How to Build a Web App from A to Z
Thursday, February 28th, 2008These are the notes I took in the awesome session from Kevin Hale of http://www.wufoo.com/ at the Future of Web Apps Conference
Any misspellings or incompleteness is my fault, and not that of the presenter.
Strengths and Weaknesses
- talking about his experiences/background
- investing, etc, wufoo
Startups
- Advice = Limited Life Experience + Over Generalization
Who? When? Where? Why? How?
- there is no “what”, these other things are more important
Workshop
- write down your questions and ask in the second half
About wufoo
- writes for treehouse and particletree website/magazine
- co-founder of Infinity Box, Inc. (Wufoo)
- 6 million page views per month
Who are the wufoo customers
- healthcare, individuals, many industries, as well as Microsoft, sony and other big names
- do all programming with 3 people, and that’s it
Story of Wufoo
- graduated with digital arts degree, english degree
- his aspirations were to take a year off, and then go get a graduate degree
- got a job in a research office (Information Specialist) at USF
- Where do you find money to do research?
- Met Chris Campbell at USF
- Met Ryan Campbell (TA for a computer science teacher)
- went to a session at SXSW and with Jason Fried of 37 signals
- made a decision to start a company at that point
- first tips: build an audience first with a blog
- made web development blog
- “single greatest thing we ever did”
- vehicle for learning these techniques would need
- Particle Tree: web dev tutorials
- inverse pyramid, most important thing at top, least at the bottom
- learned a lot about writing, programming, etc.
- thought about building a content manager (beginning of wufoo)
- came across Ycombinator
- incorporated Particletree Inc. (cost about $565)
- setup as florida S-Corp
- got contacted by CNET via an article of particletree getting picked up on slashdot
- Ryan decides to quit his job, they start talking about it. Chris splits paycheck with the three of them
- started out on 12 inch screens, terrible equipment
- gave first magazine of treehouse for free (100,000 downloads)
- Ycombinator puts up “startup school” and they apply and attend this in boston (fuck, I applied for this and got in, but couldn’t fly in)
- ycomb contacts them and likes their app a lot, they go to boston to pitch idea
Worst Interview Evar
- ycomb tells them they need a form builder
- they argued with ycomb on what their product was
- 16 teams went out there, only 8 get selected
- 10:00pm supposed to get call, got call at 11:15pm to say they were accepted
We’re Going to California
- $6,000 per each person
- rented apt, furniture
- pics of apt, and setup was horrible
- they become incorporated (In Delaware specifically, it’s tax beneficial) = Infinity Box, Inc.
- had dinner nights, still have relationships with all those groups in ycomb
- ycomb invests in 20 teams at a time now
- all the ideas of startups are so different, the collaboration between teams is awesome
- your idea doesn’t matter, what you should focus on is the users/customers. (golden rule in effect)
Wufoo
- real story: used http://instantdomainsearch.com/ (knew the guy who made that app) and got list of 5 letter domain names, and chose “wufoo” based on it’s memorable
- “girlfriend test”, made her build forms using their simple version (design) of wufoo.
- realized where things were complicated, etc. where to fix things
- put out a demo (5,000 people wanted to sign up) immediately after that
- you have a 10 minute pitch on what your product does; many people were interested in it
- accepted 50,000 from 2 different angels (figured that it would last them about a year)
- they didn’t stay, went back to Tampa (out of money)
- for the next three months they all live together with chris and fiance in their townhouse
- contacts mike arrington to see if he’d like to try it; promised exclusive of when it will launch
- launched July 5th; techcrunch gives rave reviews
- Chris becomes customer evangelist; he answered every support email
- take care of the beta customers first; and then it will benefit you
A Note on Beta Testers
- almost no one gives them feedback
- those that do give feedback wufoo gives them a deal for 50% the lifetime of the account
- gave them this deal a week before launch (so they could test out the billing system for a week)
- beta testers were not the “real” users, beta testers are just interested in the next big thing
- surveying beta testers is probably not a good idea, you can’t understand what your revenue model may be from that
- the models out there: flickr, delicious, etc. aren’t necessarily the “right” way to run a company
- the myth is that you can only start a company in california
- things learned in exercise:
- each company has been around about 3 years
- you can make a shit-ton of money regardless of the niche you are in
- location doesn’t matter
The Importance of Time
- start your company now!
