A few months ago, Joel discovered a fascinating book that he shared with all of us on the team. The title of the book was “The Decision Maker” and it proposed a radically different idea to how companies are run.

In short, it put traditional management on its head. Instead of higher-ups making decisions, often far removed from the real problems that team members face, you give the decision making power to those that are closest to the problem.

Implementing this has been an incredible transformation of Buffer as a company. Here is how we have implemented this across our whole team a few weeks ago:

Our new setup at Buffer

As a 24-person team, we have very recently split out into various product teams. Each of these teams is completely autonomous, makes their own decisions, and sets their own goals and schedule.

We have 6 product areas at Buffer currently:

The way we build product is that each of these teams has 4 stages that need to be completed for a successful feature release or experiment:

  • Growth = What are our intriguing numbers we can pursue to grow Buffer?
  • Research = What are our customers biggest problems (in the area of our intriguing numbers)?
  • Product = How can we solve this problem in the most efficient and useful way?
  • Tech = How can we implement this problem efficiently and successfully for our customers?

Each of these 4 areas needs to make substantial decisions for a new feature to be useful. We have 1 person per area that is completely independent in answering the above question for them. Here is an example of a team that works on the Buffer dashboard:

decision maker process


An example scenario of how we might work together

In an example scenario, Joel might discover an intriguing number by analyzing our data. On the dashboard, we found that some of the most common actions are changing your posting schedule, however we also know that the schedule tab causes some of the most confusion with the product.

Having identified that number, Joel passes it along to Rodolphe, who then does customers development with customers on the schedule tab. How are we causing them confusion? What problems do they have when it comes to scheduling their social media posts?

Rodolphe then presents the problem he discovered as clearly as possible to the product person with a validated hypothesis. Example “If we improve the schedule tab, then more customers will stay actively using Buffer, due to our research showing that the current confusion of how timing works causes drop-off.”

Rodolphe would also loop in “Customer”, in this case Carolyn, to see if around the intriguing number we might have already learned something from support requests so far. This is especially helpful to cut down research time and also know about the emotional state of a customer about a problem, which can help with prioritizing it better.

Joel sketches and specs an iteration of how our social scheduling tab might work. Once he’s done, he might pass it back to Rodolphe to clarify with some customers whether we are actually solving their problems.

Once we’re positive on that, we’ll pass the finished spec to the tech decision maker. They will then go ahead and decide how to best implement the new solution and roll it out to our customers.

There is no one person that “runs” this process; it’s a combined effort of decisions for each and every area.

A natural question might be, how can you make sure that the decision each person makes is a good one? A crucial element of the decision maker concept is the advice process. Here is how it works.

The most important part for our change: The advice process

Without managers, how can we ensure we’re still making sound decisions? Quite simply, by asking for advice. The advice process is one of the hardest things to get right and something we place special importance on.

In general, the larger a decision, the more people you might ask for advice. If you want to build a whole new product, you might want to speak with every single person of the company and get their advice. For smaller decision, it’s ideal to speak with fewer people and for some, possibly with no one.


The way we’re thinking about advice is to avoid a top-down approach. Someone from the growth team might get advice from someone else doing growth on a different team, on top of getting advice from leaders in the company. On “research,” for example, getting advice from the happiness team is crucial, as well as from the marketing team, since Buffer is a marketing product and our marketing team is likely the best user of our own product.

The advice process is so important that Bakke went as far as saying:

“the worst trouble you’ll get into won’t be for making a mistake. It’ll be for not asking for advice. Like I said, asking for the advice will help keep you out of trouble.”

So, in our above example, if the person owning “Growth” decides to pass along an intriguing number to “Research”, which is effectively a big decision, as it’ll affect the work of the rest of the team, she is required to get advice on her decision. This can be a short conversation with another person owning “Growth” in a different area, or an otherwise experienced leader that might have a good intuition on a decision.

Sidenote for founders and executives: One of our investors, Dharmesh Shah pointed out that this gets particularly hard as a founder: You need to be able to give people advice, with the real option that they might not take it and go a different route—and you have to be fine with it. All things considered, I believe this might have been one of the hardest things for me to do so far. Whereas before your word as a leader or manager might have been the definite way to go, it is now only one piece of the decision maker framework. While my ego gets in the way at times, it always makes me feel incredibly hopeful to being able to scale this company very far, whenever this happens.

Some current limitations and things we are aiming to improve

As you will see from the above formation diagram, Joel is filling 2 roles right now for the Dashboard team, serving as both the owner of “Growth” and “Product.” That is certainly a challenge and can lead to stretching yourself too thin, and is something we’re actively trying to improve (we’re hiring btw, check out our open positions here! :)

We also have some people that are part of multiple teams, which makes this even harder, yet it’s of course all part of the plan when you’re a startup.

