4 ms·

How much money should a startup have in the bank?

dennisgorelik 15y ago
Jason seems bitter about Google's Panda update.
- jasonmcalacanis 15y ago
  Yes, I am. :-)
  - dennisgorelik 15y ago
    1) Your opinion about quality of Mahalo content is obviously biased, so we cannot really rely on that. 2) Isn't it too expensive "burning $500K a month" just to find out what idea works? Discovery phase should be lean, right?
    - jasonmcalacanis 15y ago
      I'll start with your second question: In some categories lean can be a big number: search, making chips and (today) building a social network. Can you start with 10 people and 150k a month burn in these categories? Sure. However, if you're going to take a real stab at this type of problems in a massively competitive market with HUGE incumbents you need 30, 40 or 50 people. Also, "too expensive" is relative. Spending $500k a month might seem too expensive to someone starting out as an entrepreneur, but to VCs and founder who have been in the game for a long time they would rather risk blowing out 10m and figuring out a lot of problems quickly, than sit around and slowing figuring stuff out. The whole industry is biased toward big success. Just the nature of the beast. If you want to do something more modest--and i did on my last two companies--you raise small amounts of money and shoot for a small target. Neither model is wrong. In terms of #1, you can go check out youtube channels yourself. We basically stopped writing pages with great writers (but not experts) to creating videos with experts in them. This costs 10x as much, but the result is stunning. Not sure if you're into styling your hair, but we have one of the best stylysts in Los Angeles doing a series--think of it as a Khan academy course--on styling your hair. http://www.mahalo.com/how-to-style-your-hair/how-to-do-a-perfect-salon-blowout-at-home/ http://www.mahalo.com/how-to-style-your-hair/how-to-do-a-per... or photoshop tutorials like Lynda--except free: http://www.mahalo.com/learn-adobe-photoshop-introduction/ http://www.mahalo.com/learn-adobe-photoshop-introduction/ http://www.mahalo.com/photoshop-video-tutorial-how-to-make-a-movie-poster/ http://www.mahalo.com/photoshop-video-tutorial-how-to-make-a... Now imagine those not only as courses on mahalo, but also iPad apps. And imagine 1M videos with that level of instruction. That's what we're doing. ... and that all came from my vision of search results + content. Now, I still believe in that vision, but the truth is Google went from ten blue links in 2007 to what it is today: content + answers + some search links. we knew that was possible--that they would wake up--and they did. c'est la vie!
      - dennisgorelik 15y ago
        When I'm thinking about lean discovery, $150K/month seems too expensive too. Initial version of Google Search was developed by team of two or three people. That's less than $30K/month burn. "Neither model is wrong." -- I agree. However "lean model" is much more common and relevant to majority of founders than "$150K+/month burn during discovery phase". Another consideration: you made your previous company profitable with lean approach. But you did not make Mahalo profitable with "$500K monthly burn" approach. That should teach us something. "The whole industry is biased toward big success." - I assume under "whole industry" you imply "Venture Capital industry", not "Information Technology industry", right?
        jasonmcalacanis 15y ago
        1. The Google example is only relevant in 1999, when the market was nascent. Now that you have a big, established product there is--in all likelihood--no way to create a google killer at $30k a month. If there was one at that level, it probably would have been found by now. Established markets take bigger checks... nascent markets take smaller checks. 2. I actually did get Mahalo to profitability last December. :-) Of course, the Panda Update endend that brief, beautiful moment. 3. Actually, I do mean the entire industry. Now that the internet is global, fast and ubiquitous (mobile) everything good tends to get really big, really fast. Heck, bad things get big fast (Rebecca Black's Friday comes to mind), and so do super simple things (i.e. Angry Birds). the internet is biased toward big...
        dennisgorelik 15y ago
        1) While Google is efficient, there is no market need to create Google killer. Even if there is such a need, the goal of startup would be to create a model that at scale would overperform Google. Such model is likely to be created at less than $30K monthly burn. For example, team of two developers in couple of years may implement and fine tune some efficient social way of ranking search results. $30K/month * 24 month = $720K. 2) Jason Calacanis empirical rule: "if startup is overfunded then as soon as profitability is achieved it would be crushed by Google." :-) 3) "The internet is biased toward big..." -- I agree. But it's "big product scale", not "big organization".
staunch 15y ago
At first you want a small hard core team of dedicated people. They all have meaningful equity and are "the founding team". Maybe 2 co-founders and 2-5 employees that work really well together. They're like a special ops team that can quickly maneuver around a problem. They can probably do just fine with 12 months runway and a low burn rate. If they do figure out a repeatable/scalable business, they can raise a lot more money and grow it. If they don't, they can shut the company down. The problem with Mahalo is that they skipped the first stage completely. Calacanis was "hot" and (like everyone) he thought his idea was a sure thing. He raised $20M because he could. Even so, if he had created a small team and spent a year or two iterating it might have worked very well. With no pressure of running out of money it might have been harder to keep the the team focused, but it would have probably been fine. Hell, the company could have been breakeven off the interest of $20M. Instead, they started scaling the business from day one and never really had that hard core team to perfect the processes, product, and business model. Even if he had wanted to, the VCs would have never let him keep the company to 5-10 people for the year or two it might have taken. Due to the perverse incentives of raising that much money so early he's actually "forced" to make a bluff bet of the whole company rather than build it methodically. The first bet was made by scaling the company before they figured out the business. The second is an even more extreme version of the first. After it fails there will not be a third. tl;dr: Mahalo will fail because it raised too much money too early.
- jasonmcalacanis 15y ago
  Actually, we can't fail now.... we have a lot a revenue and 30M+ users a month. I don't thin it was a lack of focus that caused problems for us. The main problem for us was that building a search results that is--sometimes--20% better than Google's does not incentivize people to switch. This is a problem Wikia Search, Blekko, Bing, DuckDuckGo and countless others have faced or are facing. Although it's easy to change your search engine technically, it's not necessary unless your search engine is failing. The problem is that innovators look at the times when Google fails (i.e. when you see 2-3 horrible results in a SERP), but not when it succeeds (the other 7-8 search results). We took a really hard run at making better results and we proved we could make something incrementally better. In our business making something slightly better is, essentially, failure because it will not cause folks to switch. Now, in terms of educating folks, that's a wide open space. Right now there is no wikipedia, google or facebook of learning. There's Lynda for software, Rossetta Stone for language, Khan Academy for math and bunch of individuals teaching one-off things on YouTube. We are making a brand that runs from how to do your hair, to how to solve angry bird, to how to make a movie poster in photoshop to how to make a vegan shake.... and on and on and on. We have 40k+ videos now and make over 1k a week (with 90 people). We'll get to 3k a week in the next year (with 150-200 people). Then 5-6k videos (with 300-500 people). We're already, we think, the largest providers of orignal (non-UGC) content to YouTube. And we only pivoted a year ago to do learning videos. our ipad apps/courses should really drive traffic too... and forget about it when every TV is internet connected (which really starts this christmas and goes forward for the next 7 years). best j
tom100cities 15y ago
12 months of cash, if it gets less - cut burn or increase revenue

