Want to Know the Difference Between a CTO and a VP Engineering?

by Mark Suster on April 19, 2010

I recently did a post for startups on understanding sales people.

A few people have asked me to try and define the perfect startup organization chart.  I don’t believe that one exists.  Every team configuration is different.  But I do have more insight into understanding your startup team.  This time I thought I’d try and address engineering talent.  Often I’m asked by startup CEO’s about how to best build an engineering team.  I have much experience in this domain.

Because more technology people probably read startup blogs I’m guessing this post will come under more scrutiny.  The terms “CTO” and “VP Engineering” have such stigmas associated with what they are that I’m sure some people will feel uncomfortable with the definitions I’ve put forward.  Still, I believe I’m offering an accurate representation of the ideal configuration of the main technology leaders.

This post is designed mostly for non-technical founders.  I hope many will read this and have an answer for the question, “what’s the different between a CTO and a VP of Engineering?”

Let’s start with the basics.  What makes a great tech team?

1. The CTO / Lead Architect - If you want to build a great technology company, you’ll need a “rockstar” engineering lead.  Every great tech startup needs one.  Whenever I meet a team that had a consulting firm (even a great one) build their product it’s an immediate “pass” from me.  If you don’t have somebodyinside your organization who is setting the technology direction then I’m convinced you’ll never head for greatness.  I know this will fall like a lead balloon to the many people who believe it is possible to have a [insert: startup incubator or technology accelerator or technology consultant or outsource firm] build your technology.  I don’t believe it.  Either your core is innately technical or it’s not.  It’s what makes Google Google and Facebook Facebook.

So I believe that every great technology startup has the technology visionary inside the company.  This is the person who lays the foundation of what should be built.  They’re up to date on the latest platform decisions whether it’s understanding SpringHibernate and Lucene.  Or whether it’s a big data set problem and they’re familiar with Cassandra or Hadoop.  Or whether it’s a choice between using MySQL vs. Postgres.  They’ll have a view on whether Ruby on Rails is worth the hassle.  Some CTO’s swear that it is a huge improvement in development timeframes and doesn’t cause performance issues.  Others think you should never build anything highly scalable on Ruby.  I’ve heard both arguments from CTO’s.

Trying to work without this person is like wanting to build a world class sky scraper but not having a great lead architect and civil engineer.  They provide the vision for your infrastructure.

But the problem that many inexperienced startup CEO’s make is confusing these people for the people who lead the technology team.  Most often they are not.  Your deepest thinkers on technology architecture are seldom good team leaders.  They often aren’t great at planning development work.  The best technologists often aren’t amazing people managers.  Sometimes they are introverts.

In fact, it my experience the best technologists are akin to artists.  They’re highly creative.  They’re sometimes moody.  They work on their own schedule and are often hard to manage.  They may work strange hours such as 2am – noon.  They don’t love documentation.  They often don’t love testing.  Of course I’m generalizing.  But barely.  The characteristics are so prevalent.  These people are your purists.  Your Howard Roarks.

So what is the difference between a “chief architect” and a “CTO”?  Simple. Experience.  A “chief architect” is a young version of a “CTO.”  It’s your hedge.  A chief architect still has a lofty title but they still need to prove themselves in order to become CTO’s.  You still have some leeway to hire above them if need be.

The best CTO’s / Chief Architects are purists.  They care about the quality of what is build more than they care about end customers.  They should be setting the standards for how code is developed.  Let them be perfectionists – this will serve you well.

2. VP Engineering – In my view it is important to distinguish the difference between the CTO and the “VP Engineering.”  Because these titles are so often used I’m sure that some people will have hardened views about what they mean that are different than mine.  But for non-technical founders let me offer you a definition that you can use when you build a team.  The VP of Engineering is the person who still has great technical chops but prefers not to be a coding monkey (that term is meant in the most endearing of ways).

The VP Engineering aspires to manage teams.  They feel comfortable with C++ but also have a black-belt in Excel.  They are sticklers about managing unit testssystem tests and regression tests. In fact, they are passionate about automating testing overall.  They know how to estimate work units, how to manage the agile development process and how to get the most out of their teams.  VP’s of Engineering are essential to making sure the trains run on time.  The VP of Engineering is also your primary interface to your head of product management and often the VP of Engineering is somebody you would drag in front of clients to win big deals.

And first and foremost a VP of Engineering is a people manager.  They still have the respect of their team because they’re technical by training.  But they’re that rare breed that also understand the human element.  They know how to motivate their people.  They understand the different character types and which prefer carrot vs. stick.  They know how to get people to hit deadlines.  They know when it’s OK to push hard for the team to hit a deadline even if it means yet another all-nighter or weekend.  And they know when to tell YOU to get stuffed because the team has reached maximum stress / effort.  A great VP of Engineering manages as well up as they do down.

So if I were a pure startup with 5 people I’d want a Chief Architect.  As the company and therefore the team size grew I’d want a VP of Engineering.  For me the inflection point is usually when you have 5+ developers.  CTO’s max out at about 3.  Remember, “management” is often a hassle for CTO’s, not a sign that you respect them by giving them people who report to them.

