As a startup investor, I have seen many business plans. But almost all of these plans come across to me as semi-jokes because they are missing a crucial element -- in-depth customer research.

If you are a Steve Jobs fan, you know that he thought such research was bunk. He believed that customers did not know what they wanted until he showed it to them. To be fair to Jobs, this approach worked for him.

Chances are you are not the next Steve Jobs. So if you ask for capital from an investor, come prepared with a realistic sales forecast based on the feedback you received from at least 100 potential customers.

Here are six things you must do to create a compelling sales forecast.

1. Identify deep customer pain

The starting point for your business must be what is often dubbed customer pain. This means you have found a problem that affects people -- perhaps you and your colleagues -- for which no adequate solution has been developed. If you can develop such a solution, you may have the basis for starting a company.

An example that comes to mind from a student team I worked with in 2015 is the problem of high school and college athletes who play sports like soccer and football that require them to purchase expensive cleats.

Such athletes quickly discover that whenever they wear the cleats on a hard surface such as a road, the cleats wear down quickly and they need to be replaced -- costing them an annoyingly large amount of money.

The students came up with the idea of providing an inexpensive and robust cleat guard that could be strapped to the cleats before walking off the field.

When I first heard this idea, it seemed like a solution to a significant unmet need. But how would the students know there were enough people who might buy that product? And how much would they be likely to pay for it?

2. Pick a target customer group

To answer these questions, such questions, startups must consider which customer group might want to buy the product you've developed. In the case of the students developing the cleat guard, they decided to focus on high school and college athletes -- and they discovered that there were over eight million of them in the U.S.

3. Create an interview guide

Even if you find that there are millions of potential customers, as an investor I am most interested in getting a more specific understanding of how many of these potential customers are likely to buy, how much they will pay, and other details that will help me envision the potential revenues the company might be able to generate.

To do that, a startup ought to create an interview guide that includes the following:

  • A clear description of the problem you are trying to solve and how your product solves it
  • Four questions:
    • On a scale of 5=definitely to 1=definitely not, would you buy this product?
    • If so, how frequently would you buy the product (e.g., once a year, once a month, once a week, etc.)?
    • How many units of the product would you purchase each time (e.g., 1, 2, etc.)?
    • How much would you be willing to pay for the product (e.g., $100, $50)?

4. Interview potential customers

Once you know the group of potential customers to whom you'd like to sell your product and the questions you want to ask them, you still need to find individuals within that group who will answer your questions.

To that end, a startup must try to find interviewees from their personal and professional networks. That's because it's much more likely that people you know will help you out by answering your survey questions.

The cleat guard group included many students who played on the college soccer team so they were able to send their survey over social media to fellow team members and colleagues at other schools to get a sufficient number of responses to these questions.

5. Analyze the results

Once you get back the survey results, the next step is to analyze them. I advised the students to apply rules of thumb to discount the results based on the observed reality that people tend to exaggerate when they complete such surveys.

Specifically, I suggested that they assume that only 80% of the people who say they definitely will buy a product and 30% of those who say they probably will buy should be counted. For example, if 20 people said they definitely would buy and 40 said they probably would buy, I'd assume that only 28 -- 80% of 20 plus 30% of 40 -- would actually make a purchase.

Moreover, I advised them to cut the survey responses on annual purchase frequency in half. So if the average survey respondent said she would buy cleat guards twice a year, I'd assume they would actually purchase them once a year.

6. Produce a sales forecast

Having completed the analysis of your surveys, you can easily calculate the potential sales you might be able to generate. To do that, you should start off with the percentage of your survey respondents who said they would buy and multiply that by the number of potential customers in your market. From there you can multiply the number who will buy times the price times the annual purchase frequency.

My students estimated the potential sales for their cleat guard at $10.3 million. To arrive at that number, they assumed that half of student athletes -- 4.1 million -- used cleats and that 10% of that group -- 410,000 -- would buy the cleat guards for $25 apiece once a year.

This approach is simple in concept and it helps you arrive at a good starting point for the potential sales of your product. You still have to figure out how many years you think it will take to reach that potential.

But if you do this, I will consider your business plan much more seriously. And you can have the satisfaction of defying the great Steve Jobs as an extra bonus!