When it comes to testing a startup, how good is “good enough”?

May 5, 2010

The new rules of entrepreneurship favor speed, flexibility, and capital efficiency. As a founder, that means you should get a product into the hands of your customers, test their reactions, and then raise outside capital if necessary.

To get a product to market without raising a lot of money, you’ll have to bootstrap (and/or get loans or investments from friends and family). Either way, you’ll want to spend as little as possible on product development. If you try to incorporate 100% of the features you’ve envisioned on a tight timeframe and budget, you’ll probably end up with a fully-featured but craptastic site. Instead, build a stripped-down prototype, with only those features you think will be critical for your test. That’s often called a “minimum viable product”.

But how good is good enough? In other words, what does your product need to include in order to conduct a test that will let you determine whether to move forward, pivot into a new direction, or (gulp) trash your idea and go back to the drawing board?

A good rule of thumb is to start by working backwards. What results would convince you to move forward? If you need to persuade investors to get on board to do that, what results will they need to see before writing checks? How can you quantify those results so you can set specific thresholds?

Here are a few examples:

  • If you’ll be selling through retailers, such as boutiques, department stores or supermarkets, get meetings with X buyers, and see how many will sell your products on a test basis. Figure out how many total stores and / or customers you can reach through these tests, and then set sales goals. That way you can tell yourselves, or your investors, that once you have money to produce your initial inventory, you’ll be able to ship it to Y number of stores and measure sales results within some particular time-frame.
  • If you’ll be selling directly to customers online, built in the capability to measure conversion rates. For example, how many people came to your home page? How many of those people signed up for your offer, or made a purchase? You may start with relatively small numbers of visitors since you are cash-strapped, but the conversion percentages can help you measure the degree to which customers find your offerings attractive enough to make your business economically viable.
  • If you are selling directly to customers offline, look for ways to get payment in advance. If you can’t get checks, get letters of intent or other written commitments. Even just emails from a few big customers saying they’ll buy $X amount when your product is ready can go a long way.

Got advice about how to determine what’s really necessary for an effective product test? Please share…