3. Program Manager – This title almost sounds like a consultant’s job.  It is not somebody that I would have on a small startup team.  However, it is one of the more critical roles as you scale your organization.  As you head into the phase where you’ve had real customers paying real money for a period of time you’ll have a whole new set of issues.  Examples:

- every time you release new features you need to update your technical documentation

- you also need to update your marketing documents including your website

- somebody needs to be sure that customer service is alerted to the new features and are trained in how to handle these functions with customers

- new features need to be rolled into PR strategies and competitor analyses

- new features need to go into the sale people’s slides so that they know the latest and greatest about how to differentiate from the competition.

Many startups have never faced these challenges because they haven’t hit scale.  Trust me, as you grow these issues become “gating items” to winning large customers and keeping them happy.  My company never became Google but at $14 million in recurring revenue and $36 million in backlog revenue we certainly had enough big clients to necessitate a very solid program management function.

In summary.  If you’re starting a company make sure you have your chief architect.  If you’ve outsourced this to a firm that has guaranteed that they know how to launch you more quickly I’d tell you that’s like trying to launch a movie but outsourcing the script to a focus group.  People will tell you it will work, it won’t.  Don’t take the easy road.  I’d rather delay by 3 months and have the right DNA inside the gate.  As you have a 5-10 person startup you don’t need a lot of technology process management.  As the CEO you can personally help manage deadlines and the Agile process.  So no need immediately for the VP of Engineering.

But as you company grows to 10-20 people you’ll want to consider adding “technology management” skills, which means a VP of Engineering.  In my view each of the CTO and VP Engineering should report directly to you and you should remain very hands on vs. “trusting them to make the right decisions.”  As you cross the $5 million mark and have lots of customers don’t forget to add the program management function.  Coordinating new product releases into the entire fabric of your company will become vital.

OK.  Post done.  Engineering teams – feel free to attack! (or add your 2 cents)

Share and Enjoy:
  • Twitter
  • Facebook
  • LinkedIn
  • FriendFeed
  • del.icio.us
  • Google Bookmarks
  • Digg
  • StumbleUpon
  • HackerNews
  • Suggest to Techmeme via Twitter
  • email
  • Print
  • Mark Long
    So the CEOs have asked you how to build an engineering team. I'd be willing to wager, twice as many have said, "I think something's wrong with my tech team, I just don't know what...or how to fix it". Having sold our company last year and gone through the process of looking at Tech Management job opportunities (from startups to larger growth companies), I was shocked at how messed up these tech teams were, and far they let them go before making a change. So perhaps a topic for another day is, what should your growth company CEO be doing to make sure his tech team isn't going off the rails? Interestingly, I'm a few months into my new gig and there are quite a few things that need fixed and when I talk to the CEO about them he says...."you know, that's what the last 2 guys told me." I'd suggest a future blog/article on how do you know if your tech team is any good? is it metrics? we all know that's iffy, especially early. Is it an advisory group of other Archs/CTO/VP Eng types? On one hand I'm not a fan of consultant types, on the other hand there are many CTO colleagues that I think would do a great job (and I have called on them from time to time).
  • Great piece.

    How would/should the dynamics change if the purpose is to build a great technology based CONTENT company, where the technology is important to support the delivery of the content, but not the core of the business offering? Should you only focus on bringing on a "VP of Engineering" since the skill set of a true Howard Roarkie "CTO" would be wasted unless the technological aspects are in focus?
  • Harish
    Thanks for wonderful blog with some content to clear the air.

    Few questions to you;

    What is your take on startups having multiple branches. Is it advisable?.

    Secondly if you could share your experience and learning's in having worked with startups which had multiple branches ( e.g. one in US and other in Bangalore) it would be great.

    -thanks
    Harish

  • Thanks for writing this Mark. Supremely helpful for bizheads like myself
  • riccardo_larosa
    Mark,

    great article. Not to murk the waters but where does a CIO fit? I find that CIOs mostly exist in large, well established companies that have to deal with legacy back end systems and are not necessarily technical (at least not anymore).
  • jeffsolomon
    On the f'n money. This is how it went down for Leads360. The sooner I accepted our chief architect wasn't going to manage anyone the faster we would grow and the less stress i would be.
  • One item that's missing is the relationship between VP Engineering and CTO. The two need to see eye to eye on technology, vision, and even management style. It's a recipe for disaster if the two most senior technology executives don't get along well.

    On the topic of CTO has introverted tech geek.... That might be ok for a small startup, but at some point I think a CTO needs to grow into the outbound technology face of the company. A VP Engineering just doesn't have the cycles to be on the road meeting with customers, investors, giving talks at conferences, publishing, etc as often as would be required. VP Engineering is all about execution. You can't execute if you're not there and focused on engineering. Of course, everything in moderation. VP Engineering does need to get out of the building too, but not as much as the CTO.
  • Awesome stuff. Really digging the hiring advice. Maybe you will grace us with your advice on where to find - how to attract A-Team Rock Stars. As a Director of Engineering (punching above my weight class) I think its critical to be able to attract the Rock Stars (I hear Ninja is in) . VPE should be able to bring the talent.
  • jonathanchamberlin
    Mark,

    Question regarding "remaining very hands on vs. trusting them to make the right decisions.”

