Scaling A Company While Controlling Costs

Over the last decade, the costs of scaling a company in the bay area, NYC, and many other startup regions in the US has escalated sharply. We have encouraged our portfolio companies that are located in these high cost regions to build out secondary locations where they can hire high quality engineering and other talent at lower costs. Locations in North America that have been attractive to our portfolio companies are cities like Des Moines, Indianapolis, Toronto, and a number of others.

Another approach that has worked well for our portfolio companies is to have a secondary location outside of the US, where talent can be hired a much lower cost than the US.

So it was with interest that I read this in Zoom’s IPO prospectus;

we have a high concentration of research and development personnel in China

It turns out that Zoom has most of its engineering team in China, where they can hire high quality engineering talent at much lower cost.

Look at this P&L.

Zoom spends most of its opex on sales and marketing because it has kept its engineering costs lower as a result of building product in China.

This is something to think about as you scale your company.


Comments (Archived):

  1. awaldstein

    Yup–much to learn from looking at Zoom.Such a great product. Such an impossible market to crack.It is almost unimaginable that I will change using it and with every new network or project i personally extend its customer base.

    1. Matt Zagaja

      A funny thing happened at my org (government of all things!) yesterday when I learned that despite using Cisco WebEx for years the Deputy Exec Director asked our IT staff to look into switching to Zoom. It can take years to get buy in from management (especially in a gov org) for a new product but Zoom is somehow managing to do it.

      1. awaldstein

        The inverse happened to me last week as mentoring a team out of a large university innovation lab using Webex and made it clear, we are moving to Zoom.Work has changed. It is more global, more fractional, more diverse so these tools–video, large group communications (huge need), others, are simply timely.And the rate of change is accelerating.Think about the slow ponderous change from Siebel to Salesforce as against the dramatic adoption of Slack in the enterprise.

      2. scottythebody

        I noticed in the last year that more companies were hosting Zoom meetings, but I hadn’t really given it much thought. But I was at a conference of security people from all over the world two weeks ago and pretty much to a person they were all on Zoom. Not sure what that means, but it’s definitely the first time I’ve been in a professional community setting and noticed such a radical switch from WebEx and whatever else the corporate standards have been (Skype for Business? GoToMeeting?).

    2. fredwilson

      We run USV on Zoom and Google Apps and Airtable

      1. awaldstein

        Don’t know AirTable and will check it out.

        1. Richard Barker

          Spreadsheet for words. It’ll change your life.

  2. Vartika Manasvi

    Yes this is so true and with Trump’s administration more & more companies are moving engineering teams to either Canada or their own home towns. We saw major influx of U.S. founders hiring global tech talent and venturing out.

    1. Pointsandfigures

      Was true long before Trump. Zoom didn’t start in 2016

      1. Vartika Manasvi

        It was but then companies were also risk averse and now as they accept the remote culture they want to build global borderless teams with global talent. On the highlight of Trump, here’s more for the case in point –

    2. Pete Griffiths

      Curious why you see this correlated with Trump.

      1. Vartika Manasvi

        Hi Pete, thanks for the question. I run that enables global tech talent hiring and what we found was that H1B’s in the U.S. (especially Indians) are moving to Canada for more secured careers or coming back to the country. That’s why I said Trump’s administration ignited that trend even more. See more details here –

        1. Vasudev Ram

          Hi Vartika,Just took a brief look at Interesting biz model. Would like to email you about possible synergies and questions. Did not find a contact page or email id there, on a quick look. Can you share it, either here, or to my email id, which you can get from https://vasudevram.github.i… ? Thanks.I’m an independent Python (and other areas [1]) consultant and trainer.Good profile overview: [1] (Python, Unix, SQL, programmatic PDF generation (creator of xtopdf), …):Course outlines and client testimonials:

        2. Pete Griffiths


  3. Mark Mc Laughlin

    Seeing this in Europe too. Dublin is expensive for start ups with the multi national tech companies here, so Serbia, Poland, Ukraine amongst others are becoming big second location software hubs. Cost difference is closing with EU countries though. There are no borders with software development.

