When it comes to startups, products and services don’t mix

btgremlins

This post is about dancing with the devil (i.e. consulting) as you pursue your dreams and try to pay the bills.  I believe that it is near impossible to build a successful software product while maintaining a services business. (UPDATE: I should clarify that I am talking about the period *before* you have a successful product on your hands)

Keep it separate
First, what does this mean for a young, bootstrapping startup with visions of an exciting product?  Consulting work can be enticing.  While it is an unfortunate distraction from your startup, I completely understand needing to pay the bills.  My advice is to keep your consulting work completely separate from your startup. Freelance or incorporate a new S corp or LLC. It will be much more complicated to separate your product and services businesses later.

Focus
I’ll get this one out of the way although it’s a recurring theme. The basics: building a great product startup is incredibly hard and takes enormous dedication and focus.  Any distractions away from your mission slow you down and hurt your prospects.  If you can avoid doing consulting work, you will be better off. If you can’t, hopefully this post will be helpful in avoiding problems.

It is hard to turn down money
Business success can take on a life of its own, and not everyone agrees with me that there is such a thing as destructive revenue. If you are successful selling consulting projects, well you’re earning money, and as a business you’re supposed to make money right?  So let’s talk about making that money.  You need staff to both sell, execute and interface with the client on those projects.  Use contractors? Your profitability will be squeezed and quality will be spotty, leading to upset clients. Outsourcing works better in theory than reality.

Instead, resources are either going to come from your product efforts or new hires.  Actually, you need some additional skills like direct sales and project managers, neither of which can be effectively contracted out, which points towards hiring.  You also realize you need to implement new work processes to do the consulting effectively.  All this is steering you away from your core, and turning you into a different kind of business.

Conflicting Team Goals
You didn’t turn down that work because you wanted the cash, so now you’ve hired some new people.  However these new employees were hired to help with consulting work, and that’s where their skill set lies.  When you start talking about the glorious future of the business (i.e. the product), these hires start to question, “wait a minute, am I irrelevant to this future?”  That doesn’t help morale. Unfortunately, the math is simple — you’ve bloated the business with too many bodies for a lean, mean product startup machine.  Now you have extra pressure to feed the beast — a self-perpetuating problem.

Further, you have a problem in your incentive structure.  Your business is structured like a product company, with a focus on equity, but you now have a business unit which should be structured like a services company, i.e. focused on cash flow.

You also have a morale problem.  You’ve recruited people with glorious product visions, but now some of them are working for clients. Guess what, they hate working for clients.  The client staff is jealous of the product team, further worsening morale.

Skewed Priorities
Some folks think: “maybe we can build our product by doing consulting work and keep the IP?”  First, you have to negotiate keeping that IP, which probably means terrible economics.  But it doesn’t matter — you are still building a product for *a client*.  That client (the big brand you are so proud you sold) has special needs. That client wants *their* product. They are paying you; they want what they want, not what the general market needs.  If they are letting you keep the IP, they expect *even more* from you.  There is no product-market fit, only product-client fit. A glorious market of one.

(Note: one circumstance where you might be able to get away with this is when your “client” is a government organization which only cares about seeing the tech built not bossing you around on the details.)

Distraction to Death
You’ve sold these projects, so you need to make the clients happy.  However, there will inevitably be change orders, custom requests, changing client priorities, fad-of-the-day, and all sorts of little complications.  Your people end up working later and longer.

A fire drill hits and you need to complete a milestone.  With your own product, you just swallow hard and either cut features or add time. Well, you can’t do that with a client.  You promised to ship version X in three weeks and are way behind.  If you slip, at best you have an upset client and at worst you won’t get paid.  What do you do? You pull the engineers and designers working on your product onto the client gig and try to save the day.  Product? What product?  Morale? What morale?

Client services work is difficult.  If you’ve ever outsourced a software project, you know how hard it is to be happy with the results.  Do you really want to put yourself and your business on the other side of that deal? With clients demanding fixed priced work?  But again, if you have to do it, I recommend consulting as an individual or under a separate corporate entity.