    How do you define remaining very hands on? I know as a CEO you need to wear many hats and you obviously need to be aware of what's happening in your startup, but isn't the CTO and VP of Engineering the expert? And as the expert they should be able to make technical decisions better than yourself? Otherwise, why were they hired?

    I've always been of the persuasion to hire the best talent as possible and let them work their magic. There's no sense in having a CEO dictate something where he has no expertise. Limit as many barriers as possible.
  • brucekenny
    Mark, great post - you have done many a great assist by taking the time to lay it out so well. As with most things the key is timing - VP Eng is a growth requirement once you add products and customers to the mix and the CTO is critical to kick start the growth. As a company grows the needs and emphasis in each role changes. I also appreciate you calling that there is no one right model - it depends on the people, the technology, the opportunity etc. Great post. Thanks.

  • You forgot to list human resources in the lower left. :)
  • Great synopsis, and I would add:
    1. CTO is visionary, outward-looking guy. Tends to be the one to speak at confs and represent the firm's tech dreams.
    2. VP Eng is tech team manager, more inward-looking.
    3. I thought "Program Manager" was a remnant of MS Windows :^)

    Re outsourcing, I agree with *some* of the comments made so far, but would add that structures and offerings vary widely among shops.

    Our group does outsourced XP/ Agile development, mostly in Ruby on Rails. The work is done largely without spec, and things usually are not spelled out black & white. It worked for Twitter.com for a period of time. (click thru to find me, be sure to mention the word "plug" :^)
  • garydpdx
    I'm wondering about point #1 ... I recently read a paper called "The Role of the CTO: Four Models for Success" by Tom Berray of Cabot Consultants (McLean, VA) where CTO can be seen as #3) Technology Visionary and Operations Manager and #4) External-facing Technologist. Another paper labeled those as 'inward looking' and 'outward looking', and I guess that different pursuits require one or the other. In an earlier posting regarding Chief Scientists, that person would have an 'inward looking' orientation for the product while the CTO is 'outward looking' balancing their technology direction versus customers' needs.
  • David
    5+ developers need a CEO, CTO, VP, and Program Manager? No. No company should be that top-heavy. Companies (of all sizes) should have lean and clearly defined management chains. A startup with five developers needs one person in charge.
  • garydpdx
    On the other hand, in my field (electronic design automation - EDA), those 5-6 people usually have different responsibilities so, title or not, you have a CEO, a VP Sales, lead software coder as VP Engineering, lead applications engineer as Director of Applications Engineering, etc. Maybe second junior persons for software coding and/or applications engineering. As top-heavy as that appears, in this field, people expect to remain at the top of the food chain as the company grows - not have additional bosses appear above them. Most team members have done time with one or more (or all) of the Big 3 corporate EDA firms, where advancement is often difficult (EDA start-up founders who are 10-15 year veterans are not unusual)
  • CTO is just a nice title. Your definition of CTO doesn't stick. You may have different team with different needs, and the CTO will not be able to help in the way you defined. It's much better to have engineering leads who know what they are doing.

    Does Google have a CTO? No. Is there anybody at Google that matches your definition? No.
  • http://www.google.com/intl/en/corporate/execs.html#sergey

    Doesn't Sergey Brin serve the role of CTO at Google? His title is President-Technology. :)
  • Mark,

    I have been following your blog for the past couple of months and in some ways using it as a guide and evaluate if I am doing things right with my first startup. Most of the times I have reflected on your posts and felt, I have a long way to go (especially the entrepreneur DNA series). This post makes me feel ... maybe I am doing some things right and have a chance :-). My engineering team is structured the same way- have 5 developers including a cofounder / lead architect while I am the the CEO/cofounder/technical product manager that also manages deadlines and the agile scrum process. I had the same reasoning/logic as you outlined in your post for my decisions.

    Anyway thank you for all your posts and loved the understanding sales people (something I will need to start thinking about soon). Also love your This Week in VC with Jason Calacanis.
  • Mark,

    I basically agree with you here in the sense that there are two completely different (critical, depending on the size/maturity of your startup) roles you describe that people tend to confuse and blend, program manager notwithstanding.

    One question though. It is my understanding that the CTO is the most senior member of the technology team. Doesn't it stand to reason it could be the Chief Architect OR the VP of Engineering? If I could draw a simple diagram, it would basically be the CTO at the top, and depending on which role the CTO is (VP of Engineering or Chief Architect), the other would solid-line slot right underneath him/her along with the rest of the organization (e.g., developers).