- you got to have patience
- relationship and customers the most important
- Inc., 5000.: http://www.inc.com/inc5000/
You Make What You Measure
- guy of jotspot says this (just sold to google as “google sites”)
- wufoo stats
- monthly subscription rev
- increase rev
- month # users
- customers/ month etc
- “stumbling on happiness” = humans are the worst at predicting things, they are the worst at this
Chart of Destiny
- exercise:
- Q: Which is better 1000/month or 1000 + 10%/month?
- A: Over time the 1000/month is better.
- conclusion: over time, you make more, start now, and be patient
Q: When? A: Now
Bootstrapping vs. Raising Money
- bootstrapping can be done
- trade off is control
- ways to raise money:
- seed investors
- angels
- 50k - 100k
- venture capitalists
- up to 3 mill = they want part of your company
- family/friends
- business loan
- gov’t grant
- first person you talk to is an associate, not a partner (venture caps); they are no one, and have no power
Top 10 Questions Investors Asked Us
Article here: http://particletree.com/notebook/the-top-10-questions-investors-asked-us/
- Who else have you spoken to?
- How will you make money?
- How will your company grow?
- What technologies do you use?
- How easily can you be copied?
- Can we see the demo?
- Who are your competitors and how are you better/different?
- Who are your customers?
- How will you spread the word about your product?
- What will your market penetration be?
- this can be one of the weakest points for some companies
TAM = Total Available Market
- the piece of the market you’re going to be catering to
- doesn’t freakin’ matter, impossible to predict
- don’t be afraid to say “I don’t know” if it’s reasonable to do so
Resources vs Resourcefulness
Team
- Two most important things in considering for a team mate
- passion
- animal
- like a marriage
- trust
-
where?: family, friends, school, work
-
How?
- Do stressful test projects
- practice fighting
- don’t make it personal
- sleep on it
Let the Data Decide
- track things, and the data should show you the right path
- which person is more correct?; with data to back it up
2 Weeks
- do anything you want (give that flexibility to anyone)
- you have two weeks
- you can measure it
Chris Kevin Ryan
- Bitpusher = manages their servers
- CPA for the accounting stuff (every quarter and once a year)
- Lawyers
- Users (mentions book “The Ultimate Question”: http://www.amazon.com/exec/obidos/ASIN/1591397839/bookstorenow600-20)
- take our feedback based on which they are
- mentions another book “From Good to Great”: http://www.amazon.com/exec/obidos/ASIN/1591397839/bookstorenow600-20
- Customers
Profit per X
- you find the thing that drives your profit
- then maximize that profit
- for wufoo:
- stay as small as possible
- profit per employee
Customer Support
- doesn’t scale
- you cannot outsource this; have to do it on your own
- maximize your time to make customer service great
How to be a Market Leader
- Three Ways
- your product is based on the best price (price wars)
- very hard to do as a startup
- I have the best product
- Best overall value (”Small Giants” book mentioned: http://www.amazon.com/Small-Giants-Companies-Choose-Instead/dp/1591840937)
- under promise, over-deliver
- conversion rate goes up, because their trust goes way up after customer support contact
How Word of Mouth Works
- give them a great customer experience
Marketing is what you do when you didn’t make something remarkable
- Would you recommend me to a friend or coworker?
Wufoo Support
- only 3 people support 60,000 users
- great customer support doesn’t scale easily
- just as much time on support, as working on the application
- do everything in gmail (support@wufoo.com)
- label with “C”, “K”, “R” with who should answer it
- assigned different days to answer support: all day, from when you wake up, until you go to bed
- built tools to help them manage accounts, things that users never see
- wiki, forums, screen casts HD!, FAQs, knowledge base
- they use TextExpander to help with support emails
- promote your documentation in your support emails
- use google search for their documentation (Google Custom Search) ($100/year if you don’t want any ads, and they provide a lot of tools on statistics, what’s being searched, etc.)
- hand-wrote christmas cards to customers, made postcards, and hand-wrote those too. (wow!)
