more people to chase.

That’s the other part of organisational management art. All I want to say is whoever cares the most, they should follow up the most. When the policy can’t follow and make things work, you can only work harder to get the result. You will definitely feel tired and aggrieved, but this is your professionalism so you have to do so.

Take away: similar to decrease the number of low-effective meetings, to make remote-work works in your organisation, these are likely the things you will have to do for your organisation, whether you are running a start up or internal innovation in a large enterprise XD.

  1. Verify the problem (sounds like start up huh?)

  2. Form a common agreement

  3. Follow up with execution

It might be painful, but hey, it’s the work we love. And we make things work.