    I'm guessing I might just be pointing out semantics, but just wanted to share my thoughts.
  • robertsnell
    Nice article Mark. I'd quite agree with you, having seen great tech leaders work well in a startup. Part of the trick is finding someone with a passion for the job, team and product. Someone who clearly ticks all the technical boxes and has proven themselves. And someone who might even be slightly obsessive about getting it done right, whilst remaining pragmatic and not being too much of a "code snob" to the extent that it affects product deadlines. I've also seen at least one case where a decent technical team was put to task on developing a product without a super strong tech lead. The product owner wasn't entirely comfortable that they were doing the right thing. So the PO asked a trusted expert from outside the startup to cast an eye over what's going on and give constructive feedback, which worked quite well.
  • Thanks for sharing your thoughts Mark! I am myself in the early days of a start-up and often get confused of what title should I use Dir Technology (or VPE) or a CTO - As of now I see myself shuffling between the two roles, there are times when I go deep into silos and there are times when I have to look around and make sure my team is sailing with me.

    My thoughts ...

    I feel VPE is a path to a good CTO / Chief Arch, esp for start-ups. The Chief Arch needs to set the technical direction and the one taking decisions: should it be MS Exchange vs Google App, Rackspace vs EC2, Python vs Ruby vs PHP etc. Its like defining the boundaries within which the VPE can play / act. All such decisions have a large impact on the team skills (esp tech), productivity, cost etc - Chief Arch does not need to worry about tracking to the plan but should be able to help VPE deliver on time - by identifying solutions - alternative ways to solve a problem, short term hack vs long term fix, performance / scalability hit vs functionality etc. As a VPE one is learning and being groomed to handle multiple issues - always in the thick of things - from people, scope, customer, tech etc but not necessarily taking decisions outside the defined boundaries or identifying which features to be built. VPE will often need to be pulled out of the weeds and I feel this is one of the critical roles of a Program Manager (in addition to what you mentioned).

    Program Manager is someone who co-ordinates, manages and prioritizes scope - this person has a good view of what features need to be built, by when and why and in a way should be asking the VPE to build ABC features in XYZ time. The Chief Arch needs to be step in on how the VPE intends to "build" those features ...
  • Great post. Found a link to it in the AVC.com comments section.
  • When I was at Amazon.com, we had a role called Technical Product / Program Manager (they watered down the role to Program Manager in 2008 because it was so difficult to find TPPMs). A TPPM is someone with the technical skills to be hired as an entry-level developer, the product vision to define new features / roadmaps, and the people + process skills to manage the design, creation, launch, and maintenance of features on the site. (I was the TPPM for the Personalized Recommendations team, which does "Because you bought this book, Amazon recommends...")

    I think it's a key skillset that needs to exist within a young startup. I've seen some startups founded by an MBA + CTO pair and the lack of strong product skills is obvious.

    WRT CTO vs. VPE, I agree with you generally. I like to think of CTO as Chief Architect + Technology Evangelist. Someone who is up on the latest technologies and is thinking about their application at a theoretical level, then can go out and present it at a conference and via white paper.
  • This is great! I have had this conversation with so many technical founders who think somehow that the VP Engineering role is "more important" than the CTO role -- or vice versa -- and so are hell-bent on taking the role they are least suited for. I immediately sent this off to several of my clients in the throes of these kinds of decisions. (It has such credibility coming from an entrepreneur and VC)
  • Laurent Boncenne
    Great post again Mark ! Thank you !
    One thing I'm not sure I understand perfectly is the work expected from both positions...
    The way I understand your post and how I understand the difference between the two is that a CTO is the one who would be heads down into the documentation of any new technology (or say framework or language, if you were a web technology startup), and break it down for the VP of Engineering to process it for the devs. He would be the one who would try out said technology to see how it can work/apply/scale with the project in development.
    The VP of Engineering would however study how he would apply that new tech to his team of devs to come up with a piece of code. (My understanding again)

    Then again, the CTO and the VP of Engineering would work both at the same level with the CEO (or product guy with the vision of the product who just happen to be the founder of this startup) to streamline every process needed to get the job done.

    For instance, If we take Koral (your last company if I recall correctly), you were the CEO and the CTO would analyze what sort of technology would be best to handle the users problems (file sharing, version control, multi-user, centralized company dashboard, etc.). He would be the one to test if php could handle the job of delivering such data or if python would be more appropriate.
    Then he would meet with the VPE to put things into action, ensuring what is expected is handled the right way before any code is written.
    According to my understanding, the VPE would break down this vision tied with one technology to assign it to dev teams to get the job done.
    Do I get the whole issue right ?

    I've been recently talking with two friends about a project I have (software development) where both have coding skills but one is more on the heavy coding side and the other one seems to try to get away from coding, and I wonder if a team of founders like that could work. Like, one business guy who connects to the dev guy through another (in my mind, to illustrate : CEO to CTO to VPE would be how it would work, the CTO would translate and refine and double check the expectations of the CEO).