Productivity Experiments
Paid Incentives
- you meet certain quota on writing for particle tree, and you get money
- if you don’t meet it, then you don’t get the extra money
List of 10
- every 6 months, what are our major goals
- 15 things you’d have to do to make that happen
- first person to finish at least 10 things, would choose the company trip
- last person would be the “Trip Bitch” ha ha
Scrum
- everyday have a quick meeting
- what did you accomplish yesterday
- what are some things in your way and how can I get them out of the way
- write it all down, accountability
- wufoo
- they do this once a week (video chat)
- highly recommend this approach; what tasks need to be done
Behavioral Economics
- Dane Ariely “Practically Irrational”: http://www.amazon.com/Small-Giants-Companies-Choose-Instead/dp/1591840937
- Efficiency vs Efficacy
- Chocolate / Intangibles. How do you quantify? How do you price quality?
- SSN
- take the last 2 digits of your SSN
- take a list of items and write this number next to each item
- consider that number the price of the item
- wager to see how much items: correlation between high number and what they bet
- once you get a price you usually don’t stray to far from what the price should be
- price is placed within context and that matters
- so when you price your app, place it in context
- free plan, 9/month, 69/month, 100/month
- don’t doubt putting those higher prices out there
- do tiered pricing
- candy
- the “free” plan is very very enticing; beware of it
Document Your Story
- talk to everyone
- remember the golden rule
- wufoo:
- took a lot of screenshots
- lots of pictures
-
one guy kept a journal
-
you can do a startup and have fun; don’t take yourself that seriously
- don’t be in it for the money
Bibliography
- Small Giants: why companies choose to be great instead of big
- Blue Ocean Strategy
- Made to Stick
- The Best of Software Writing
- Prioritizing Usability (get this one)
- CSS Mastery
- Landing Page Optimization (SEO book, really good)
- Game Design Workshop (user experience, fun, and engagement)
- Harvard Business Review (only magazine they get: get subscription)
Questions
- how was ecommerce and setting that up?
- Answer: Sucks. Don’t have recommendation for any particular payment systems. Built their own recurring billing. Documentation sucks.
- they are considering these guys: http://www.paysimple.com/
-
still developing this service, may not be ready for months. wufoo is working with them
-
would you consider taking venture capital when getting the contacts?
-
Answer: Taking the seed money was so worth it, just in the contacts you make. Invaluable.
-
be aware of people that are doing angel investing for the first time. Get someone who has done it for awhile. Talk with the people the angel investor in.
-
Outsource your IT?
-
Answer: No one will be good as hiring your own. Set up a three stage push system, they helped them set it up. They are really good than working with them. They find it to be completely worth it in paying a little more, and not having to do everything yourself. They use Bitpusher.
-
Outsourcing your code?
-
Answer: They have two people they are considering hiring, they gave them some projects to work.
-
How did you initially figure out the pricing plan?
-
Answer: Copied jotspots pricing plan. They researched their competitors. Haven’t raised their prices.
-
Should invest in some kind of marketing or PR?
-
Answer: Focus on your product in the first 1 -2 years. Then transition to marketing and advertising, that’s what they’re doing right now.
-
What would you do differently if you could start over?
-
Answer: Start earlier! Regarding legal stuff: No reason you should pay a lot of money for ToS or Privacy Policy. Go copy ours! Be aware of where you ask for legal advice.
-
Have you leveraged blogs to promote wufoo?
-
Answer: Yes. Sold some web dev ads at first, but then found out that all conversions in promoting wufoo worked much much better.
-
Build audience first?
-
Answer: Yes. Blog worked out very well. Joined Web 2.0 workgroup.
-
How much time goes into research (reading blogs, etc)? Percentage of time goes into this?
-
Answer: Fridays they hang out. Share books all the time. Reads 3 books a week.
-
How often new features come out?
-
Answer: Wrote the form builder 7 times. Rolls out new versions to new users about 1000s at a time. See how they react.
-
Feature requests?
-
Answer: Feedback form. Don’t let people know what you’re timeline is, setting them up for disappointment.
-
Work in same location?
-
Answer: Work separately, remotely from home. Only have monday video conferencing. Always have instant messenger open for real-time communication. Always have phones on talking about it.
-
How does workflow go?
-
Answer: Design driven development. Design gets taken to have the data serve the data, then the JS/interactivity. If you see something cool in front of you, then you’re more motivated to get it going. Kevin does all the HTML/CSS. Don’t make a lot of spec requirements, just get a design, and then get the work done.
-
Best technology in 5 years?
- Answer: Ha! I said humans are bad at predictions. HTML/forms hasn’t changed in 10 years. For form building nothing much, hasn’t really changed.