The wrong investors
Sophisticated early-stage tech investors are usually turned off by services revenue.  They want a scalable business, not one that requires an hour of work to get an hour of pay.  So you end up with a different kind of investor. These new folks probably do fewer deals in tech, but they are excited about your space, the blue chip brands you are working for, and they appreciate the fact that you have revenue coming in the door (not the right revenue, but they don’t necessarily know that).

In this situation, you’ve just pulled in an investor with a very different risk profile to your dreams. You can’t do a bait and switch. When the time comes and you want to focus the business entirely on your product, you might find them to be a serious roadblock: “Wait, you want to just drop our lower-risk revenue, bet it all on an unproven product, and risk complete failure?” (of course the irony is that consulting revenue is not always that much lower risk!)

God forbid you might, like most startups, need to pivot.  These aren’t investors who understand “pivot”. You’re already in the doghouse because you want to turn down money, slim down and focus all resources on a product, but now, even worse, you’re a bozo who can’t even get the product right.  It’s not even their fault; it’s yours, because you chose the wrong investors.

final thoughts…

So a consulting business can’t create a product?
That is indeed my argument: a consulting business has the wrong structure, processes, priorities and people involved to pull off a great new product.  However, a consulting business can be fertile ground for good *ideas*, because you are exposed to lots of client pain points.  If you want to be effective at creating that idea (or whatever it pivots into) and bringing it to market , you are better off spinning out a new company with a small, dedicated, product-experienced team right at the start.

But there are lots of product businesses with strong services arms!
Lots of companies built services revenue streams on top of a product — a successful product — but I cannot think of many examples where it happened the other way around.

If you have seen examples where a services business built a great product business, please let me know.

UPDATE: great response from Scott Francis

(image from this awesome BT ad)

  • stephanwehner

    Hey there!

    I don’t find your advice convincing. Mainly because you are not offering much backup.

    Look at this one: “A true hero, indeed, Sandra [Schmirler] […] did all this while holding down a job and being a world-class curling champion” from http://www.myhero.com/myhero/hero.asp?hero=s_schmirler

    Stephan

    • Hi stephan – glad to get a skeptical voice and so thanks for speaking up. The short answer is that I’ve actually seen all of this happen. More than anything it is meant as a warning for those who decide to try to balance both, and to give them some things to watch out for.

      As for backup, let me know if there are particular sections where you feel like it is clashing with your own experience. The need for focus, the need for different skill sets to do services for big clients (compared to a lean product team), the pressures of client demands and deadlines, the problems with outsourcing, the morale problems of mixed teams with mixed goals… these are not obscure issues.

      Regarding your example, my point isn’t that you can’t do more than one thing at once or that you have to sacrifice all balance in your life. But there is no question in my mind that the more you can focus on a few key problems and opportunities, the better chance you have to do well.

      In startups we take on a lot of risk and uncertainty. Things get messy so why make it more messy than you need to? If you need to do consulting work, so be it, but keep it separate is my advice — to be taken or ignored!

  • stephanwehner

    Hey there!

    I don't find your advice convincing. Mainly because you are not offering much backup.

    Look at this one: “A true hero, indeed, Sandra [Schmirler] […] did all this while holding down a job and being a world-class curling champion” from http://www.myhero.com/myhero/hero.asp?hero=s_sc

    Stephan

  • Hi stephan – glad to get a skeptical voice and so thanks for speaking up. The short answer is that I've actually seen all of this happen. More than anything it is meant as a warning for those who decide to try to balance both, and to give them some things to watch out for.

    As for backup, let me know if there are particular sections where you feel like it is clashing with your own experience. The need for focus, the need for different skill sets to do services for big clients (compared to a lean product team), the pressures of client demands and deadlines, the problems with outsourcing, the morale problems of mixed teams with mixed goals… these are not obscure issues.

    Regarding your example, my point isn't that you can't do more than one thing at once or that you have to sacrifice all balance in your life. But there is no question in my mind that the more you can focus on a few key problems and opportunities, the better chance you have to do well.