  4. Matt Zagaja

    1. Assuming a typical startup, what would we normally “expect” them to be spending on engineers (/R&D)2. How does an org handle the time zone, communication, and other issues? Do the Chinese engineering teams work on a US schedule? Do design and product people also work out of China or is that more US based?3. How does recruiting work in foreign markets? Lots of travel? Does the company establish a “beachhead” in a country first?4. Taxes. My girlfriend for a company and is spending a year in their London office. Dealing with local tax law when having US employees work internationally is more than an adventure. Competent counsel/lawyers, compliance people, etc. are worth their weight in gold on this front.5. At what point in growth does this usually make sense? After a Series B? When headcount reaches 100?

    1. Pointsandfigures

      Also IP issues and legal issues.

    2. Dave Kruse

      Great questions. Major Disclaimer: We help companies and startups offshore in India. We also partner with startups. Offshoring isn’t always easy, but the gains in productivity are helpful.2. You need a change in mindset. If you’re used to having your tech team around the table, this will feel different, awkward at first. You need to get your team onboard before trying to offshore.Then you hit the major issue. Communication. There is nothing fancy about communication. You just need to be in regular contact, have daily standups. Communicate throughout the day on slack or skype. And have a very good git system set up. Our India teams have about a 3-4 hour overlap, which works out well. Have a daily standup in the morning and touch base over slack the rest of the day.For design and product people, we’ve generally seen them work in the USA. They need to be closer to the customers and understand the USA nuances. Although it makes sense for them to visit the offshore teams. Offshore teams often need clear direction on what to build. If the product team and designers can do this, then that is helpful. You can still practice agile just make sure the tasks and direction are clear.3 and 4. These are both a pain. One option is to work with the many firms that offshore. They help recruit and manage your team. You also don’t have to set up a company and pay taxes, It costs more per hour to go through a firm but it’s easier to scale your team and try out offshoring.But at some point you can set up your own team. Your offshore vendor might be able to help with that too.5. That’s a great question. For sure by series B you could afford to set up your own office. Even a larger Series A. Like above, you could partner with your offshore vendor to make it easier. Or hire someone in the USA with connections to the country and speaks the language.Offshoring isn’t for everyone. What is often overlooked is that there is amazing software talent across the world. They do need to be managed differently but the increase in productivity per dollar makes a big difference.

  5. jason wright

    Digital colonialism?

  6. Guy Lepage

    Wouldn’t it be wise to just HQ the company in a location that makes more economical sense? ie. Austin, Houston, Miami? You also get to create your own culture.

    1. awaldstein

      Building a company where the market is is always a consideration.Lots of ways to slice this but I personally like to be where the market is, with satellite offices to get pockets of work done.

      1. Guy Lepage

        When I had my digital ad agency (Vancouver based), I had only two [edited] Canadian clients. All of my other clients were in LA, NY, Baltimore, London, Germany. And that was an industry that requires constant contact with the clientele. Many face-to-face meetings per month.I don’t see how a tech company needs to be in SV or NY. I’m just not understanding why.The teams that I’ve seen move to the valley from NYC say that there is a bigger talent pool there. I suppose there is but the “talent” that they’ve hired is not top notch to be honest. To me the word “talent” means above average skill set. Which, as a small company, I can’t really afford. Also, if they’re above average talent, I’m ok if they live in Mongolia if they want because, I know they will get things done where ever they are in the world. It’s the “average” teammate that needs 80% of the hand-holding and attention. Hiring average developers can be done almost anywhere.Maybe I am missing something.

        1. awaldstein

          It’s never simple.It’s always case by case.In the last year two international accounts have decided to form NY presence around me as a foothold in their markets,(Big fan of Vancouver, went to UBC for grad school and need to make another visit soon.)

          1. Guy Lepage

            In the last year two international accounts have decided to form NY presence around me as a foothold in their markets,Gotcha. Nice. I guess if you’re planning on doing anything enterprise related it really makes a lot of sense. Basically you’re looking at it from the business opportunity side. Hm. I can see that. I wonder if someone has done a study on cost benefit analysis for startups.. I mean it is also easier to bump into great marketers and sales people just through osmosis of being in NYC. I’ll have to think about that a bit more. (Big fan of Vancouver, went to UBC for grad school and need to make another visit soon.)Nice! You should.. But you’ll need to open that pocket book.. Things are wildly more expensive than NYC. I lived in Kits for 15 yrs. Ran to UBC and around the sea wall every morning. Such a beautiful part of the world.