Another thing that I personally struggle with a lot is that it’s often not that obvious when you’re actually making a big decision. Being self-aware and at times going a tiny bit slower than you’d like can substantially improve the decision making of everyone in your company. I expect that this might take a bit of time as we are all getting fully familiar with the new concept.

Scaling a company where everyone enjoys their work

Although we are convinced that the decision maker framework is the most efficient and productive way to run a company, this isn’t necessarily the root of this change. At the very heart lies the idea of giving people control. Dennis Bakke writes that:

“Only if we’re in control, we can have fun doing what we do.”

That was also why Joel and I started Buffer in the first place – to have complete freedom for how we work and what we decided to do with our lives. A few years into working on Buffer, although we personally have achieved that vision, we haven’t extended this to the rest of the team. The decision maker framework is an opportunity to give the whole company the same amount of freedom that we feel.

I believe it’s also allowing us to make better decisions that will create a better product for our nearly 2 million customers. (Ready to save time with social media while driving more traffic and engagement? Try Buffer for free!)

I remember a Tweet from Danielle Morill once where she asked, “will there ever be a 1,000 person company who people really and genuinely enjoy working for more than anything else?” With the limited knowledge that I have, I believe that the decision maker framework might have something to do with the answer to that question.

How do decisions get made where you are? I would love your thoughts on our slightly crazy new way of working as a company and what you’ve found to work well within your workplace!

Interested in joining us on our journey? We’re hiring!

Free up your day with our Social Media Tools

Buffer can save you up to an hour a day and grow your traffic too.

Learn More
Written by Leo Widrich

Co-founder and COO at Buffer. I enjoy writing about Buffer’s lessons learnt, social media tips and updates to Buffer. For some more personal posts, check out leostartsup.

  • http://peter-rocks.it/ Peter Piskun

    Fantastic insights Leo (and a wonderful approach, I should add the book to my wishlist) – just on one thing I would like to have some more details: How are you managing your projects / advice processes – how is the workflow and which tools do you use, f.e. Trello Boards or something else?

    • LeoWid

      Hey Peter, ah yes, great point. Yup, Trello and Hackpad mostly!

      • http://www.sahilparikh.com/ Sahil Parikh

        How do you’ll use Hackpad? We use Brightpod (created by us) but an external notes app would do wonders. Have you integrated Hackpad with Trello?

  • http://www.kibin.com/ Travis Biziorek

    It seems like the real benefit of this composition is to have small teams where everyone is able to make decisions throughout the process. As a result, you move faster. It seems like this decision making process is the root to employee happiness (I’d agree).

    But I’d be curious what happens as these teams begin to inevitably grow. What happens when you need 5 engineers on one particular team, or 10? From what I understand, there’d still be just one “tech decision maker”. How do the other engineers get the opportunity to make decisions.

    TL;DR it seems like this would work really well with small, focused teams, but what happens when these team grow larger and larger?

  • hdc77494

    What’s interesting to me is that your essay doesn’t mention cost/benefit analysis nor allocation of resources, i.e., with limited resources, what’s your decision process to determine which projects get fully funded and which don’t. Once you move past those decisions, to me the best approach is financial transparency to team members and granting said team members the authority to determine how to deliver results combined with direct accountability for delivering them.

    I strongly recommend the text “Execution, the discipline of getting things done” by Larry Bossidy. One of the greatest strengths and greatest dangers of small teams (24), is that it’s fairly easy to put together a group of like minded people, and like minded people can miss seismic shifts in the market because they all see the market the same way. Good luck with your experiment.

    • LeoWid

      Great point, I think that’ll be a great idea to dig in and see if we can exactly measure how much we win/lose. In general, knowing whether people have more fun, is a good measure and seems to trickle down to productivity and all other results naturally.

      Awesome recommendation, will take a look!

  • João

    Great article Leo! Wondering where the Costumer Support and all the Happiness heroes fit in this new setup?! Onboarding? They are providing feedback from users to multiple product areas? Or you have different heroes assigned to different areas?

    • Hugo Bessa

      Wondering that too. They might be split into many teams!

      • João

        Not sure how efficient that would be. Since Buffer likes the “all hands on deck” approach and low response times. And to achieve that on twitter and at other costumer support platforms they need all the happiness team available to everything costumers could need. (Android, iOS, Chrome, business)

    • LeoWid

      Hey João,

      Great to see you stopping by here and that’s a great question!

      I just mentioned to Jason above, that I should have added a paragraph about how the Happiness and Content teams fit into this. We have the same approach in these two teams, there are decision makers for various areas (blog, social, livechat, email support, phone support, etc), that aren’t managers.

