Thursday, June 3, 2010

Fat Startup: Raise $100 Million For Your Next Internet Company

(Side note - some of the thoughts below are based on conversations with Josh Hannah of Matrix Partners, one of the smarter investors out there).

One of the things I hear quite commonly is that the "old days of the Internet" were totally different from today.  The claim is that multiple distribution channels were wide open (email, SEO, SEM), so it was trivial to build a business without spending any money on customer acquisition.

Although there are many companies that grew without spending a dime acquiring customers (Google, FB, Twitter), when I think back to the original Internet bubble, there were lots of companies that seemed to thrive due to their raising tens of millions of dollars to build brand and acquire users in what was a very noisy environment.  

Indeed, there are a number of companies that were able to scale *because* they spent tens of millions of dollars buying users.

Examples include: 
  • PayPal (who paid $10 per user to join the PayPal network)
  • (remember the superbowl ads?)
  • (went public in 1997 and wasn't profitable until 2001, with costs here were spent on, amongst others things e.g. logistics and market share)
I think there is a fundamental class of business that could be very successful emulating the bubble days approach to building a userbase.  In particular, there are a number of entrenched businesses with strong network effects (e.g. eBay) or brands (e.g. Amazon) or logistical advantages (Amazon).  The lean startup approach may never overturn these apple carts.  The way to challenge these incumbents is to spend a lot of money to build brands or acquire users in order to overcome these effects.

Where To Spend $100 Million?
  • Buy enough of a network to flip it.
    • What are business with network effects/liquidity needs?  Examples of these are eBay & PayPal.  Try and calculate the costs - how many users do you need in a network for the network to "flip" to your product?  If you buy 1 million or 10 million users at $10 each can you create enough of an effect to drive the rest of the network to you?
    • Totally crazy example idea - chose a vertical on eBay and buy all the goods on that state/vertical pair and then resell them at a slightly lower price point on your vertical marketplace and lose $n per item.  How long do you need to do this bulk buying before people start coming to you instead of eBay for this vertical?
    • Crazy example idea 2.  Or, take advantage of media buying/intent buying and outbid everyone in a specific vertical or product category and saturate the vertical with your ads.  You may lose money on every transaction, but in parallel may build enough of a brand and liquidity to sustain yourself long term.
  • Branding.  
    • When Google powered Yahoo! search, Yahoo! lost market share to Google.  But the results from an end user perspective were basically identical.  Calculate the SEM or other media buying/lead/intent buying budget you would need to drive in N users, and then compare that to the brand building exercise and the number of recurring users you would drive in via additional channels with a big budget.  What is the more economical long term strategy?  
I think the whole lean startup movement has some good aspects to it, and when I was working on my startup Mixer Labs (which Twitter acquired in December) we ran very lean and mean.  But I do think too many people "think lean" which often ends up also meaning "think small" these days.

So, I ask you this - who is going to raise $100 million to build the next massive Internet business?

You can follow me on Twitter here.

Startups: When & How to Fire an Employee at a Small Startup

You have spent the last few months hiring a great team and finding people who are a good fit for a startup, but unfortunately even the best hiring strategy may yield members of the team that you will need to let go. In a small startup (e.g. less then 10 people), it is crucial to act quickly to get rid of team members who are not performing or are a bad fit.

If you have a 3 person engineering team, having 1 bad person saps much more then just 1/3 of overall team productivity - they take up your time to manage them more actively, they make other people less productive, and/or you dont give them the hard/crucial jobs out of fear they won't get it done or will argue endlessly about it while making excuses.