Jason seems bitter about Google's Panda update.

jasonmcalacanis 15y ago
Yes, I am. :-)
- dennisgorelik 15y ago
  1) Your opinion about quality of Mahalo content is obviously biased, so we cannot really rely on that. 2) Isn't it too expensive "burning $500K a month" just to find out what idea works? Discovery phase should be lean, right?
  - jasonmcalacanis 15y ago
    I'll start with your second question: In some categories lean can be a big number: search, making chips and (today) building a social network. Can you start with 10 people and 150k a month burn in these categories? Sure. However, if you're going to take a real stab at this type of problems in a massively competitive market with HUGE incumbents you need 30, 40 or 50 people. Also, "too expensive" is relative. Spending $500k a month might seem too expensive to someone starting out as an entrepreneur, but to VCs and founder who have been in the game for a long time they would rather risk blowing out 10m and figuring out a lot of problems quickly, than sit around and slowing figuring stuff out. The whole industry is biased toward big success. Just the nature of the beast. If you want to do something more modest--and i did on my last two companies--you raise small amounts of money and shoot for a small target. Neither model is wrong. In terms of #1, you can go check out youtube channels yourself. We basically stopped writing pages with great writers (but not experts) to creating videos with experts in them. This costs 10x as much, but the result is stunning. Not sure if you're into styling your hair, but we have one of the best stylysts in Los Angeles doing a series--think of it as a Khan academy course--on styling your hair. http://www.mahalo.com/how-to-style-your-hair/how-to-do-a-perfect-salon-blowout-at-home/ http://www.mahalo.com/how-to-style-your-hair/how-to-do-a-per... or photoshop tutorials like Lynda--except free: http://www.mahalo.com/learn-adobe-photoshop-introduction/ http://www.mahalo.com/learn-adobe-photoshop-introduction/ http://www.mahalo.com/photoshop-video-tutorial-how-to-make-a-movie-poster/ http://www.mahalo.com/photoshop-video-tutorial-how-to-make-a... Now imagine those not only as courses on mahalo, but also iPad apps. And imagine 1M videos with that level of instruction. That's what we're doing. ... and that all came from my vision of search results + content. Now, I still believe in that vision, but the truth is Google went from ten blue links in 2007 to what it is today: content + answers + some search links. we knew that was possible--that they would wake up--and they did. c'est la vie!
    - dennisgorelik 15y ago
      When I'm thinking about lean discovery, $150K/month seems too expensive too. Initial version of Google Search was developed by team of two or three people. That's less than $30K/month burn. "Neither model is wrong." -- I agree. However "lean model" is much more common and relevant to majority of founders than "$150K+/month burn during discovery phase". Another consideration: you made your previous company profitable with lean approach. But you did not make Mahalo profitable with "$500K monthly burn" approach. That should teach us something. "The whole industry is biased toward big success." - I assume under "whole industry" you imply "Venture Capital industry", not "Information Technology industry", right?
      - jasonmcalacanis 15y ago
        1. The Google example is only relevant in 1999, when the market was nascent. Now that you have a big, established product there is--in all likelihood--no way to create a google killer at $30k a month. If there was one at that level, it probably would have been found by now. Established markets take bigger checks... nascent markets take smaller checks. 2. I actually did get Mahalo to profitability last December. :-) Of course, the Panda Update endend that brief, beautiful moment. 3. Actually, I do mean the entire industry. Now that the internet is global, fast and ubiquitous (mobile) everything good tends to get really big, really fast. Heck, bad things get big fast (Rebecca Black's Friday comes to mind), and so do super simple things (i.e. Angry Birds). the internet is biased toward big...
        dennisgorelik 15y ago
        1) While Google is efficient, there is no market need to create Google killer. Even if there is such a need, the goal of startup would be to create a model that at scale would overperform Google. Such model is likely to be created at less than $30K monthly burn. For example, team of two developers in couple of years may implement and fine tune some efficient social way of ranking search results. $30K/month * 24 month = $720K. 2) Jason Calacanis empirical rule: "if startup is overfunded then as soon as profitability is achieved it would be crushed by Google." :-) 3) "The internet is biased toward big..." -- I agree. But it's "big product scale", not "big organization".