      Let me know if that helps! Excited to share more about our progress here!

      • João

        Thanks for the insights Leo!
        Maybe we could have an article from someone at Happiness Team as a followup from this one!

  • Thomas Gegenhuber

    Interesting Article! Check out Gary Hamel’s take on organizations without managers: “First, Let’s Fire All the Managers”: http://tinyurl.com/km9djgn

    • LeoWid


  • http://www.jasonfpeck.com JasonPeck

    Fascinated by this. I’d love to hear more in general as it relates to some of the other questions folks have. Also, I wonder how you determine where content creation (and all the great people who write on your blog) fits into this?

    • LeoWid

      Hey Jason, thanks for stopping by and that’s a great question!

      I should have added a paragraph about how the Happiness and Content teams fit into this. We have the same approach in these two teams, there are decision makers for various areas (blog, social, livechat, email support, phone support, etc), that aren’t managers.

      Let me know if that helps! Excited to share more about our progress here!

      • http://www.jasonfpeck.com JasonPeck

        Definitely. Anything else you want to share on this whole topic would be amazing. So. much. awesomeness.

      • http://www.GoCanvas.com James Quigley

        We are actively researching this and potentially rolling out a similar structure. We have stayed “Flat” for years now and close to 50 people and sustaining Hypergrowth year over year. Fascinated with your approach and it may just fit for us as well.

      • http://www.jasonfpeck.com JasonPeck

        Would love to hear how you handle QA/testing in this framework. It’s something we are trying to tackle so are curious about how other folks are doing it. This would make an awesome blog post :)

  • http://pimp-my-conversion-rate.com pHx

    About the theoretical question:

    “will there ever be a 1,000 person company who people really and genuinely enjoy working for more than anything else?”

    I believe that this is possible only if you can hire 1000+ people with very similar characters and good amount of skill. Even 2-3 slackers can ruin the whole idea.

    I know what is to grow from 15 to 150+, how people change, how many problems can occur and so on. The model that you are implementing is great, but I am very sceptical about scaling it above 50 people.

    Good luck ;)

  • https://twitter.com/PascalClarysse Pascal Clarysse

    “While my ego gets in the way at times, it always makes me feel incredibly hopeful to being able to scale this company very far, whenever this happens.”

    This! This is the key point.

    – I opened my first start-up when I was 20 years old, around 1998. It was a small start-up and I pretty much was “in charge”. I would listen to advice but my decisions would prevail. It went well for four years. Then my hubris drove us into a dead-end.

    – age 24, I joined my biggest supplier. It was a medium-sized start-up with much more room for fast-pace growth than the one I had before. I was in charge of Marketing, but someone else was owning operations and yet another leader was in charge of development. Like you at Buffer, we set out to run the company as equals, as a triumvirate, in consensus. The hardest part during the first year (especially for me, I would say in retrospective) has been to swallow our ego. Once we did, we grew and thrived for the following years. And it was an awesome ride. And it made us better men too. And better leaders, not only for this venture but for the rest of our careers.

    So! In my humble opinion, a boss transforms into a leader, like a roach turns into a butterfly, when and only when he checks out his ego at the door. You’re on the right path! Keep on it :-)

  • Culture Amp

    At Culture Amp – one of our values is giving decision makers the power to make decisions. Anyone can own something and take action.

    However, our model is (1) if you’re taking a decision, it’s all about transparency to the team and solicit as much feedback as possible, (2) if you’ve given feedback you then back the decision maker, even if it’s against your advice and (3) we always have someone “verify”.

    3 is pretty important. The aim here isn’t to contradict or challenge the decision, but to throughly validate the inputs and the result. Someone proofs the blog, someone does a code review, takes the design past the customer or validates the numbers (and so on). Then final result and managing the outcome is then co-owned by the maker and the verifier.

  • sam_rye

    Nice work on this Leo!

    We’ve been experimenting with decentralised, self organised organisation for awhile now over at Enspiral in NZ. We’re up to roughly 150 people across locations, with a variety working on venture teams, open source projects, and services businesses.

    One of the first things we bumped up against was decision making, which spurred the development of Loomio – a decision making tool which is now being used by tens of thousands around the world – from businesses to social activism groups (its available in something like 25 languages) to governments.

    Check them out – it could be handy for you: http://loomio.org

  • https://twitter.com/costayolo Costa

    Spotify uses a similar paradigm; they call their teams “Squads” that are very independant: https://labs.spotify.com/2014/03/27/spotify-engineering-culture-part-1/

    • https://pawzii.com Alec Matias

      Thanks for sharing this, very cool.

