5 steps to a good market analysis for your startup (aka Market analysis 101)

I met an entrepreneur on Monday who is a young (fresh out of college), enthusiastic and a quick learner. Over the last 6 months, he and a friend from college have started working on a very interesting idea in the eCommerce space.
Both came in and presented “what they had done so far” – to get feedback from me and also get some pointers on how to prepare their pitch and company so they can get outside investment. The conversation got to “how big is their market”.
I wanted to document what I shared with them so you can help me with some feedback if I have missed anything. I have used these metrics consistently for a few startups and keep adding as needed an on going basis. Its not “quick and dirty” but at the same time its not meant to be “paralysis by doing immense analysis”. 
This is market analysis 101 which I suspect most seasoned entrepreneurs already know about.
First two metrics: Total market  and addressable market. The total market is the total revenue generated in a particular segment of the economy.The addressable market is the total amount of revenue that your company could generate if it acquired every potential customer. (source)
1. Understand the total market: Its usually a good start to look for numbers at a general macro level. Say you are looking to build a eCommerce site to sell widgets online. Top down analysis of your market looks at overall demographic, economic and other trends to find out how many people buy widgets (of all types). This is less about how many widgets are sold and more about what’s the total need for widgets overall. Subtle difference, but an important one. 
2. Compute the Addressable market:  Currently widgets are sold offline via physical stores. You should be able to get current total number of widgets sold. This is your addressable market. This assumes every person will buy widgets online and give up buying widgets offline. This is a highly unlikely scenario, but in certain segments, such as travel, increasing number of people are going online than to a travel agent to book tickets.
3. Determine the possible Market share: Take into account factors such as internet penetration (low in India), penchant for people to buy online versus offline and you will arrive at another number that will tell you the market for widgets online. Assuming excellent execution, your market share of that number will be anywhere from 30% (you are one of several players in the space to sell widgets online) to 60% (you are the dominant player in the space).
So, if the total market for widgets is $5 Billion, of which currently $2 Billion worth of widgets is being sold, assuming 10% of widgets will be sold online, you arrive at $200 Million. This is the amount of widgets sales that can occur online. Of this if you assume 30% market share with a first mover advantage, then you can possibly get about $67 million in sales, assuming you execute well.
4. Ensure you double check with price and unit sales. If you can only source 10% widgets, and assume you charge a lower price than offline vendors, this also lowers your potential revenue. So if offline stores sell 500 Million widgets at $4, and you can only get 5 Million widgets, sell them at $3, then you can get a max of $15 Million revenue.
5. Analyst reports based projections: Usually I dont give much credence to market projection by analyst firms. Many go through the same process but get lot more data points. Usually more data points is better accuracy, but the projection business is tricky and paying someone to get “more accurate” numbers is not worth the money in my opinion.