          2. awaldstein

            Was a TA in Eng Lit and used to take my group down to Wreck Beach to teach poetry.Time to take my son who was born in Lumby to visit regardless.

          3. Guy Lepage

            Haha.. Crazy Wreck beach.. A lot of the beaches in Vancouver are now nude beaches. So many wealthy europeans. It’s pretty much the norm there now.

          4. awaldstein

            Was a bit back then as well.Other spot–since we digress–was preopening poetry readings at the museum of anthropology as well. Magical.Going to book a trip!

          5. Guy Lepage

            Nice! Safe travels. ✈️

    2. Adam Sher

      Only if the founder would want to locate there. Do you want to do that or have the HQ in your town and then manage your remote teams? That may increase your (founder) travel, which may not be feasible.

      1. Guy Lepage

        I look at it like this.. A founder should want to be where ever the business is going to be best located. I would move anywhere that makes it easier for the business to succeed.

        1. Adam Sher

          Do you have an opinion on what the min. level of local team you need to locate your HQ there? CEO + CTO, more?

          1. Guy Lepage

            I would think it’s best to have the majority of decision makers in one spot, but I’m not convinced of that either.

    3. JamesHRH

      Location is strictly personal choice + confidence.NYC founders want to live in NYC. How do you think Austin got to be Austin, Seattle got to be Seattle and the SV got to be SV? Go where you want to be.It is not a material issue, never was. You are overthinking it a bit, b:c it matters to you.No harm, no foul.Need proof?How about the founder of RM Auto Restoration & RM Auctions, who was born, raised and stayed in tiny Blenheim, Ontario. He eventually sold out to a fairly well known acquirer.

      1. Guy Lepage

        Well, some places are definitely not wise. I would slightly tweak your statement.. Founders who actually care about being successful, should go where it will be easier to grow their business. There are sooooo many places on the planet that are not great for starting a business. One can try in Mongolia and succeed but it would be more difficult.I had a startup in Vancouver. Not a wise place to start a startup. Most people do not work there. They are mostly ultra high net worth individuals. The high majority of my friends from there do not need to work.So there are definitely bad places. But there are also a lot of great places. IMO.

        1. JamesHRH

          No argument there.With the exception of the people it tends to attract as citizens (not a universal, but, say an 80/20 statement), Vancouver is awesome.

  7. Pointsandfigures

    Yes and be prepared to manage it. One of the tough things can be time zones or communication. Had companies create engineering teams in India, Ukraine, Ireland.I also think you need to really pay attention to what your sales/marketing team is telling you from front of the house feedback so the back of the house engineering team can respond rapidly.All about communication.

    1. fredwilson

      Which, coincidentally (or maybe not) is the business Zoom is in!!!!

    2. awaldstein

      Always felt that having engineering in reverse time zone was super useful if not also crazy to manage.Can think of three companies where I ran global sales and marketing and at the end of every day meeting with customers, would debrief and get things built/fixed while I slept, then often share changes, directions, patches.It’s a bitch on you body though.

  8. William Mougayar

    Well, it’s global labor arbitrage. And it’s a competitive advantage to those who use it well.

    1. fredwilson

      Yuppp. Toronto benefits from this too

      1. Jevon

        I think Canadian companies are generally behind the (new) offshoring wave in comparison to what I am seeing in competitive US cities these days (NY/SF and even Austin and Chicago). We’ve had low wages and an abundance of talent here for so long that I think we’ve rested on our laurels. Now that reality is hitting in cities like Toronto, it feels like we are going to be playing catch up for quite a while.

        1. William Mougayar

          Behind- how specifically?

          1. Jevon

            There is not a lot of offshoring expertise in the startup community yet.