      • LeoWid

        Love what Spotify is doing here, taking a lot of inspiration from them!

  • Galaxion Vulderdrip

    In order for anyone to be truly happy, they need to part of a company! Or at least feel like they are!

    I remember a Tweet from Danielle Morill once where she asked, “will there ever be a 1,000 person company who people really and genuinely enjoy working for more than anything else?” With the limited knowledge that I have, I believe that the decision maker framework might have something to do with the answer to that question.

    To answer this question?

    If you have 1000 toilet cleaners, the one on top is doing the crapping! So no, they won’t be happy!

    If however the 1000 toilet workers can make the person on top go when they want, then they have control of any given situation!

  • https://pawzii.com Alec Matias

    How do you handle features or tasks that affect multiple teams? For example, let’s say the iOS team decides a speech-to-text feature is what they’ll build next. How does the Android and Extensions team prioritize the work to add the feature to their platforms?

    As always, huge thanks for sharing your internal insights.

  • http://www.dangoldin.com Dan Goldin

    This is very similar to another book – Reinventing Organizations – http://www.amazon.com/Reinventing-Organizations-Frederic-Laloux/dp/2960133501

    The idea there is also to push decisions towards the edges where the people who are best able to solve them make the relevant decisions. But in order to get the process working many organizations end up adopting this advisory model as well. Worth taking a look at if you found The Decision Maker an interesting read.

  • ceceliajernegan

    I am a retired baby boomer. I have worked for several large corporations. This story makes me once again believe in America and our values. Keep doing a great job at Buffer. I just found you today on line and LOVE your concept. If I can help more I am based in Arizona. I have a little booklet on Amazon with best practice tips working in a virtual office. (cliff note type) Keep up the great work!

  • http://emidlife.com/ David Pethick

    This is really interesting. It’s certainly better than the default hierarchy that exists in most teams (he who is most senior makes the decision).

    I’m not sure if it will scale. By default, when you have ten engineers but only one decision maker in a group, isn’t that person the manager? Obviously you can keep splitting the decision pie and perhaps that is what is really powerful in this management framework – it’s empowerment in a box!

    Can’t wait to read more as the experiment continues. Thanks for sharing Leo!

  • Yiannis Mavraganis

    Congrats for creating a working environment where any employee feels empowered. This is the single most important thing in order to find joy at work. As Max De Pree said: “Not having the chance to make decisions within the organization in which one works is a great tragedy.” Have a look at the following article describing Boyd’s OODA loop:
    Effective organizations should decide and act as a whole like a fighter pilot.

    At AfterSearch, a very early stage startup (@AfterSearch), inspired by agile and lean principles, we are building a tool to facilitate a team’s decision making process.

  • Rob Lennon

    I view one of my chief responsibilities as a manager to grow and nurture my reports and ensure their success and engagement with their work. While Buffer does seem like a stand-out workplace where everyone is engaged, how do you account for the things a manager does that don’t relate to making decisions, like making sure people are happy on a track to develop their skills and career?

  • http://www.jasonscottmontoya.com/ Jason Montoya

    Thanks for sharing. Our company did something similar. Where you used the role title “decision maker” & “primary assister” we used “authority” and “driver”. By adding the clarity, we discovered many problems when this type of visibility does not exist. It’s affirming to hear about you guys doing this :-)

  • udidahan

    We’ve been moving in this direction in our company as well.

    One of the processes we’ve introduced is Innovation, which allows for certain ideas to percolate and develop with limited investment until such time as they’ve collected enough evidence to justify bringing them into the core of the company.

    Although we’re also moving away from “managers”, that role has been broken into 2 parts: Expert and Mentor. Anyone in the company is capable of developing into an expert in some area as well as being a mentor. One of the goals of a mentor is to help their mentees grow.

    When a taskforce is making a decision on a certain topic, an expert on that topic will probably already be a member of that taskforce.

    In general, I’d say that changing our organizational structure has driven us to be even more specific about what we want to achieve and how we go about that, all leading to good results.

    In any case, it’s great reading about your journey. Maybe one of these days I’ll start writing about ours :)

  • Piotr

    Hi Leo, we’re thinking of implementing similar model in our startup. I was wondering if you could share how is that working out for you guys after over half year since you implemented it?

  • Amanda Madorno

    Could you please develop some intelligent software responses for what people want to post? I get way too many suggestions that don’t fit me or my business. For me, the one thing Buffer is supposed to do is make my life easier that way. It would be great if you could have intelligent something-or-other that can determine, ‘oh, she likes these kinds of articles.etc,, so we’ll send her more of that.’

  • Andy Edwards

    I love how transparent you are with all your strategies and company changes.