Ted Barnett, a silicon valley veteran and longtime product & operations guy, wrote a post laying out his arguments for why product managers are becoming less necessary (tedbarnett.typepad.com)

1. It's always a good idea to reduce the length of the chain from customer-to-developer.

2. Specs are much lighter now. We are shipping Minimum Viable Products and learning from the market.

3. Higher-level development tools (e.g. Ruby on Rails) and powerful, cheap services (e.g. Amazon Web Services, UserTesting.com, Curebit, Gigya) free engineers to think on a higher level.
4. Because of these low-cost tools, consumer-grade services can be built for very little money.

let's discuss! (and ted, if you send your email address, i'll invite you too - hunterwalk at gmail)

I find Ted's blog post representative of the debate that was going on circa 2006 wrt product management. In fact, it was my very first meeting with Reid Hoffman in 2007 where we discussed this topic for 4 hours.

I think fundamentally there are three viewpoints that need to get represented - technology, customer/user, business/strategy. It's not impossible to have one person represent more than one of these, but it's not typically scalable. Startups get away with it due to exceptional talent, passion & focus.

My thoughts on product leadership in the modern consumer era are here:
blog.adamnash.com

I'm not sure I buy Ted's thesis that instilling eng/UX with "product management thinking" somehow alleviates the need for one person to own the product as it relates to the customer experience.

On a broader note, to answer Hunter's question: yeah, I think we do need PMs. I always loved Todd Jackson's "shit umbrella" theory of product management. Well-functioning product teams *love* having someone act as a buffer. It lets them build without distraction, it lets them achieve consensus on what it is they're building, and it helps them interpret the feedback coming from the users/data/market. techcrunch.com

I see merit in Ted's assertion to get the team closer to the user; I always wanted the Blogger engineers monitoring the help forums and replying directly so our users knew we were listening and were committed to fixing legitimate issues. But that's different than tasking them with prioritizing the next deliverables, managing the process, and shipping v.next.

Bad product management taints the lot for everyone. When it becomes about process instead of outcome, you should take out all of that product management and throw it in the dumpster.

Great product managers are behind the scenes, taking the founders' vision and working with the team to define it into concrete steps (and representing that vision when any questions or roadblocks come up along the way), and doing whatever it takes outside doing the design and writing the code (working with pr, marketing, ops, legal, etc) to make sure everything is ready for something new to ship. And making sure the engineers know what needs to be done and how to get there.

Maybe one thing Ted and I would agree on is that if you can't get closer to great product management, you should potentially get rid of it...

It all depends on how capable the person with the vision is, and how complex the vision is.

In many companies, the founder(s) own the vision. I personally believe the best companies are started by product people. As the vision/company grows, it's not a simple single line roadmap. This is where product managers come in, to own different parts of the lines of the vision.

My process is a little more unique, where product, design, and engineering are all jumbled together in my head and spit out in various piles and then reorganized time and time again. It's a process that works for me. For others, having a leader say "I want it to work like this" and then having someone who can interpret that vision and build a product that accomplishes the key goals is key.

The title of product manager might go away, but not the role. The role just might shift to the UX designer, or CEO, or "product engineer". But the need for someone who knows how all the pieces fit together will always be there.

One of the refrains I tend to repeat is that if product management didn't exist, we would spontaneously regenerate the role.

I agree with Josh 100% - I believe mediocre product management is common, and counter-productive. I also this is why most engineers & designers tend to have bad historical experiences with product managers.

The role of Product Manager as spec-writer or even product designer is pretty out-dated. Spec-writing has become a relic (on the product definition side) and product designers increasingly come from any number of functional roles.

To me the essence of Product Management is communication. In many ways it's internal marketing and internal PR. It's the role that reminds the team where we are and where we are going. A good PM finds a way to say the same thing 100 times a day without going insane. Product management is the grease that keeps things like "Engineering doesn't know what Design is doing?!" from happening.

And I very much agree with Adam's point that "if it didn't exist we would create it." When done best, it is a medium in which great work happens and hardly even noticed.

I think the world definitely still needs product management, but depending on the organization and product, it may not need Product Managers.

The holistic role of product management (note the lowercase) generally has consisted of two things:

1. Product Responsibility: Responsibility for the user/customer's experience with the product. Feature set, design, brand, how it's discovered, how it's understood, how users understand the product, support, etc.
2. Project Management: Responsibility to help everyone internally to understand what they're doing, when, and why.

Specs or PRD's are the traditional way that Product Managers communicated #1 to chart the course for #2. Seems outdated and overly bureaucratic now.

Many companies have broken apart #1 and #2 into other roles (Design, Developers, Project Managers, etc.). I think that's fine as long as it's the most efficient way to make sure product management (lowercase) is getting done.

