Tuesday, July 16, 2013

Startups: Legal, Pricing, and Marketing Issues

This post is a recap of what I have learned in the Global Founders' Skills Accelerator I am participating in with my startup, NarwhalEDU (narwhaledu.com) over the last week. I am writing this up in the hopes that it will be useful to other new startup founders as a reference point. Some things are only relevant if you are open sourcing, if you have more than one founder, and other caveats of being the experience of a single person in a single company at this accelerator.

Lawyers
Founder's Agreement and Equity Split:
Determined by factors such as replaceability of founder (e.g. technical knowhow), how much a founder is investing in the company monetarily, basically a really fluid of trying to determine how much a founder contributes to the company over time.
Vesting schedule -- so that the equity / shares are earned over time, in case one founder decides to leave. Recommend 3 years, although investors will push for 4 years
1 year cliff -- in case a founder leaves early on, their shares do not vest

C Corp vs LLC
C Corp in Delaware -- recommended because that is what investors expect
vs. LLC -- for tax purposes, can offset business losses against personal income for tax gains, but investors will not accept it if you ever take outside investment (so more suitable for lifestyle businesses, e.g. never plan on taking outside investment, not really aiming for exponential growth, longer-term are all used as criteria but I think the main one is not planning on taking outside investment.)
It is much easier to go from LLC to C Corp (about 5k costs) than the other way around

Fees
We've gotten quotes ranging from \$2000 to \$5000 for the standard incorporation and founder's agreement and a set of template documents. There is an additional 250-500 in filing fees. e.g. if you do a Delaware C Corp every year there is \$500 in fees to have a registered agent there.

Packages
Some firms offer this, some offer "flexibility" and haven't started using the term package yet. But basically, firms will offer anywhere from full deferment of fees until you are "successful" (e.g. have raised a major round, are making over a million in sales, etc.) to 50/50 split (so pay 50% upfront, keeping you honest about controlling your spending) to a flat fee. So in a sense you are pitching to the law firms that offer packages.

How it works
What tends to happen is you will interview with the super expensive lawyer (ranging from \$250 - \$800+), then depending on your company you will get a team of 2 to 3 people: a junior associate (in the \$500 range) and paralegals / secretaries (in the ~\$100-300 range). These people charge less per hour and so the law firm tends to pitch this as being cost efficient.

Why a lawyer when there is legalzoom.com?
We asked one lawyer this question. She said that lawyers can provide business and legal advice, especially after working with startups for over a decade. When there is more than one founder, make sure that everything is agreed to properly. In general, it seems to be the thing to do at the accelerator so we are running with it, although I suspect there is a bias towards "startup" and not "lifestyle business" that a legalzoom LLC might be appropriate for.

Why incorporate so soon?
Put your IP into a company instead of into individual hands, and then can decide to open source it or license it however you want. Prevents one founder from forming their own competing company. (?? still a bit confused as to how this applies to open source hardware).

Open Sourcing and EdTech
http://stallman.org/articles/online-education.html The end conclusion is to NOT use a creative commons noncommercial use only license.

Capitalization
After incorporating, have to put some money into business to make it an actual business. Within 30 days, need to fill out 83B form.

Pricing
Often engineers go for cost-based pricing. Not the best idea. Consider instead the amount of value provided to the customer and getting an appropriate fraction of that. In particular, I have heard that in software the most money comes from subscription-based models. This is because the customer does not know the full value of the software initially and actually gains a lot of value over time, and the subscription model captures more of this value than an upfront pricing model.

Pricing! Okay, let's go into how to do pricing surveys. In an ideal world, you do randomized questions with at least 30 responses each.
DO NOT ask the customer "how much would you pay for this product: \$100, \$150, \$200?" because that is not how we make decisions in real-life -- this feels like bartering, so the customer's answers will be influenced by their bartering. Instead, randomly present them with one choice and ask them a simple yes/no.

To do this, I used qualtrics software, a trial version.




Then, to get survey responses, I used facebook.



The going rate seems to be about \$1 per survey answer.
I put the max bid at 50c per click.
Click through rate: ~1%, e.g. if 1000 people see your ad, ten people will click through. Of that, maybe 2 people will actually answer the survey (that would be pretty ideal).

I had this 1 dollar/answer my first and second days but on the current day (day 3) it seems to have shot up to almost 3 dollars per answer ;__; hopefully the next few days will even this out, or perhaps weekends are just better for people clicking through and deciding to actually answer the survey.