          2. William Mougayar

            Well, we have Nova Scotia and New Brunswick 🙂

        2. Vartika Manasvi

          Not really true, did you know In 2017, Toronto created more tech jobs than the Bay Area, Seattle, and Washington DC combined. I run and have 000’s of skilled immigrants on our platform moving out of U.S. More research and facts here –

      2. Jim Peterson

        In the end, we are all “talent finders” who can meet and find people who can help us get things done- no matter where they reside. We are in California and, using Zoom, we were sharing a screen with a partner in Dallas and it was like we were in the same room.

    2. scottythebody

      One of the greatest rigged markets ever. Capital is free to find cheap labor but people aren’t free to move to great workplaces.

  9. Tom Labus

    A 12 hour time difference can make it easy to hide when problems come up. Keeping everyone on the same page to the nuances of a project can be extremely difficult.

  10. Adam Sher

    We employ an off-shore software team from Ukraine to complement our local team. In selecting that group, I interviewed local, national, and off-shore people. The low cost of off-shore makes it practically a free option.We use Slack, begrudgingly, and Zoom to communicate, and have weekly stand-ups. Prior to working with them, we fine-tuned our work requirements so we could be as clear as possible as to the scope of work and its priority. What’s silly is that in the past, I’ve been less strict about work requirements, SoW, and priority when looking locally because I assumed the daily contact would draw out those things naturally. What a mistake.

  11. sigmaalgebra

    R&D in China? If Zoom wants that, okay.Gee, now that I have an EIN and a company name, I’m roughly, sort-a in business so have to think about such things:(1) Communications.If they are in China and I’m in NYS and we talk via Skype or some such, do we do this when they are asleep or when I’m asleep?(2) What language do we speak? Or, how reasonable is it to have people in China who can read, write, speak, and understand English WELL? As I have screamed often enough at AVC, it’s tough enough to get nerds in the US to do well with English — tough to believe it would be easier in China.(3) For the R&D, just how deep are those? If they are very deep, then I’d be concerned about protection of intellectual property.(4) The crucial core of my startup is some proprietary intellectual property, secret sauce, difficult to duplicate or equal — and so far has saved my tail feathers and tail as no one else has been able to do what I’m doing, although there are some weak, kindergarten efforts that show that they are starting to see the importance of the problem I’m solving, while I got delayed by unpredictable external total nonsense. No way do I want even hints of how my secret sauce works outside my own, close control — hmm, right, maybe a Faraday cage??? Having such hints in China? NOT A CHANCE.(5) The China team would need developer computing comparable with some of the best in the US. They will get that from where? I’d ship it to them, complete with spare parts, etc.? That would be more expensive than here in the US.(6) What the heck would the China team be researching? If it is anything at all original or deep, then that would be company intellectual property, and no way would I want it in China or even out of my HQ. If they are doing just routine Web page programming and SQL queries, hmm, then I’m not seeing the potential for big savings. E.g., so far I’ve written the code for the Web pages and SQL queries, and that code should be good for revenue north of $1 million a month. For more, will have to think about scaling:(7) For scaling, that needs consideration of the server farm. So, I just call up Amazon and dump all the server farm work on Bezos and AWS, i.e.. so that the China group could access my company’s work on AWS also? Hmm …. My guess is that for good scaling, the software people and the server farm and network people should communicate a LOT, and I don’t see a role for China in that or even AWS at all.(8) Eventually I might need to go face to face with the China group. So, I’d have to fly over there. Travel. Long distance travel. Expensive. To a strange land. No thanks.(9) I’m not so sure I want to use AWS as my server farm. Instead maybe hire a consultant for some weeks to advise me on how to set up and manage my server farm. AWS, Microsoft, Facebook, Google, no doubt all the big banks, etc. have done that; how to do it has to be well enough known. Reasons? Close, hands on location. Good, face to face communications with everyone involved. Full control of everything. People management down to the last guy pulling cable and system management down to the last virtual machine and LAN packet. Security in general. Security for the intellectual property. As simple as possible legal situations.(10) I already know where I want my company to be, in the suburbs of a really nice mid sized town in one of the fly-over states. For now, it’s Upstate NY, and that will be okay for now, maybe also later.E.g., in such a town, there are grass and green trees. There is a community that cares. People can own houses and raise families. People travel by driving gasoline powered cars. No subways. No Greenies. No AOC. No irrational, quasi-religious, NYT hysteria and guilt about CO2 and global warming total nonsense. No worship of electric cars; Electric golf cars and fork lift trucks, okay. No one cares about Pelosi, Schumer, AOC, Beto, or Mueller. Plenty of land for baseball, soccer, football, town BBQs. Small enough that I can have some influence on the local school board and am first name with the mayor and town council, Lions, etc. The town likes it that my company is bringing in money. Taxes are LOW, LOW, LOW. No one cares about the condition of the catalytic converters on the cars. It’s okay to burn trash in a wire incinerator in the back yard. In the suburbs, there’s no water or sewer system — people use wells and septic tanks. There’s plenty of cheap land. Some of the land has cows and horses. Some people raise chickens. There’s a farmer’s market with locally grown fresh vegetables. There are some good churches. People know each other. The streets are safe. People don’t have to lock their doors. Even a plumber or electrician can buy a house and raise a family with a stay at home wife. When it’s time to put up a nice, new building for 100 more employees, with nice architecture and landscaping, a pond with ducks and geese, covered employee parking, etc., the town zoning board tells me that the project was approved yesterday.China? No thanks.