    Is there anything I don't get right ?
  • philsugar
    Why outsourcing doesn't work:

    1. The requirements for a software company are never in black and white.
    2. The ability to have somebody to hear direct feedback in person from both sales and support is priceless.
    3. The ability to have somebody tell you that no, you're wrong, there is a flaw in your reasoning versus yes boss, I'll code that right away boss is just as priceless.
    4. Frankly you as small company are going to self select into the bottom quintile of programmers. Think somebody in India for example is going to want to work for Microsoft, GE, Intel, Huge Investment Bank, or you??
    5. Cost savings in which I do not believe....but I'll take on anyway.....as the company matures development will be 20% of costs or less. Assuming, and again I totally disagree, that you can shave a bit off that...is saving a percent or two overall worth not having a core team?? The sales needle is much more important.
  • Thank you, you've written my post for me! Seriously, these are the reasons.
    - water cooler matters
    - customer proximity matters
    - team productivity / learning in one place
    - requirements evolve too fast
    - etc.
  • Spot on! I agree with you and strongly feel outsourcing your startup product development just does not work. You need to have The tech person in house who can show technology leadership and all the attributes you mentioned. I personally was involved in a startup, joined them as CTO after they got their funding from a top VC firm, who had outsourced their complete product development right from day one. Needless to say the product could not really scale, had a tough time making the non-tech founders understand the issues and what was needed to get it fixed. Eventually the company could not scale and folded down.

    Having the The tech person inside is very very important especially if you are building a tech product.
  • First time I haven't been nodding my head in complete agreement with a Suster post. I can only speak anecdotally from working at a handful of venture-backed startups, but our CTO's were never the cliche of the artiste coder/hacker. They were brilliant technologists (with coding productivity worth 10 normal devs) and led the charge on technical vision and solving the gnarliest technical challenges (I dislike "CTO's" that are all vision, no coding), but they were all well-spoken and great evangelists with customers. They understood both technical issues and the business issues. Brilliant developers who couldn't cover those bases stayed "lead/senior developer".

    Whether one calls your startup's technical leader "CTO" or "Chief Architect" doesn't really make a difference to me -- I think the title just needs to be strong enough that they can talk to prospects/partners without people wondering who else might be more important.

    I do agree that the VP of Dev is all about process, and it's not a hire you need with few developers. You need people who just roll up their sleeves and get stuff done. At an early stage, process can be managed by your CTO/Chief Architect, or by one of the other coders, or even an ex-coder product manager if you have one.
  • well, giff, count yourself as lucky then. I find that more often than not the deeply technical person isn't perfect in most customer situations and often doesn't want to do it. also, having them do a lot of customer meetings can be a drain on their productive time. It's true that there is no "one size fits all" but I wanted non-tech founders to have a distinction in their minds between the deeply technical leader and the technical leader who is a process jock.
  • Yeah, while none of those companies turned into Google, but I have been lucky to work with some smart people. Agree re: customer meetings -- I didn't mean to imply a pre-sales like role, but swing into important deals to provide some needed "wow, these people are brilliant!" factor.

    "I wanted non-tech founders to have a distinction in their minds between the deeply technical leader and the technical leader who is a process jock"

    Right on. On a related note, I've met a lot of star programmers who faced a career dilemma in bigger companies: they want to feel like they are "getting ahead" but then found as they got more senior that they had to do less coding and more people-management and process. Which is why great hackers should all do startups, so that it's their *company* that gets ahead and they can still code! :)
  • kamadoll
    Really liked your article - being an ex entrepreneur/CTO and now a VC, it touched a lot of chords with me. Personally I think of myself as the CTO in the startup gene-pool so was lucky to find a good VP Engineering to team up with. We scaled up to do sales, marketing and program management in the early years.

    Further, instead of program management I have always favored a product marketing and product management team which like CTO/VP-Engg require different skill sets.
  • Yeah, I like strong prod mgmt and prod mktg teams - a future post since these are often confused, too. sometimes having just a "planner" that coordinates across the company can be worthwhile as you grow. the Prod mgmt and prod mktg people are usually not interested in this internal only role.
  • kamadoll
    I wrote up a post a few weeks ago on trying to disambiguate product management and product marketing and marketing programs. http://bit.ly/9TS9MT
  • Al
    There are several inherent issues in most startups though that prevent the structure you describe above.
    a) If the startup was started by a non-tech founder (who then hires a CTO) - that founder generally does not understand and does not want to deal with technology and hence ends up having the VP eng. report to the CTO. This happens much more than you would like - and does happen in a lot of well funded startups as well. When I worked for my first startup, that was what happened. As you can imagine, the classic problem of a CTO not being able to manage, now having a VP eng. under him.

    b) If there is a CTO founder as part of the founding team, often the problem is reliquishing control. As you can imagine, people feel a certain power by having other people reporting to them and CTO's generally do not want to reliquish that control to the VP engineering especially if they are founders. That causes the classic internal battle of ownership.

    c) The last classic problem I've seen is the task of hiring a VP engineering where the CTO plays a big role. He often hires for the things he lacks, which most often is the process/people orientation. So the VO eng. ends up being more closer to the "program manager" quadrant in your diagram i.e. lacks the technical expertise.