At first you want a small hard core team of dedicated people. They all have meaningful equity and are "the founding team". Maybe 2 co-founders and 2-5 employees that work really well together. They're like a special ops team that can quickly maneuver around a problem. They can probably do just fine with 12 months runway and a low burn rate. If they do figure out a repeatable/scalable business, they can raise a lot more money and grow it. If they don't, they can shut the company down. The problem with Mahalo is that they skipped the first stage completely. Calacanis was "hot" and (like everyone) he thought his idea was a sure thing. He raised $20M because he could. Even so, if he had created a small team and spent a year or two iterating it might have worked very well. With no pressure of running out of money it might have been harder to keep the the team focused, but it would have probably been fine. Hell, the company could have been breakeven off the interest of $20M. Instead, they started scaling the business from day one and never really had that hard core team to perfect the processes, product, and business model. Even if he had wanted to, the VCs would have never let him keep the company to 5-10 people for the year or two it might have taken. Due to the perverse incentives of raising that much money so early he's actually "forced" to make a bluff bet of the whole company rather than build it methodically. The first bet was made by scaling the company before they figured out the business. The second is an even more extreme version of the first. After it fails there will not be a third. tl;dr: Mahalo will fail because it raised too much money too early.

jasonmcalacanis 15y ago
Actually, we can't fail now.... we have a lot a revenue and 30M+ users a month. I don't thin it was a lack of focus that caused problems for us. The main problem for us was that building a search results that is--sometimes--20% better than Google's does not incentivize people to switch. This is a problem Wikia Search, Blekko, Bing, DuckDuckGo and countless others have faced or are facing. Although it's easy to change your search engine technically, it's not necessary unless your search engine is failing. The problem is that innovators look at the times when Google fails (i.e. when you see 2-3 horrible results in a SERP), but not when it succeeds (the other 7-8 search results). We took a really hard run at making better results and we proved we could make something incrementally better. In our business making something slightly better is, essentially, failure because it will not cause folks to switch. Now, in terms of educating folks, that's a wide open space. Right now there is no wikipedia, google or facebook of learning. There's Lynda for software, Rossetta Stone for language, Khan Academy for math and bunch of individuals teaching one-off things on YouTube. We are making a brand that runs from how to do your hair, to how to solve angry bird, to how to make a movie poster in photoshop to how to make a vegan shake.... and on and on and on. We have 40k+ videos now and make over 1k a week (with 90 people). We'll get to 3k a week in the next year (with 150-200 people). Then 5-6k videos (with 300-500 people). We're already, we think, the largest providers of orignal (non-UGC) content to YouTube. And we only pivoted a year ago to do learning videos. our ipad apps/courses should really drive traffic too... and forget about it when every TV is internet connected (which really starts this christmas and goes forward for the next 7 years). best j

tom100cities 15y ago

12 months of cash, if it gets less - cut burn or increase revenue