  12. Jim Grey

    I’m in Indianapolis in software engineering. I guess it’s less expensive to hire here than on the coasts but I work with engineers in China, India, and Colombia and they’re far less expensive than any engineer I hire here.I strongly, strongly prefer to work with colocated engineers but I can’t ignore the costs.The engineers we have in China and India do good enough work but the time zone differences are a struggle. The engineers we have in Colombia are in our time zone and so are much more responsive.

    1. sigmaalgebra

      What do you have them doing?

      1. Jim Grey

        The same kind of work as any Indianapolis engineer: building product features.

        1. sigmaalgebra

          I remember Indianapolis: For two years, I was a grad student in the math department at Indiana University (IU). I was there to learn the math for theoretical and mathematical physics, especially quantum mechanics and relativity.They actually had some good people on the faculty, but the courses they had me in were a total bummer: They had me in Kelley, General Topology. The prof started off saying that the homework was to work all the exercises. Then I knew it was a joke — come ON, prof, can YOU work all those exercises? As an undergrad senior, I’d given a prof lectures on the book, one week a chapter and the next week about 1/3rd of the exercises in that chapter. The prof just sat and listened but once called in another prof to listen also. So, at IU I didn’t waste time listening to lectures less good that what I’d already given. I did another 1/3rd of the exercises and handed those in; I was honest and didn’t submit the exercises I’d already done in ugrad school. Waste of time course.They had me in a beginning course in abstract algebra. I’d taken such a course in ugrad school and written my ungrad honors paper on group representations theory. Early in the course I asked the prof if he was going to cover group representation theory, and he was shocked and scared and responded “That’s deep.” So, at the end, there was only one topic in the course I didn’t know well, Galois theory. So, I learned that in a weekend (I hate Galois theory); he gave me an oral exam; his tricky question was about duality and adjoints, and I’d covered that topic in a much more advanced way in functional analysis the previous summer, and that was the end of that course.The third course was on real variables from a poorly written book. He started with some set theory, and then gave a test. After class he called me saying my solution was no good. It looked like he wanted me out of the class. It happened that the previous summer I’d also had a course in axiomatic set theory. At the last moment I saw a solution to his problem and wrote it down quickly. In a hurry, I used little omega without defining it believing that the notation was standard — it was. When I gave him the definition of little omega, he said my solution was good, one step shorter than his. Right, it was and was. Then I asked him what he wanted to talk to me about, and he shook his head, waved me off, and said “Oh, now, nothing.” Bummer. I never saw that turkey again.I did a really good job on the teaching they were paying me for, did some independent reading, started violin in the excellent music school there, and found my wife.I went to DC and had a good career going, did a lot of studying on my own, the keys to my Ph.D. later.So, I remember some things about IU!There were some good math profs there; I just needed to do what the Princeton math department has at times had on their Web site — students prepare for the qualifying exams on their own. That’s all I wanted to do; being free to do just that is all I needed. When I was in DC, that’s what I did do.But IU had a fantastic music school: The music school grad student who gave me a first lesson in violin was a Stern protege. The student I took the violin course from next played the Brahms concerto in Toronto. He gave me the basics, and with that I made it half way through the Preludio, the famous one, Bach also used in one of his favorite cantatas, in the E major Partita and about half way through the Bach Chaconne. Those two should count! In DC I took a few lessons, and as I played the first page of the Preludio in time, in tune, effortlessly with good expression and at one point some novel fingering, he exclaimed “You could play in an orchestra!”. Well, not yet then, but there was promise that if I continued I would be able. My violin is still in its case, but I’ve gotten too busy to play it.For my startup, I’ll certainly consider Indiana, maybe Warsaw? There are a lot of towns in Indiana about the size of Warsaw — I’d consider those, too. My father in law was head of the REMC (electric power cooperative), on the bank board, on the school board, in Lions, doing REMC work in DC, etc. His wife was a nurse and taught nursing in Indianapolis. One of his daughters was a nursing student there.I would regard Indianapolis as too big. Sure, Bloomington would be a candidate. But I’d say that there are also good candidates in Ohio, Iowa, Michigan, Pennsylvania, West Virginia, Kentucky, Tennessee, etc. Upstate NY? Okay at least for now.