    In startups we take on a lot of risk and uncertainty. Things get messy so why make it more messy than you need to? If you need to do consulting work, so be it, but keep it separate is my advice — to be taken or ignored!

  • brianknoblauch

    It really depends. Obviously a short blog post is going to have to deal in generalities.

    Some startups may have people who are not fully allocated and have extra time to bring in some cash by doing some service functions (prime example is a sysadmin at a small programming shop). There’s just not enough to keep that person busy all day internally, but getting GOOD part-time sysadmins is nigh impossible. So, pimp out (errr, offer services from) your sysadmin and take the extra cash! No need to walk away from it!

    • having been through the extreme pain of trying to hire a good sys-admin (that fit the company), I feel this one 🙂 But wouldn’t you, as a lean product team, rather handle sys-admin tasks within the existing team, or hire part-time help, rather than taking on a full-time specialist before you need it?

      • Something happened with DISQUS, so going to paste brianknoblauch’s response, and my response here:

        brianknoblauch: I’ve seen that road taken too. Having non-sysadmins attempting to do sysadmin tasks (especially on a part-time basis) is building upon a faulty foundation. You may not even notice the issues until the pain to fix them is extreme. I can’t recommend that route if you want a sustainable infrastructure!

        Giff: my challenge, in trying to be lean, is that I’ve worried too much in past too early about getting infrastructure really solid, when really I needed to focus all efforts (and money) on getting the product fit and solving my technical sins a bit later. it’s a balance I’m trying to figure out.

  • brianknoblauch

    It really depends. Obviously a short blog post is going to have to deal in generalities.

    Some startups may have people who are not fully allocated and have extra time to bring in some cash by doing some service functions (prime example is a sysadmin at a small programming shop). There's just not enough to keep that person busy all day internally, but getting GOOD part-time sysadmins is nigh impossible. So, pimp out (errr, offer services from) your sysadmin and take the extra cash! No need to walk away from it!

  • 37signals grew up as a services company doing web design. Now they make a living from SaaS products such as basecamp.

    • I was wondering about their roots. They are an interesting case of building tools to make their own jobs (of delivering services) more efficient, and then sharing those tools. Actually, as I think of it, Pivotal Labs is another example.

      So that’s one good exception area — tools built for internal collaboration or efficiency. Thanks!

      • Yeah – FogCreek here in NY did the exact same thing: http://www.foundersatwork.com/joel-spolksy.html

      • I think there are more exceptions in the SaaS / ASP model than there were in the traditional enterprise or shrink-wrap software models. These subscription offerings tend to focus more on “service” which plays to a consulting team’s core competencies a bit better.

  • 37signals grew up as a services company doing web design. Now they make a living from SaaS products such as basecamp.

  • I was wondering about their roots. They are an interesting case of building tools to make their own jobs (of delivering services) more efficient, and then sharing those tools. Actually, as I think of it, Pivotal Labs is another example.

    So that's one good exception area — tools built for internal collaboration or efficiency. Thanks!

  • having been through the extreme pain of trying to hire a good sys-admin (that fit the company), I feel this one 🙂 But wouldn't you, as a lean product team, rather handle sys-admin tasks within the existing team, or hire part-time help, rather than taking on a full-time specialist before you need it?

  • Brian Knoblauch

    I’ve seen that road taken too. Having non-sysadmins attempting to do sysadmin tasks (especially on a part-time basis) is building upon a faulty foundation. You may not even notice the issues until the pain to fix them is extreme. I can’t recommend that route if you want a sustainable infrastructure!

    • Giff

      my challenge, in trying to be lean, is that I’ve worried too much in past too early about infrastructure, when really I needed to focus all efforts (and money) on getting the product fit and solving my technical sins a bit later. but i’m not sure we’d disagree — all depends on stage of company

  • Yeah – FogCreek here in NY did the exact same thing: http://www.foundersatwork.com/joel-spolksy.html

  • Giff,

    It’s do-able but it is much tougher than your traditional VC funded startup. Bootstrapped companies cannot go after markets that require tons of capital, for example. Doesn’t mean it can’t be done, but I think you’re spot on that you shouldn’t start out building a product company if you’re consulting. Start by building the consulting business and once you have that stable think about what kinda product you’d want to build and go from there.

