How to justify ROI?

You're barking down the wrong track, here. (Did I just mix a metaphor?)

There are too many possible reasons why a particular business idea may or may not take off.

Indeed your list of five factors is a little bit silly. The sales volume depends on

(Number of people in the world)
*
(Percent of those people who could use this product) // unknown
*
(Percent of those people who find out about this product) // unknown
*
(Percent of those people who can afford this product) // unknown
*
(Some big random number)

When you multiply a really big number (6 billion) by all these other little numbers that you don't really know, you get: nothing. bubkis. Meaningless drivel.

There's not much point in modeling something like this, and smart investors get it.

Instead you need to show two things.

1. The size of the opportunity.
What happens if everything goes right? This needs to be a really big number. If your business plan consists of selling T-shirts on the 9:36 commuter train from Darien to Grand Central, well, the size of the opportunity even if everything goes right just isn't large enough to get an investor interested. Typically VCs are looking for opportunities in the $billions. There are plenty of great businesses (like bug tracking software) that probably could never be a billion dollar business even if everything went perfect.

2. A list of reasonable goals on the path to that opportunity, with success tests at each goal.
You need to be able to say: "In order for this idea to pay off, we need to prove these theses which are unknown. Here is what we will do to test these theses."

Since I'm in the process of launching a new business at careers.stackoverflow.com, I'll give you an example from my business. I have two early goals and tests for whether or not they are successful:

1. Enough programmers will submit CVs so that employers should be willing to pay to access them. I picked a metric for this: there have to be enough CVs that I can find qualified people to work at Fog Creek in New York. I proved that thesis about two weeks ago.

2. Enough employers will see the value to pay for access to the CVs. We did an early test, asking a handful of hiring managers what they thought. That went well enough. But today we're rolling out the live service, and crossing our fingers to see if anyone pays. That's the second test we have to pass.

Later, as the business expands, we need to do more tests. The other two theses I need to prove next:

1. There is a cost efficient way to get more programmers to put their CVs into the system. For example, we have to be able to find and motivate a programmer to put their resume into the system for less than the $29 that they're going to pay us.

2. There is a cost efficient way to market and sell to hiring managers, using some system that has a positive ROI. For example we need to be able to sell a $500 job search for less than $500.

Neither of those goals has been achieved yet.

So... what I'm saying is, you have to show the investor that the size of the opportunity is large enough to justify an investment, and that you understand well what all the assumptions are that you made in asserting that your business idea will work, the so-called "entrepreneur's thesis," and that you have a plan and a program for proving them one at a time.

The nice thing there is that even if the investor passes, you can send him little progress reports once in a while as you prove each of your theses.

By the way, the more of your theses that you can prove before you take money, the better a valuation you will get. In other words, StackOverflow Careers will be worth a heck of a lot more money next week when we've shown that recruiters will pay for it than it was worth this week when that thesis remained to be proven.

The only other thing I can add to what Joel wrote is that you SHOULD NOT prepare answers for what you THINK the investors want to hear. It almost sounds like that is what you mean when you say, "I am trying to build a presentation strategy", however, I may be off base.

They'll see right through the BS. These guys (and gals) are not stupid. Don't try to tailor your pitch/presentation in any way other than meeting the usual guidelines regarding startup "pitches" and business plans. (you can find those on angel group sites, VC sites, etc)

You're better off being open and honest about what you don't know, what the risks are and how great your competition is. That will at least show them you understand the market and are not trying to BS them.

The business model is sound, and the product will add significant values (benefits/costs, integrity of services, etc.) to the end-users.

That doesn't necessarily mean that the product will sell. E.g. another company might be faster to offer a comparable product, or you might just have a hard time to convice potential users that those advantages really exist. Don't underestimate the cost of marketing a sound product.

i.e the product will sell.

Even if, that doesn't necessarily mean that there will ever be enough profit for your startup to survive.

Basically, you must be able to calculate the expected costs and the expected revenue. Otherwise, you cannot even say if you are on track or not. If you just start with "the idea is good, let's see what happens", chances are that you get nowhere.

Having done some angel investing and works with many angels, I've never heard anyone talk about investing in companies as an ROI.

So I have to agree with Joel -- wrong premise.