  13. Pete Griffiths

    Two risks spring to mind:1. culture2. intellectual property

  14. Adam Parish

    I feel like we could accomplish similar savings by leveraging the US more efficiently. For example, there are talented software and hardware engineers in Chattanooga, Tennessee. Just one example. BTW, I don’t live in Chattanooga.

    1. sigmaalgebra

      Yup, and Chattanooga did a lot with 1 Gbps Internet access.

  15. Vasudev Ram

    Might have said this before here, but recently started using Zoom due to a training client who recommended it. Working out well so far, except: they should make it possible to change the chat font size.On a related note, the same client also paid me for the training via Remitly, a (newish?) online payment service. Payment came through fast (Canada to India) in a day or so (unusual, except from a few (marketplace, not payment) services like Codementor and Gumroad, both of which I am on too). Remitly founder is ex-Barclays, IIRC. Another good experience.Edit: The Remitly About page:

    1. Vasudev Ram

      Also, just saw this video interview of the Remitly founder:The Business Debate – Remitly – Transformation in Digital Banking:…No affiliation, just think it is worth checking out. Check out the fee differences, for example.

  16. Robert Goldberg

    Fred I agree with you AND many of the other commenters. Challenging to scale in NY, SF and other high cost cities. But we are finding that a number of companies can scale in secondary cities quite effectively, and not just as a second site, but the whole company. As you know I’ve been involved in Nevada economic development for years now. Reno is in an incredible Renaissance as are a number of similar cities. Better quality of life, lower cost of living, tax and business friendly……Give it a try! 🙂

  17. jason wright…”I am closing comments to this post because I don’t want to turn this blog’s comments into a soapbox for certain people and we all know who they are.”I think you’re beginning to overcook this issue. People are intelligent enough to be able to accommodate the broad range of voices (and their delivery) to be found here.

  18. Gen Kanai

    Software developers in China are no longer inexpensive. Baidu, Alibaba, Tencent, Toutiao, Meituan, Didi, Bytedance, etc. These companies are all inflating salaries of Chinese software developers faster than most markets.The other issue that Zoom doesn’t specify in the IPO prospectus is that Zoom currently operates in China at the mercy of the Chinese government wrt access through China’s government firewall. If the Chinese government decides to block Zoom, like they have blocked most other Western communications tools like WhatsApp or Viber, etc. then that not only affects their market access to China but also their internal operations if they have 500 developers in China.At the moment, only Skype and Zoom work through China’s firewall without a VPN. We should assume that this means that these services share data with China’s government.

  19. Betty

    @fredwilson:disqus Is there a way to search in the comments section? I hate duplicating comments/questions.My question was: why Des Moines and Indianapolis? (Can understand Toronto.)

  20. awaldstein

    Agree.Can’t see any reason not to have employees and functions spread out in most cases.Need to be where the market is. Need to cut costs and do part of the engineering elsewhere. Need to have a core team culture in an affordable place.Gotta wonder if this is simply not the new norm especially as we have more and more tools to make it possible.