    Most bootstrapped companies keep the services business alive till they hit some amount of sustainable revenue in the product business, but the product business comes about as part of seeing a problem that they can fix.

    Some interesting bootstrapped examples include peepcode, asmartbear, and bingocardcreator.

    • thanks for comment — jason cohen (asmartbear) said he has a post brewing on the topic, so I’m looking forward to hearing his perspective

    • Yes Smart Bear was bootstrapped, but it’s an example that supports Giff. We don’t do consulting, never did it on the side, and considering the more-than-full-time effort required to make Smart Bear work it’s doubtful that it would have been successful with the distractions of consulting.I do agree with you that for VC-funded companies it’s *especially* unwise to pursue consulting and that for bootstrapping it can be a nice way to self-fund. However I competely disagree that this is “tougher than VC,” especially since the VC failure rate is 91%.I admit that if a person ends up consulting and never gets a product going, that’s probably still a better life-place than grinding away at a big company or failing at a VC-funded company!But there are a LOT more examples of consulting-funded bootstrapped companies who failed at product because of all the reasons Giff listed.

      • Jason,Thanks for being so epic. 😉 Love your blog.How do you bootstrap without any revenue? Savings till you’re ramen/cash-flow positive, right? That’s a luxury most people don’t have. So, it’s either consulting on the side (or a full-time job) and/or raising VC. I suppose getting to cash-flow and raising VC are similar amount of effort, so point conceded. Like Mark Pincus said at startupschool, it is freakin’ 2009 so you shouldn’t really be building companies that don’t charge people or have a solid revenue model. 🙂

        • Well I tried to say that if you *have* to consult on the side, keep it on the side — don’t blend it with your product business.

          Further, I don’t think you *can* raise VC anymore without some basic proof points on a product, not unless you have a big exit already under your belt. So that route still requires you to buckle down and focus on getting a crude version of the product right first.

          • Agreed. However, as I noted here: http://news.ycombinator.com/item?id=1083507, it is tough to keep it on the side… do you know of any exampes of people that have done this? It’s hard enough to start one company, that starting two (one consulting, one product) is going to be even tougher, no? 🙂

            Right, raising VC or even early stage capital without a prototype of some sort is almost impossible – the running joke is that as a first-timer, if you don’t have a prototype don’t bother, then if you don’t have traction, don’t bother, then if you don’t have revenue don’t bother, and once you do have revenue – why bother? 🙂

          • Well it’s not hard to *start* a company, it is hard to build it. Freelancing is probably an easier way to help pay the bills. If you really need to build a consulting shop, you should ask yourself where your passion is. If it’s on the product side, find someone else to drive the new consulting shop. I know it’s all easier said than done 🙂

            and I LOVE that phrasing for VC. Hadn’t heard it said that way before.

          • BTW, Giff I think you’re right -keep it “freelancing” under separate legal entity. Then it shows up clean on the books as additional investment you’re putting into the company, and not as a revenue stream that investors will expect you to keep investing in, nurturing, and growing.

          • BINGO

  • Giff,

    It's do-able but it is much tougher than your traditional VC funded startup. Bootstrapped companies cannot go after markets that require tons of capital, for example. Doesn't mean it can't be done, but I think you're spot on that you shouldn't start out building a product company if you're consulting. Start by building the consulting business and once you have that stable think about what kinda product you'd want to build and go from there.

    Most bootstrapped companies keep the services business alive till they hit some amount of sustainable revenue in the product business, but the product business comes about as part of seeing a problem that they can fix.

    Some interesting bootstrapped examples include peepcode, asmartbear, and bingocardcreator.

  • Something happened with DISQUS, so going to paste brianknoblauch's response, and my response here:

    brianknoblauch: I've seen that road taken too. Having non-sysadmins attempting to do sysadmin tasks (especially on a part-time basis) is building upon a faulty foundation. You may not even notice the issues until the pain to fix them is extreme. I can't recommend that route if you want a sustainable infrastructure!