This is called the "taxi meter" effect, something you have to beware of in pricing, that the customer is not constantly focused on whether they are getting the value they paid for out of your product (e.g. if you priced weekly or monthly instead of yearly).

Marketing
Pick up the phone and call!
Have the whole company focused on the same metric -- something everyone can work towards. e.g. number of sales
Always be helpful, e.g. even if you are not immediately knowledgeable on a topic, you can answer forum posts / emails with a bit of google searching.
Should be more time than money initially, as you do labor-intensive one-on-one direct sales.
Personalization of emails is important (e.g. with their name) (for mailing lists, use mailchimp)
Other recommended software: HARO, grasshoper, Highrise CRM, google analytics, mailchimp
Testimonials are important! If they are above the fold, can lead to 10-15% increase in conversions.
Start a blog right away! Short posts just to get over the hurdle of getting started. Basically free marketing by establishing yourself as an expert right away.

Conclusion
There's lots more, but I think that's a good amount of information for one post. Please feel free to contact me or leave a comment if you have any questions.

Oh! One thing -- if you're affiliated with MIT, sign up for the MIT Venturing Mentoring Service. They will offer free legal advice sessions (make sure to sign up for the mailing list) and a whole host of other benefits (e.g. matching you up with mentors and you set the schedule as to when you meet).

Sunday, July 14, 2013

trip to nyc / adafruit industries / founder of hackaday!

trip to nyc / visit adafruit industries / meet founder of hackaday and talk to ladyada herself!

with charles guan and cynthia lu and hanna lin
10 am: depart
2 pm: arrive
left to right: charles, cynthia, and hanna
arrive in CHINA wait sorry flushing district of New York City

look what we found in chinatown! an AVR manual IN CHINESE. @___@

and a MasterCAM X book. This was at the WJ bookstore
we ate at Taste of Shanghai on prince street. delicciousss
Cynthia sketches ladyada chibi! Ladyada was super happy about this when she got it.
Photo
this picture from Charles! Not mine!
We arrive!
The magical workings of adafruit heavy metal industries. Err. Adafruit industries.

$150k+ pick and place machine! IT IS SO SHINY. Apparently it is or is a relative of the ones Samsung uses to manufacture phones, and that this is something Foxconn / Apple would never do, sell you the machines that make machines at a conference
Assembly station! At adafruit industries they do it by product and not as a procedural task with multiple people per product. The guy is Philip Torrone, founder of hackaday and editor at large at MAKE magazine while he works with adafruit on everything awesome.
IT'S BINS OF AWESOME BADGES.
The ladyada shrine :)

Hydraulic wooden arm spotted! d'aww
And finally, on the air on ask-an-engineer! We are sitting in the background, as Charles is the main guest (we sort of surprise showed up). Left: Phil; Middle: Ladyada; Right: Charles
The episode is archived at
http://www.adafruit.com/blog/2013/07/14/ask-an-engineer-references-equals-zero-by-charles-guan/ and
http://www.adafruit.com/blog/2013/07/13/show-and-tell-7132013-e-ink-badges-flora-coats-laser-asteroids-twitter-api-neopixel-code-and-on-air-notifier/

Phil and Ladyada were amazingly open about everything we could ever possibly want to know. I learned that:

  • adafruit industries took on no outside investment and was entirely bootstrapped
  • it started in just one apartment, then two apartments, and finally they moved to a warehouse
  • their first pick 'n place was picked because it was the only one that fit through their apartment door, and was $35k. they are now donating this to nycresistor with the condition that no questions come their way about how to use it or anything
  • after 8 years, they are at 50 employees, 1/3 in shipping, and at 10-15 million dollars in revenue and they are tripling every year. 
  • They wrote their barcode/shipping software in python and ship ~1000 packages a day. The software notifies the buyer and charges their credit card when items ship 
  • In fact, they are the largest shipper in lower Manhattan.
  • they grew specifically so that ladyada could do all the engineering, and it sounds like they have a few consultants from around the world that help with the engineering but ladyada essentially solo beasts all the products (it sounded like)
  • they chose to keep all manufacturing in house instead of outsourced
  • they used a black and decker IR oven, equipped with arduino, for reflow for a long time. In the new space, with the real reflow oven and new pick 'n place, they are finally able to keep up with demand
  • Their latest video show has 3 year-olds making lemon batteries o____O Each 2-3 minute video takes over 80 hours to produce. Phil is the one manipulating the giant plushie LEDs / figures under the table
  • they are thinking about adding translations to learn.adafruit.com, hopefully crowdsourced, as there is a lot of demand from germany, japan, china, and italy
  • adafruit.com gets about 11 million visits (either daily or monthly) after 8 years of relentless quality documentation
  • the new move from docuwiki to their own system has cut down documentation time by about a third
  • but each tutorial still takes anywhere from 6 hours to 2 days
  • Their PCB stencils are done by hand. They found that people can do it better when trained and get over 99% yield, such that they almost think they don't need an optical checker
  • Testing rigs for circuitboards! Ladyada actually designs products with testing in mind
  • They worked with companies like Eagle to make girl scout/boy scout-esque badges, and Eagle was very happy about it
  • They helped Jay Silver with MakeyMakey, who really just needed encouragement that a market exists
  • Ladyada did not market ressearch nor business plan. She believed in the product strongly enough that she created a market / demand for it. Execute.
  • ladyada reminisced about her time at MITERS and how it used to very much by Tim Anderson's shop, and also about her Bridgeport mill which is still in active use at MITERS
ALSO. I GOT A 555 PLUSHIE. d'awwwww it is an octopod ^__^ heart cousin of hexapods

And a final note on fail.

I wanted to draw ladyada's face with our robot arm but sadly we were not at a stage to make a recognizable drawing and I think ladyada was just like, WTF. ;__; SORRY LADYADA one day we will draw better faces with our el cheapo 9g miniservos and then give you a portrait. we probably should have stuck to the adafruit logo or something.

edge detected from the WIRED magazine cover
draw out... yea... it does not resemble ladyada at all x___x

http://youtu.be/q73vKiPvaPA?t=59s
Ah well. So it goes, so it goes.

That's all for tonight folks! Getting ready to teach Intro to Robot Arms, class #2, in a few hours (for MIT HSSP. Let me know if you all have any questions.

Oh, an easy one -- I don't know how this came about. I think Charles emailed Ladyada saying he would be in NYC and wanted to visit, and he was then invited as a guest onto the show. He was going to by in NYC for a mikuvan trip including us, so we tagged along to visit adafruit industries. All quite strange and befuddling.

I finished a hexapod instructables, get featured, and then mope some more

I finished my second instructables ever! This one was a lot more "successful" than my first one -- it was featured on the front page for a day or two. :D yay!

This is the instructables for my 18 degree of freedom (18 servo) hexapod:
www.instructables.com/id/Simple-18dof-Hexapod-Arduino-nano-optionally-wit/



It probably took slightly less time than my 7 minute video on the design process (which took at least 2 or 3 full days) yet garnered about 10x the views. The little star in the upper right hand corner stands for featured.


All that it means is that your instructables gets put on the front page for a little while. At least it did much better than CNC nyancake. Apparently nyancat is a thing of the past for everyone except me XD;

day 1
day 2
It was featured within hours of posting, crazy moderators.
instructables published jun 24th

It is gratifying to see that at least some of the thousands of people who visited that instructables clicked through to my took-ages-to-make design process of a hexapod video. See spike in traffic at the end of June.

I still don't know the secret to getting comments though. I want to interact with people virtually! At least for now.

As usual -- then I mope about how I focus too much on these sorts of stats instead of going out and "enjoying what I do" or whatever.

But yea, all in all. Four years ago I never would have imagined being on hackaday and having a featured instructables. These were all things I saw the peers I looked up to do, not myself. Yet I find myself qualifying these accomplishments -- it was just for a project I basically copied off the internet, it is just for a really derpy hexapod I never really finished, etc. I have an awesome friend who was published in science as an undergraduate, yet she always qualifies her publication, and it's obvious that she's missing how amazing this whole thing is. Maybe I am doing that?

Lack of self-confidence is unattractive and can make other people difficult to work / high maintenance with, yet I can't get rid of this in myself. Grr! At least there is hope for me. I can reasonable list three things every day that I am proud of myself for, unlike some friends of mine. x___x must spread positive energy

Well, that's life in the first world. I am in an amazing spot for myself currently -- working on my own startup with two very good friends who are still my friends so far, assisting with a go-kart class for pay, no financial issues for at least a few months -- yet I still feel not-legit, like I haven't really built anything really cool or robust. I'm not sure what it'll take, since I certainly won't catch up anytime soon to people I look up to in the areas they specialize in.

I guess that is just something I will have to get used to. Or I could finish kiwikart...