Reasons to Fire an Employee at a Startup: (not mutually exclusive)
  • They are not getting done what they need to get done
    • The employee fails to get their work done in a timely or complete manner
    • The code quality is low or shoddy
    • It is not fair to the rest of the team to carry someone who is not contributing equally
    • It is not fair to the employee to work somewhere they are not productive - productivity is sometimes a function of environment, and some people are just not productive in some environments. You should help them find a better home.
  • They are a bad culture fit, whether they are getting things done or not
    • The employee creates a working environment such that you and other team members don't enjoy working with them
    • Thy argue incessantly rather then get stuff done (in a startup, if it takes more time to argue about a feature then to implement it, you should just implement it and try it). They may be religious on technological choices, or just someone very gruff who is hard to work with. This will sap overall team productivity while you burn through your limited cash
    • You and the other team members are walking on eggshells around them, and you worry about asking them to do stuff because it will inevitably result in an issue
  • They are dishonest or unethical. I will not spend more time on this one, but these people should not be given second chances and should be terminated quickly.
Some Signs You May Need To Fire Someone
  • You have spoken with them repeatedly about their issues and they make short bursts of progress right after the talk and then lapse back into more or less the same behaviors
  • "If there is a doubt about a person, there is no doubt you should let them go"
  • You find yourself giving them unimportant or secondary tasks because you worry that they will not come through on (or take too long with) the important stuff
  • You spend more time worrying about the person and managing them then focusing on other key elements of your startup
  • You worry about asking them to do something, because you know they will start an argument with you about the feature, whether it is needed, etc.
  • You and other team members are walking on eggshells around them.
  • They are constantly making excuses about why *this time* things didn't pan out
The "But We Need To Get X Done First" Fallacy
There are some rare instances where you need to wait for some major milestone or release before letting someone go. If the employee is non-productive or unethical, you should let them go as soon as you realize there is a problem (that is not being corrected based on your feedback to the employee.)

Once you let a low performer go, you will frankly feel a wave of relief. You will also notice that even if the team initially reacts in surprise to the person being fired, the overall team will be much more productive. It will drive home how much this one person has been holding everyone else up, and other people will take on (and actually complete) the work that has been stalled.

If you wait too long, the team will form strong relationships or friendships to the underperforming employee, and letting the person go will be even more jarring to the people left behind. You need to act quickly and decisively.

Typically, if you are a first time manager, it will be well past the time you should have taken action that you actually decide to take action. There are always excuses and reasons not to do the hard thing (firing someone is very very non-fun).

How To Do It
  • If this is your first time letting someone go, talk with your lawyer as well as get advice and tips from an experienced member of your startup's network (e.g. angel, advisor, etc.) who was a professional manager before
  • Talk to the employee about what issues you have observed and give them concrete examples of behavior you think should be corrected.
    • On an ongoing basis, you should give your employees feedback on what you think is good/bad about what they are doing and be up front about issues as they occur - so if there is an ongoing performance issue this should not be a surprise
  • Work out a plan to correct this behavior and document it in e.g. an email.
    • You should give them a chance to correct an issue - set a specific timeframe for things to change and specific goals if it is a performance issue
    • You should document the issue to (a) ensure you both have an understanding of what is going on and (b) to protect yourself if you let someone go in case of a lawsuit around the termination (talk to your lawyer about this)
  • If the behavior is not changing, you will need to let the employee go
    • Talk with your lawyer about the documentation needed to terminate an employee
    • You may decide what sort of severance you want to give the employee and whether the severance is tied to e.g. the employee signing certain legal documents associated with their termination
  • If you let the employee go:
    • Be cordial, polite, but firm about the reasons for letting them go. Thank them for their contributions to the company up to that point. This is not personal and you should not make it personal.
    • They may ask for one more final last this-time-it-will-be-different chance to prove themselves. Don't do it. This never works out.
    • I think it is usually best to let someone go the same day rather then e.g. have them stay for 2 weeks to "wrap things up". Having someone who is leaving the company stick around can have a negative impact on the rest of the team. If they were not productive before, there is no reason to think they will be productive suddenly now.
    • Talk with the remaining team about it when the person is gone. Be up front about the reasons the person was let go but do not dwell on details or rat hole. Reassure the team about the process you went through with the employee, and explain why you think this is the right and fair outcome for everyone (including them as the people who were working much harder and getting their work done).
Letting someone go is one of the hardest things you will do. It can be an emotional experience on both sides, but ultimately it will be a key factor in building a great team and executing the startup vision you have put your heart and soul into.

Comments? Any thoughts or tips on the best ways to identify hiring mistakes and to correct them?

You can follow me on Twitter here.