[jack's message came in while i was composing mine; it's very similar]

For me, product management is both a skill and a process. It's one that can be shared/sharded among other functional roles, but I haven't seen a company grow past the size of ~20 people that hasn't needed some dedicated person, even if that's not necessarily the individual driving the "product vision."

I do think people in the product management role need to understand two things about their function:

1. product the skill is not solely the domain of product the role. holding to some notion that only product managers know what's best or have some unique capability is bs

2. product managers have the least structural authority of any role. that is to say, your success and impact is solely dependent upon creating value for the team. you are the caretaker of something bigger than you are.

We actually tried running without "PMs" at a previous social gaming startup of mine, largely because of my disdain for top-down style management, and as a way to keep the team engaged about the entire product.

We had engineering pods (2-5 people per pod), and then one "lead" per pod. But those leads were all engineers, and needed to spend 50% of their time still coding. We obviously cast people who were more customer centric engineers. The idea being they would spend more time simply "showing" the team what they meant instead of making 30 page powerpoint presentations.

The result was not a disaster. The best benefit was a real tight camaraderie, a sense of everyone being on the same page. But, the negatives were obvious. Everything that happened when you under-fund product management by 50%+ happened.

Product management isn't a role or a function, it's a set of skills. Those skills help remove obstacles and grease the wheels so that the functional experts can do their jobs best. Product management also balances the needs of users, the business and the team and makes the difficult tradeoffs needed to keep pressing ahead. In that way, Product Managers are very similar to CEOs. Very few would argue that a company doesn't need a CEO. Product managers are simply CEOs of their products. No organization should be without someone who has "product management skills" and works to make everyone else's lives easier.

Another point worth mentioning, building on Satya's point about making everyone else's lives easier. There are any number of gaps that exist; if you're a small team, you don't have enough people tasked to each functional role. If you're a large team, there are a dozen (or more) groups that need coordination to ensure that the entire effort stays on track.

The PM's job is to *at least* ensure the gaps are filled, but more often it's her job to actually fill in the gaps. If you distribute the responsibility across a team, it's too easy for everyone to point fingers and say, effectively, 'not my job'. The PM can never say it's not her job.

Being the CEO of the product means the buck stops with the PM. When that role is jointly owned by a bunch of people, it's impossible to assign ultimately responsibility, which makes it far more difficult to assign accountability.

Satya - completely agree, and that's basically the most succinct summation. An excellent product manager is keeping in mind the long term vision while driving short term results, has the customer intuition to get there, and has the authority/integrity to lead the team along the way - very much a mini CEO when it's done right. That is, of course, why there a so few good product managers.

For all the reasons a CEO function exists and we don't have a thread here on getting rid of the CEO (yet?) you should not short change product management.

There needs to be a central point, not of command/control necessarily, but of accountability.

The world needs product management, but that product management doesn’t necessarily need to be done by product managers. If a team is blessed with “full stack” engineers and designers then it’s certainly conceivable that they could meet in the middle and cover the responsibilities typically covered by product managers.

In reality they usually don't and we end up with dedicated product managers for one of the following reasons:
1. Interest - Designers would rather design. Coders would rather code. Even if Non-PM's could cover PM responsibilities successfully, they would rather not.
2. Ability - Non-PM team members are not truly "full stack" and there are high-leverage PM responsibilities that existing team members can't cover successfully.
3. Productivity - Other things being equal, if you're running a team you'll get higher productivity out of the team if people spend a higher percentage of their time on responsibilities in their wheelhouse.

Related thoughts at b.qr.ae

great discussion guys. I'm going to conclude the conversation over the weekend, so any last thoughts today pls :)

I love this quote from Ian: "Designers would rather design. Coders would rather code. Even if Non-PM's could cover PM responsibilities successfully, they would rather not."

It's an important point, because I've noticed that one of the effects great product managers have on their teams is that people feel "safe" focusing on what they love and are great at, rather than keeping one eye open for "other things" that aren't being taken care of. Let's everyone run forward as fast as possible.

Gets to the heart of a really important motivation factor across life - it's not a question of what you can do, it's a question of what you want to do and what lets you do your best work.

Organizational systems that scale have to, in the end, have to allow people to spend a majority of their time and effort on what they are actually passionate about. That's what differentiates organizational theory from what actually is scalable and maintainable over time operationally.

As my former CTO Craig Lancaster is fond of saying, "Specialization is for insects."

The earlier the startup the more likely a high percentage of team members have opinions about more than their specialty ("designers want to design").

We pursued a non-PM strategy as a method of empowerment, in reality it was unhelpful. The root here seems to be about empowering the rest of the team, which feels critical especially early - I'm curious what tactics others have used to address that issue of empowerment.