Ask HN: How does one size a market? | Hacker News Reader

Ask HN: How does one size a market?

5 pointsimechura4 comments

I have always read that when validating a possible new venture one needs to know what the size of the market is for the product or service you will offer.

This seams like a simple task if you are opening a gas station in a neighborhood because you have easy access local demographic information but I have had difficulty coming up with realistic numbers for previous situations I have been in.

Take this product for instance. "A centralized log monitoring tool for J2EE applications". One would have to know how many companies in a target region develop or run applications based on J2EE technology. I have to this date not been able to find an accurate number for that.

Or another "A web based staff planning application for small businesses". In 2008, there were 29.6 million businesses in the United States but what percentage of those businesses would this product even apply to.

I am starting to think that the market analysis included in most business plans is somewhere between magic and pulled from thin air.

Can anyone offer some resources to help me better understand how you can accurately size a market for a particular product or service.

I am starting to think that the market analysis included in most business plans is somewhere between magic and pulled from thin air.

There's probably a lot of truth to that, but there are sources of information... you may just have to do some digging. For starters, the big analyst firms (Gartner Group, Forrester, IDC, etc.), do a lot of research on this stuff and publish reports and whitepapers that you can use. They do publish some information for free, but a lot of it is kinda pricey. You can also commission custom research / studies to answer specific questions. The problem is - as you might guess - this stuff isn't necessarily cheap.

Outside of those guys, there is market research you can do using information from the US Census Bureau, various State government agencies, and local / regional trade associations. You may have to extrapolate and make some guesses, but you can probably come up with something that's a little bit better than complete BFM / "pulled from thin air."

This book:

http://www.amazon.com/Successful-Business-Research-Straight-...

is fairly inexpensive and gives some useful - if basic - advice on doing preliminary market research.

Note that there are also smaller market research firms (sometimes as small as one individual) that you can hire to do custom market research. I've never used any of these outfits, so I can't really speak to how expensive or how effective they are. But I know they are out there.

You may also find some useful information by trawling through the "Market Sizing" and/or "Market Research" topics on Quora:

http://www.quora.com/Market-Sizing

http://www.quora.com/Market-Research

Firstly there are market research firms which specialize in doing this kind of research, it tends to be expensive but if you have a local business library you may be able to use it to get free access.

For your J2EE example I'd have a look at the existing products out there and see how many users they have (if it's open source look at their download figures, if it's commercial see if the company accounts are publicly available). Remember you don't need a precise figure, an order of magnitude will do for judging a market size.

For your web based staff planning application, I think you need to be much more specific about what your product does. If you can't figure out which business your product will apply to then you'll have a hard time selling to them. Start by thinking how you'll identify customer to sell to and the generalize from that.

If an investor quibbles market size its usually due to lack of traction and velocity of traction.

Total number of people who are potential customers * amount your product could sell for / number of realistic buyers * 10% == more realistic outlook.

--

there are 7500 hospitals in the US which could buy my product for $100,000 dollars per hospital, but only 25% of them would actually be able to purchase it for various reasons of which I only have the ability to produce/call/sell to 10% of those in the first year == ((7500 x 100,000) x .25) x.10== ~$18,000,000 addressable market.

Then I would get even more pessemistic and assume that you'll have an exceptionally long sales cycle getting out of the gate and that you'd only land say 10% of those customers in the first 12-18 months. while not hitting your stride fro say 36 months.

Calc out your opex and this gives you an idea of the funds needed to support your runway.

Reply on news.ycombinator.com