    All of the above issues cause the startup to go haywire from a technology leadership perspective unless corrected and I have personally experienced that.
  • Totally agree, Al. It's why I like teams where the CEO is at a minimum a great product guy with technical experience and I like teams where there is an emphasis on engineering talent more than biz dev talent in the early days.
  • Rahul Chaudhary
    Mark,great post. Here are some of my thoughts

    A CTO would be a visionary and a technologist who can see the future. A CTO would be doing more of a architecture, technology, vision, evangelism etc. Where as a VP of engineering would be more involved in delivering products, managing process, schedule and resources etc.

    In early stage of the startup, one person can do the job of both CTO and VP Engineering until the development team reaches to a certain number of people.
  • absolutely. and you make my point that both of them are actually technical. but one is tech visionary and the other is tech delivery. those are different things and people. but, yes, in the early days wrapped in one.
  • I think the confusion also holds with technical people too. For many architects, the CTO role that many aspire to be, includes the engineering management responsibility.

    The confusion comes from starting first with "titles" and then backing into the responsibilities.

    Your map serves well to clear this up. On my team, I like to separate the responsibility by roles first and titles later, here's the product development trifecta,

    - Technology Leader - what you call lead arch->CTO, but could come from any level, from rockstar developer, senior developer, principal architect, chief architect etc

    - Engineering Leader - VP eng but could be at various levels of experience - from process-oriented developers to development managers to Eng Director

    - Product Leader - for web/media products where tech+product are entwined/symbiotic, this fits into your Program Manager segment. This also works particularly well where teams are doing agile development, as the engineering function is self-organising (in terms of project mgmt), based on the priorities of the product roadmap.
  • Mark, these are great insights from the ground floor of good startups !

    btw: we fit the charter well ;-)

    -best, Miten
  • This article makes me realize one thing: 99% of companies hiring senior or experience technical staff confuse the terms.

    Preamble - until now, I have considered myself a CTO, but in reality, I'm closer to a VP Engineering. I can code in various languages, and I'm proficient in different technology fields, but I'm not a rockstar PHP developer, for example, or a kick-ass firmware developer. I see the VP Engineering as someone who can manage a technical team as he cannot be bullshiitake'd by developers with techno-babble, and can also talk to marketing and sales. Ideally every company should have one of these, but in many cases it is missing. This creates the huge gap often seen between the tech and sales/biz camps.

    A personal experience regarding the confusion of terms. I was contacted by a headhunter for a job as CTO at an established leading online outlet. Right off the bat, I was confused as to why an organization with close to 200 employees, cash-flow positive, and expanding to other countries, didn't already have a CTO. The requirements for the position read:

    "Extensive experience in software development including roles as CTO or head of development. Must have been a CTO for at least two years."

    and

    "Experience with open source technology and the LAMP stack. (Linux, Apache, MySql, Php). Java development experience."

    This was mixed with requirements about being a technological leader, proven experience managing 50+ developers, project management skills, and so on. After the initial phone call with the headhunter, it became clear that they were not looking for a CTO, but a VP Engineering. They wanted someone who had already built, in their own words, "an Amazon or eBay". Clearly I didn't have what they were looking for, but they were asking the wrong question. They wanted someone who "will be the leading technology executive in the company and will play an integral role in the company’s strategic direction, development and future growth." This is not a CTO by your definition.

    I have seen this same scenario over and over again, I hope your post is read by hiring managers, HR people, recruiters and headhunters, so that they get their messaging straight.
  • ramakrishna tumuluri (aka rk)
    Excellent post. The picture captures it all.

    To take it to the next stage, I would add "concentric circles". The larger, outer circles would have more "roles" such as "R&D chief", "QA", "Support" and other dedicated functions. Would be usefull if you map it out to include "sales", "marketing", "biz-dev" and other vital functions needed for successfull ventures.

