Ask HN: Estimating market sizes

You can buy reports at <a href="http://www.marketresearch.com/" rel="nofollow">http://www.marketresearch.com/</a> or you can cobble the data together from free sources, such as the US gov't sites:<p><a href="http://www.census.gov/main/www/access.html" rel="nofollow">http://www.census.gov/main/www/access.html</a>, <a href="http://www.bls.gov/data/" rel="nofollow">http://www.bls.gov/data/</a>, <a href="http://www.sec.gov/edgar.shtml" rel="nofollow">http://www.sec.gov/edgar.shtml</a>

I'd be careful about believing any numbers, and just because some big market research company once said it doesn't mean it's true. Certain things spring up as being the truth after they are quoted enough, e.g. the Internet traffic doubling 18 months (<a href="http://www.lightreading.com/document.asp?doc_id=18702" rel="nofollow">http://www.lightreading.com/document.asp?doc_id=18702</a>).<p>In my experience the best way is to analyse a market in several different ways and get bounds on the market size. E.g. for baby monitors in the US, the market size must be less than the total revenue of all the companies that make them, and the units sold per annum must be less than the number of babies born per year in the US. You then may find one company that claims to have shipped over a million baby monitors in a year (e.g. inside their SEC filings or even marketing material), which is therefore a lower bound.<p>A combination of google, asking people (e.g. either industry experts or in your local shop ask the shop assistant roughly how many of those do they sell a day) and you should be able to get to an order of magnitude of an established market place.

There are also usually trade organizations that will report information. They have an obvious incentive to make the size seem bigger, but if they are big enough organizations investors will also probably use them as a reference.<p>The other thing is to consider Total Addressable Market (TAM) vs (SAM) Served Addressable Market. Basically, it is the difference between a complete market and the portion you would be able to serve. This is helpful to keep in mind if you are targeting a niche market. The SAM niche may be small, but if there is applicability to the TAM it could be viewed more favorably.<p>Here is a good link describing SAM vs. TAM:<p><a href="http://windowmanager.blogspot.com/2005/05/wth-is-tam-sam-som.html" rel="nofollow">http://windowmanager.blogspot.com/2005/05/wth-is-tam-sam-som...</a>

<i>where do hackers look to discover the size of their target markets?</i><p>Its a whole lot of voodoo and has little to do with anything that matters to your business, since for a market of any size you're going to be limited by how much of it is addressable anyhow.<p>I swear, the only reason people even bother checking is so they can present the "if we got 1% of $350 million..." argument that everyone says they know is useless.

They're very often wild-assed guesses. The introduction of a revolutionary new product often changes the market size significantly. For example, the market for computers got a lot bigger when the Altair was introduced (a world market for <i>6</i> computers? Really IBM?), then bigger still when the IBM PC was introduced, then bigger still Compaq and Dell and clones came about, then bigger still when Win95 provided the first decent GUI for clones.<p>You can make some rough estimates by figuring out what substitutes your product has, and then researching the combined revenues of all the products you'll be replacing. This also degenerates nicely: if you can't think of any substitutes, then it's fairly likely that you have a $0 market.

How sensitive are the decisions you are trying to make to the actual market size? A factor of two? A factor of ten?<p>Does the market exist? Does the need exist?<p>For startups, these last two questions are most useful in my experience: trying to attack a large market because it is well established is not normally a recipe for success unless you have some way to take advantage of a new trend or a new technology.<p>I would also focus much more on what's required for you to reach break even operation in a niche or segment since failure to do so will kill you more quickly than many other things.

I think market research companies use existing sales data to estimate current market size; figures can then be confirmed with light checking. The market size estimate can be helpful to provide a reference on which to base rough figures (which usually mean nothing) or hone an idea. Investors that may not understand an industry do understand numbers, so that's why such figures are popular. I'd keep in mind, though, that a market size can change up or down for various reasons.

If you have public competitors you can often find out from their annual reports. For private companies published interviews with senior staff are often useful. Generally for most markets you have some idea of who the major players are and how the market share breaks down, so even if you can only get data for a few of the companies you can form a decent estimate.

In this particular instance you could use semiconductor market statistics, they often break the component usage down to market. They are usually for free and give some kind of indication. But as mentioned before there is lots of smoke around the market size subject and someone has to take these stated numbers with a grain of salt.

If you're trying to put together a business plan, investors will want to hear your "total addressable/available market", which means what's the most $ your company, specifically, could make.<p>TAM is size of your company's market * price of 1 unit of your product.

To get a rough idea, sum up the rough revenues of the public companies in that industry.