Phil Libin (@plibin, CEO of Evernote), Alexei Erchak (CEO of Luminous Devices) and I did a session with the Boston chapter of the Founder Institute last night. Towards the end of the program we listened to several startup pitches. Every startup was tackling an important problem for a large audience and they all said something along the lines of:

• My target audience is […]. There are […] thousands/millions of these businesses/people in the US.
• Our goal is to have [a few dozen .. a few tens of thousands] of users in the first few months after launch.

In working with startups, I hear this all the time. What I don’t hear enough of is how a company is going to select its very first customers. Unfortunately, that’s because most early stage startups don’t select their early customers. Instead, they are happy to have any customers. They are simply trying to get customers as opposed to select them. In that lies the root of many problems startups have with product/market fit and building scalable go-to-market models.

Let’s set the record straight:

All customers are not the same. Some are good, some are bad.

Therefore:

All revenue is not the same. Some is good, some is bad.

One of the worst things that can happen to a startup is dealing with and especially taking money from a bad customer. The definition of a bad customer varies based on the nature of the business you are in. Some common examples in the enterprise space include businesses who’d love to have you as a custom development shop as opposed to buy lots of product from you. In the consumer space, those can be users who have a negative product experience and then start spreading the word about it to their social network. These are cut and dry examples but there are much subtler ones.

Startups listen closely to their initial customers. That’s good, up to a point. Customers are the lens through which a startup sees the market it is going after. If the lens distorts reality in some way, especially if the company is unaware of that fact, this can lead to a series of product and business decisions with lasting and often irreversible implications for a company. Someone once said that customers own the company’s product roadmap. If a company cares about its roadmap, it better care about who its early customers are. It should therefore spend time thinking about how to select who its early customers are. That’s especially true if the company is going after a large target market.

Here’s an example of why this is the case. Say your target market is moms with kids under X years old. Let’s say there are 10 million of them. Your plan calls for having 1,000 moms use your super-cool product in the first month growing to 10,000 moms by the end of the third month. Further, let’s say that you’re doing agile development with two week iterations. In three months you’ll do six iterations collecting feedback from 0.01% to 0.1% of your ultimate target demographic. If all moms are the same and they all use your product in the same way that sampling rate would be sufficient for you to evolve the product in the right direction. Chances are that that’s not the case for your product. That creates two problems. First, you won’t have large-enough sub-samples to understand the various sub-groups within your target audience. Second, and what often hurts startups much more, is that your feedback pool will be a mishmash of what’s important to different sub-groups. If you listen and implement what the aggregate group of your customers asks for you may not be able to create an outstanding product experience for any one sub-group.

If you are building something that’s targeting a problem for a large audience (10M moms) the go-to-market question you should be asking yourself is not whether you can find some early users (1,000 out of 10,000,000). Instead you should be asking how do you select (or, to use inbound marketing terminology, make yourself discoverable by) the right 1,000 moms and what to do with any other mom that wants to use your product.

It is not often easy to figure out the profile of the best early customer for a startup. There are lots of trade-offs, for example:

• Someone who’s cheap to acquire and may not love the product vs. someone who’s expensive to acquire but will likely love your product?
• Someone you can learn from vs. someone you can get money from?
• Someone you can get money from now vs. someone you can get more money from in the future?
• Someone who has influence (Walt Mossberg) but whose negative opinion can hurt your brand and even fundraising ability?
• Someone who has a big brand but who’ll be slow to sell to, overly demanding and difficult to support or a nobody who you can make really happy?
• Someone who helps you reach critical mass (in a geography, interest community, etc.) at higher acquisition cost vs. anyone who wants to be a customer?
• Etc.

The right combination of trade-offs depends on many factors unique to any given startup. My call to action for entrepreneurs is to do the following:

1. Make a hypothesis about who your best early customers should be based on current product, roadmap and resources.
2. Validate that using your favorite techniques such as customer development
3. Develop a mechanism for determining whether a potential customer is of the right type
4. Create and validate an acquisition plan for that type of customer. Repeat 1-4 as necessary.
5. Explicitly decide how to handle potential customers who are not of the right type

The last point is very important and something startups typically find very difficult to do. Saying no to customers (revenue, etc.) is not easy. But remember: you don’t want the wrong type of customer. It can lead to problems with product/market fit, cost of sales and support, product and brand perception, capital efficiency and fundraising, etc. It can cost your team countless hours of work down the road and tons of money. Stop thinking of your customers as merely a source of benefits (validation, information, revenue) and think through the reduction in your ability to be an agile startup and real end-to-end costs associated with your ongoing commitment to deliver an outstanding product experience to them for years.

You don’t have to say yes to all potential customers. There are effective strategies for diverting and capitalizing on unwanted interest. Two that I’ve been using with the startups I’m involved with through FastIgnite are to either be honest and tell interested parties that you are not ready to have them on but you’ll let them know when it’s the right time or to set up qualification hurdles to gather data and validate fit and interest. I’ve worked with some companies that used smart automation combined with techniques from behavioral economics to not only make this a low/no-touch process but also to build up demand for their product by having non-customers market it. How’s that for turning lemons into lemonade?

What if you can’t find your ideal customers? That’s a real problem but the best solution certainly isn’t ignoring the root cause and acquiring lots of random customers. A good topic for another post…