    Giff: my challenge, in trying to be lean, is that I've worried too much in past too early about getting infrastructure really solid, when really I needed to focus all efforts (and money) on getting the product fit and solving my technical sins a bit later. it's a balance I'm trying to figure out.

  • Hi Giff; great post. Maybe it’s because we were in the same business, but I feel you on all of these points. I think there is/was the additional challenge of being a service-based business in a market that wasn’t already productized, thus trying to build tools and standards while meeting client requirements, even if you did have a service-oriented startup.

    • thx Adrienne — hope metaversatility is doing well 🙂

  • Hi Giff; great post. Maybe it's because we were in the same business, but I feel you on all of these points. I think there is/was the additional challenge of being a service-based business in a market that wasn't already productized, thus trying to build tools and standards while meeting client requirements, even if you did have a service-oriented startup.

  • thanks for comment — jason cohen (asmartbear) said he has a post brewing on the topic, so I'm looking forward to hearing his perspective

  • thx Adrienne — hope metaversatility is doing well 🙂

  • Excellent post! To consult or to product that´s the question, it´s the software entrepreneur’s dilema of the xxi century.

    The book “the business of software” makes a good case for both business models and concludes that wisdom is somewhere in the middle, since product companies become fast highly dependent on one hit product, and usually have issues reproducing the hit (with few exceptions called microsoft, google, adobe). It also can be argued that building a VC financed product, in an ivory tower with absolutely no point of contact with “reality”, following some biz plan to death is not the way to go either.

    There are some concepts that at least in theory appeal to me, one of them is the term “consultingware”. If you´ve read the Micro-ISV you must be acquainted with it. The theory is you narrow down your scope of business to one biz domain, facing the same kind of problems repeatedly, one technology of choice, plus some clever developer which gets tired of reinventing the wheel time and time again, so he comes up with this package that ends up reusing in most implementations and is now a corporate asset, voila the consultingware, halfway product, demi framework, which if taken cared of properly can grow into a product.
    To me the key is narrowing the scope, knowing how to say NO to evil cash, trying to accept only gigs that add in your strategy/direction, as long as your cashflow allows it, of course, and meanwhile continually invest in your productline.

  • Excellent post! To consult or to product that´s the question, it´s the software entrepreneur's dilema of the xxi century.

    The book “the business of software” makes a good case for both business models and concludes that wisdom is somewhere in the middle, since product companies become fast highly dependent on one hit product, and usually have issues reproducing the hit (with few exceptions called microsoft, google, adobe). It also can be argued that building a VC financed product, in an ivory tower with absolutely no point of contact with “reality”, following some biz plan to death is not the way to go either.

    There are some concepts that at least in theory appeal to me, one of them is the term “consultingware”. If you´ve read the Micro-ISV you must be acquainted with it. The theory is you narrow down your scope of business to one biz domain, facing the same kind of problems repeatedly, one technology of choice, plus some clever developer which gets tired of reinventing the wheel time and time again, so he comes up with this package that ends up reusing in most implementations and is now a corporate asset, voila the consultingware, halfway product, demi framework, which if taken cared of properly can grow into a product.
    To me the key is narrowing the scope, knowing how to say NO to evil cash, trying to accept only gigs that add in your strategy/direction, as long as your cashflow allows it, of course, and meanwhile continually invest in your productline.

  • Yes, 37signals and Fogcreek are *Always* trotted out as counter-examples.

    Which is why Giff’s post is correct. For every 37signals I can show you 10 consulting companies who have a product they’re half-assed trying to take to market, but are failing for the very reasons given here.

    Of course it’s possible! Just hard.

    Also knowing the hardships as listed here makes it more possible to succeed! Because if you accept them rather than ignore them with crap about how “If Joel Spolsky can do it, anyone can,” then you’re more likely to actually succeed with it yourself.

  • Yes, 37signals and Fogcreek are *Always* trotted out as counter-examples.

    Which is why Giff's post is correct. For every 37signals I can show you 10 consulting companies who have a product they're half-assed trying to take to market, but are failing for the very reasons given here.

