How do you validate your startup idea and start gaining early traction, without spending a ton of time and money building software? Delay complexity as long as possible by doing things manually. “Fake it till you make it.”
Below are some examples of how successful companies smoke / ghetto / wizard of oz tested their ideas:
In 2000 it wasn’t clear that purchasing shoes online was a great business. Shoes have different fits and comfort levels and many thought they needed to be worn before they were sold. The riskiest question for founder Nick Swinmurn’s fledgling company was: “Will people buy shoes online?” One way to answer this would be to go out and raise millions of dollars in financing, build a huge warehouse, fill it up with shoes, build a comprehensive ecommerce system, hire a bunch of folks, cross your fingers and pray people placed orders.
Nick realized there has to be an easier way to de-risk his business. Instead he went to Foot Locker and took pictures of their inventory. He put photos of the shoes online. Every time someone placed an order, he went to Foot Locker purchased the shoes and mailed them to the buyer.
Is this model scalable? Nope. Did he make money on each order? Nuh uh. Was he able to prove that people would by shoes online and get some early traction? Yes. (source: The Lean Startup)
In 2003, Marc Cenedella saw the difficulty executives were having finding new jobs online. Wouldn’t it be great if there was one place that listed only high end (in this case $100k+) jobs? Is that a service that job seekers would pay for?
Marc’s initial “prototype” involved going out once a week and browsing HotJobs, Monster and other job boards to manually collect $100k+ jobs. On Monday mornings he would send a newsletter to job seekers containing only these 6-figure jobs. He charged $25 to subscribe to the newsletter and his audience quickly grew through craigslist postings and word of mouth.
After we had been doing that for about nine weeks, I missed the 9 a.m. Monday deadline… At around 9:10 I started getting e-mails from people asking where the newsletters were. Those e-mails kept coming. That’s when I really knew I was onto something. That was the moment of validation.
Having passionate users who care when you mess up is an awesome sign of product-market fit. Once a manual approach no longer scales you are in a great position to start automating these pieces with software. (source: Sramana Mitra)
Google is really bad at answering questions like: Where should I grab a drink in SoHo after work? Should I go to business school? Which DSLR camera should I buy? On the other hand, people are really good at answering these questions. Aardvark was started as a social Q&A service. You would send Aardvark a question over instant messager. Aardvark would do some magic and get 3 people to answer your question and send it back over IM.
The most technically complex piece was the algorithm to find the right 3 people to answer your question. The folks at Aardvark realized that although there was a large technical hurdle, the riskiest question to the success of their business was not can we build it, but will people use it. Rather then spending time initially to build out the algorithm they used a Wizard of Oz approach:
Aardvark employees would get the questions from beta test users and route them to users who were online and would have the answer to the question… While it used humans “behind the curtain,” it gained the benefit of learning from all the questions, including how to route the questions and the entire process with users.
(source: Dean Eckles)
Taken from my previous post on: 8 Lessons Learned from Zynga about Virality
In the last 5 minutes of the video below Zynga’s founder Mark Pincus is asked what’s the best way to do market research. His answer – “Ghetto Test”. If someone wants to build, let’s say, a hospital simulator he creates an FB ad that says, “Ever wanted to run your own hospital?” which leads to a survey (or if it’s really ghetto a 404 page).
All Zynga has to do is track CTR and compare it to previous historical rates to get a pretty good idea of demand.
This works great if you are comparing multiple game concepts, product ideas, taglines, names, etc… though isn’t a good fit for testing out a new concept (without a comparison).
The folks at seamless allegedly spent their first few months without a web product. They called up law firms in NYC and asked what they wanted for lunch. They called the restaurant, placed an order, managed delivery and billed the firms at the end of the month. (source: anecdotal)
Andrew Mason and the team at Groupon launched the first version as a wordpress site. Besides posting a deal, everything else was manual.
It was totally ghetto. We would sell t-shirts on the first version of Groupon. We’d say in the [write] up, ‘This t-shirt will come in the color red, size large. If you want a different color or size, email that to us.’ We didn’t have a form to add that stuff… It was enough to prove the concept and show that it was something that people really liked… It got to the point where we’d sell 500 sushi coupons in a day and we’d send 500 PDFs to people with Apple Mail at the same time.
There are obviously situations where a lean approach doesn’t apply (i.e. google the search engine or an iPad). For the large majority of startups today the biggest question is not can we build it, but should we. Before you go ahead and spend a ton of time and money building complex technology, try to manually execute parts of your business for a small target audience. Once things are working and you are unable to scale, then invest in technology to automate.
Any other good examples that we missed?