    /rk
  • Thank you. I would like recruiters and n0n-tech managers to read it. It's a source of major confusion.
  • You never stated a reason why you don't believe in outsourcing, only made a movie analogy. Why don't you believe in outsourcing? In what scenarios have you seen it fail?
  • It's a big topic - I'll pick it up more fully in a future post. Sorry.
  • Great, thanks.
  • Ryan
    Mark, I agree on the distinction between different types of engineering talent, regardless of the actual titles used. One thing we struggle with is balancing the 'perfectionist' nature of our high-tech guru with the need to detail schedules and timetables for prospective investors.
  • Doesn't everybody! The true definition of an artist. Image how book publishers or music labels must feel! Or the owners / managers of professional sports teams. I think these analogies hold well.
  • Another axis would help here: vision. Your quadrant essentially "favors" the VPE, and doesn't highlight why you often also have a CTO. That maverick, free-thinking, unmanageable type (which you've blogged convincingly about on the sales side) is just as powerful and effective on the technical side. The management burden all too often weighs down/hinders technical forward thinking, so separating them works well.
  • Hey Well. I hope I didn't imply that the CTO/ Lead Architect was unimportant. Quite the opposite. I would hire the chief architect well before the VPE. My point is EXACTLY yours. Too many non-technical founders think this is one role, titled either VPE or CTO. My point is that both are different and both are hugely important.
  • yeahbaby
    I agree with the point that you should be skeptical of outsourcing especially at the start of the startup. Actually I feel the same way for a medium size to large company. Skepticism doesn't mean don't do it , just think about what you are trying to achieve and what the motives are. If the motive is barebones cost related, then ask yourself if the technology really plays such an important role in your startup. If it does, then cost is not the absolute number 1 criterion and you should really hesitate before moving the operation to say India or some far off land. I have long experience with this and the grueling nature of maintaining context and work in process between two widely divergent timezones, as well as questions of quality because of the frequent breakdown in context on the part of the remote site is not only a productivity killer but a business killer.
  • As a simple framework this should be helpful to non-techies. More importantly, this lays out the skillsets that are necessary for tech companies, especially as they mature out of their uber-scrappy very early days. I don't think it's true that you need a person to fill each box so to speak, or that people's personalities always pigeonhole them into one, but each of the attributes needs to be covered by someone. Nice synopsis, Mark.
  • thanks, Aaron.
  • Love the Howard Roark reference. Every entrepreneur should read the Fountainhead.
  • Every person should read it!
  • My 2.5 cents from the trenches

    At it's height (alas it's been and gone), the previous company I founded had 12 full-time developers.

    When it came time to appoint a CTO (for both investment and management purposes) we went for someone as experienced as we could afford with a heritage of study at Cambridge University and years spent at the IBM advanced internet lab. Great knowledge, loved documentation, investor-friendly and highly organised.

    At the same time we had a young hacker, junior PHP developer, highly-motivated, hungry as anything, lived, slept and breathed technology.

    He was the youngest in the team by far, the most inexperienced but probably hands down the most talented. He had a raw, unpolished ability which in the years I have been working with developers I had never seen before.

    Needless to say the older more experienced team members were drawn to him and before long he set direction for major tech decisions, not from force of personality but from rational insight and thought.
    He didn't possess the pedigree, experience or credibility which the CTO did but his passion, momentum and dedication were contagious.

    I don't know about official org charts or management structures, but I do know that in the early days of a startup whilst your really swimming against the tide and hustling to make your mark and break through, a young dedicated hacker is worth a hell of a lot more than a suited resumed CTO
  • I couldn't agree more. It's why I always tell young companies to hire people who want to "punch above their weight class" rather than senior guys with big resumes. It's also why I've said in the comments that if I were a startup I'd hire a "chief architect" rather than a "CTO" in the early days. The implication for me is somebody who on paper is slightly more junior but probably delivers more immediate value.
  • Interesting distinction, but as usual with titles, they are stereotypes and stereotypes while often being accurate, are also regularly inaccurate. I'm a CTO and fall inline with everything you wrote on that job description except:

    1) I care more about the customers and the end product than the code
    2) I have been told by many that I'm the best development leader they've had; have been a mentor to people (according to them), and love working with people and helping lead them all to the finish line. I've managed developers my entire career and hope to always be in that position. I consider myself the 'head coach' of development teams.

    I fit everything you wrote under VP Engineering as well and have frequently been pulled in to work with the client, but its true I hope one day to no longer be a code monkey, at least not one that is regularly relied on to churn out code, but I could never give up coding. I hate testing, but I love documentation. I keep crazy hours, I like setting up a team structure, doing team training, team building, managing daily, but hate setting up a development schedule (mostly because I don't agree with the methodology most companies use for this). I also love setting coding standards and the processes behind them, but all of that is in the name of building a great product to make customers happy, not to build great code to make developers happy.

    I'll think you'll find that David Heinemeier Hansson from 37 signals is very similar to me as well in how he fits into these categories and I would imagine we aren't the only ones. So I have no idea what title you would give us. For now, I use CTO.
  • Brian, I accept that there are talented people that can do both. I also accept that in order to come up with a generalized "framework" you need to stereotype and the description doesn't always hold. My objective with the post was to explain to non-technical founders / CEO's that there is often a distinction between the "prototypical" CTO and VP Engineering. All too often I find people confuse these individuals. Thank you for your input - I appreciate your adding to the debate. I know it's not as "black-and-white" as my framework might suggest.
  • Yep, I understand. And I completely agree with your stance on outsourcing. I've seen so many startups get hurt by that, and eventually have to bring talent in (like as you said one of those main positions) and often dig their way out of the hole the outsourcing created.
  • John C. Smith
    If this is dealt with elsewhere in the comments I'm not seeing it but I'd be curious to hear your thoughts on how product leadership fits into this matrix, particularly for early stage companies.

    I've seen a number of start-ups where the CEO is the product person but doesn't have time to really do the PM job. Thus tech dev't happens without proper business side input but rather contentious weekly mtgs between the CTO / VP Engr + CEO.