    Of course it's possible! Just hard.

    Also knowing the hardships as listed here makes it more possible to succeed! Because if you accept them rather than ignore them with crap about how “If Joel Spolsky can do it, anyone can,” then you're more likely to actually succeed with it yourself.

  • Yes Smart Bear was bootstrapped, but it's an example that supports Giff. We don't do consulting, never did it on the side, and considering the more-than-full-time effort required to make Smart Bear work it's doubtful that it would have been successful with the distractions of consulting.

    I do agree with you that for VC-funded companies it's *especially* unwise to pursue consulting and that for bootstrapping it can be a nice way to self-fund. However I competely disagree that this is “tougher than VC,” especially since the VC failure rate is 91%.

    I admit that if a person ends up consulting and never gets a product going, that's probably still a better life-place than grinding away at a big company or failing at a VC-funded company!

    But there are a LOT more examples of consulting-funded bootstrapped companies who failed at product because of all the reasons Giff listed.

  • Jason,

    Thanks for being so epic. 😉 Love your blog.

    How do you bootstrap without any revenue? Savings till you're ramen/cash-flow positive, right? That's a luxury most people don't have. So, it's either consulting on the side (or a full-time job) and/or raising VC.

    I suppose getting to cash-flow and raising VC are similar amount of effort, so point conceded. Like Mark Pincus said at startupschool, it is freakin' 2009 so you shouldn't really be building companies that don't charge people or have a solid revenue model. 🙂

  • Well I tried to say that if you *have* to consult on the side, keep it on the side — don't blend it with your product business.

    Further, I don't think you *can* raise VC anymore without some basic proof points on a product, not unless you have a big exit already under your belt. So that route still requires you to buckle down and focus on getting a crude version of the product right first.

  • Agreed. However, as I noted here: http://news.ycombinator.com/item?id=1083507, it is tough to keep it on the side… do you know of any exampes of people that have done this? It's hard enough to start one company, that starting two (one consulting, one product) is going to be even tougher, no? 🙂

    Right, raising VC or even early stage capital without a prototype of some sort is almost impossible – the running joke is that as a first-timer, if you don't have a prototype don't bother, then if you don't have traction, don't bother, then if you don't have revenue don't bother, and once you do have revenue – why bother? 🙂

  • Well it's not hard to *start* a company, it is hard to build it. Freelancing is probably an easier way to help pay the bills. If you really need to build a consulting shop, you should ask yourself where your passion is. If it's on the product side, find someone else to drive the new consulting shop. I know it's all easier said than done 🙂

    and I LOVE that phrasing for VC. Hadn't heard it said that way before.

  • Pingback: Two zinger quotes — giffconstable.com()

  • sfrancis

    BTW, Giff I think you're right -keep it “freelancing” under separate legal entity. Then it shows up clean on the books as additional investment you're putting into the company, and not as a revenue stream that investors will expect you to keep investing in, nurturing, and growing.

  • sfrancis

    I think there are more exceptions in the SaaS / ASP model than there were in the traditional enterprise or shrink-wrap software models. These subscription offerings tend to focus more on “service” which plays to a consulting team's core competencies a bit better.

  • BINGO

  • Pingback: Product Man does professional services - The MurkyGrey blog()

  • SandeepS

    “Outsourcing works better in theory than reality”. Giff, I’m supporting a proposal to create a Q&A site on stack exchange. would you like to be a supporter? I feel you can be a very good contributor of your expertise for the community. Please take a look at: This proposal n

  • SandeepS

    “Outsourcing works better in theory than reality”. Giff, I’m supporting a proposal to create a Q&A site on stack exchange. would you like to be a supporter? I feel you can be a very good contributor of your expertise for the community. Please take a look at: This proposal

  • SandeepS

    “Outsourcing works better in theory than reality”. Giff, I'm supporting a proposal to create a Q&A site on stack exchange. would you like to be a supporter? I feel you can be a very good contributor of your expertise for the community. Please take a look at: This proposal

  • Pingback: Love Mark Suster’s Blog on Crappy Little Services Companies » Process for the Enterprise()