    So I guess the question is how do you view the product role in an early-stage company, and where does it fit into the overall rubric you lay out in the above post?
  • It's a big post and a topic that has been on my mind for a long time. I promise to write about it soon-ish. Next 60 days or so.
  • ES
    Mangled sentence: "A great VP of Engineering managers as well us as they do down." I think you meant to write "A great VP of Engineering manages up as well as they do down."
  • Yes. thank you! darn that 1am blog posting! I'll have to stop that.
  • Tony Karrer
    Mark - I think there are lots of startups that outsource their CTO - especially initially. A couple that you will know in So Cal are eHarmony and MyShape. And there are lots of others.

    I would argue that trying to hire an actual CTO or even Chief Architect out of the gate, especially full-time is NOT a good choice unless you have lots of dollars. It makes a lot more sense to outsource that to a part-time CTO and only hire the hands-on developers. I've written quite a bit about this on my blog:

    Part-Time CTO


    Technology Roles in Startups


    Technology Advisor


    CTO Founder


    Startup CTO or Developer


    Acting CTO



    Granted there's often a gap between non-technical founders and developers, but that can be closed by the right person.

    I think your "immediate pass" philosophy might be something you want to think about.
  • Tony, I understand you feel that way. You position yourself on LinkedIn as a "an experienced CTO for hire." http://www.linkedin.com/in/tonykarrer so that makes sense.

    But I fundamentally disagree with you. We'll have to leave it there because on this topic we're going to be "oil and water."
  • Tony Karrer
    So if someone hires me to help them, they should hide it from you? Or they should list me as an advisor and put their Lead Developer front and center?

    I've heard about this perspective before, but it's surprising to see it in black and white. I actually appreciate that you are open about it.

    Curious what you would think about a CTO advisor type person helping Launchpad companies? I know a couple of them need that kind of help.
  • Yeah, for me it really is black-and-white. I'm not opposed to your acting as an advisor. In fact, that could be quite helpful. But I simply don't believe in funding a company that has a "CTO for hire." At least not as an early-stage VC. It's nothing against you - it's a sign for me of a company that doesn't have the DNA I'm looking for.
  • philsugar
    Spot on.

    I would add that they need to be a major equity holder and they need to have experience working on products that are sold, i.e. not internal development or consulting projects.

    I don't really care what you call the people, you need the CTO first who can also code like crazy.

    I don't think outsourcing is ever an option..... If you are a technology company you need to be just that.
  • CTO / Non-founder
    Can you define major equity holder? As a CTO / Chief Architect I was the early rockstar that grew the company's product from nil to a platform that allowed us to grow from $1M -> $10M over the past 3 years - providing both the technology leadership and coding a bulk of the product. As a non-founder I have a relatively small share of options that represent what I think is an unfair total % of the company. I haven't been able to find much out there in the ecosystem that talks about this in an open forum. I'd love to hear thoughts on the topic.

    Oh and this article is spot on. Almost a little eerie how close to reality this is for us.
  • philsugar
    I certainly can't respond to your specific situation, but in my world there's really three legs to the stool when starting up a technology company....

    A CEO that can sell, understands finance, and technology,
    A rock star programmer that loves to code, has built professional products, and uses source control, and knows how to build real products,
    An organization/customer relations person, which probably was the idea person.

    I've always felt the equity should be equal, but vest over time. I would not take this as the common view however.

    I do have a saying that everybody is replaceable, and the way to ensure you are replaced is to try and make yourself irreplaceable. That being said I think the person I'd least like to replace at a technology company is the CTO, and that is why they should be tied to the business with a major equity stake.
  • SZ
    "It’s what makes Google, Google and Facebook, Facebook." should read: "It's what makes Google Google, and Facebook Facebook." Otherwise it sounds like Google and Facebook make Facebook :)
  • OK! I'll try it your way ;-)
  • Zernan
    Why was my "Sales...you need a kick ass sales guy" comment deleted??
  • It wasn't.
  • Zernan
    Sales...you need a kick ass sales guy.
  • The VP of Engineering provides the focus that brings to life the CTO's vision. Without focus, the grand ideas, eloquent architectures and revolutionary plans are just that, plans. Both are tough jobs because it can take a big leap of faith to put your trust and career into a vision. Good VPE's are pragmatic and know how to focus the team on the important stuff while still maintaining the artistic vision of the CTO.
  • Thanks, Jarie. I hope my post didn't imply that the VP, Engineering wasn't a vital role in a startup. To the contrary! It is very critical. Especially as you scale. Thanks for your 2 cents.
  • Mark - great post. We had a great CTO at my last company - brilliant, loved to evangelize/teach, deeply technical. He loved being in front of clients, which he was great at and clients loved but often led to the overcommitment problem. As long as we had someone along that could pull on the reins when the hyperbole got overboard, it worked great.

    Totally see the beneficial yin/yang between CTO and VPE.
  • John W
    Nice piece. My one qualifier is that you are talking very specifically about small start ups. In a larger start-up (say more than 50) CTO alternately means the person for whom all things engineering reports to and is the representation of technology at the executive level. Otherwise engineering/technology gets represented by the COO which never goes well.
blog comments powered by Disqus

Previous post:

Next post: