Venn To Get Funding: The Entrepreneur’s guide to TAM (total addressable market) on the Cheap (or free)

Venn To Get Funding: The Entrepreneur’s guide to TAM (total addressable market) on the Cheap (or free)

This next post is a three-parter going straight down into the maw of one of the single most important part of getting any project or company off the ground: The dreaded TAM (Total Addressable Market for those engineers who are trying to make something work without the brain pollution of standard marketing types.), or "So how big is the market for this, really?"

It’s sooo much easier when you are working for a big company.  Just call the folks who keep the corp. library or competitive research group and ask for all the market size data you need.  (Ok…it’s not that easy since most new ideas don’t have direct market data support but require bank shots to figure out, but still a lot easier.)  But when you’re in a small company, start-up or start-up wanna-be, you are in a serious catch-22.  You need credible market size numbers to get funding, but without the numbers, you have no extra cash to spend money on analyst reports.

What we’ll cover here is why the hell you need to build the TAM in the first place, how to figure out how to use mental sonar to ping around the shape of credible market sizing  when you are blazing new trails, and where and how to find good market data for nothing, since that's usually yer budget.

Part 1:  TAM the Torpedo-full speed ahead (Or why do you need a Venn anyway?)

Part 2: Market sizing as a word problem

Part 3: Where to sneak free market data

Part 1: TAM the Torpedo:

In its core, a good Total Addressable market is there for two reasons. To convinced people with purse strings that what you have to sell has a big enough market to be interesting if it’s a winner (Here's a great VC, Brad Feld on the issue). VC’s and other $$ folks have seen a painful number of these, most of which one way or another indicate that everyone who’s heart is still beating is a potential customer.  As potential entrepreneurs, we all are always given mixed messages of being realistic about our numbers, while if we show a market that just is a solid businesses we get ash-canned for not having a big enough potential to be interesting to bet on.  Layer on top of this the fact that most of the sources of data are frankly not that good. (Having been an analyst for part of my career, it’s one of those cases of knowing how sausage is made.) The reality is that the market will never end up anything like what your numbers say.  Having said that, getting a good TAM does have a few very valuable aspects, both for the potential investor and for you. 

Loosing money but making it up on volume:

The two best values that doing the TAM exercise comes when you plug it into your napkin business model and discover that even if you smack it out of the park, the damn thing doesn’t make any money, or in using it to figure out what you need to tweak in order to get that $$ flow to actually work.  When I was at Abobe working on the precursor to Photoshop.com, we ended up doing a business model for the service that ended up showing that even if we could make the sucker work, that the $$ just didn’t make it work.  Of course, in order to get to that point we had to make a bunch of assumptions, how many people of our TAM would actually use the service, how many page views would they have, how would the viral spread work, how many would be willing to pay for a service, how much would the ads bring in, etc.  The interesting thing, is that when you do these models, having a lot of assumptions isn’t as bad a thing as it appears, as long as you list them out clearly in the prezo and business plans.  The people you are trying to loose from some coin will have their own feelings about each of those assumptions and as long as you’ve set up the business model to easily play with those numbers to see what happens, you are in good shape.  More importantly, playing with those dials and seeing what happen tells you a hell of a lot about how to actually engineer the offering.  If you see that increasing the number of page views is critical for getting the numbers to work, then that influences the design in one direction, if the number of people that actually put down money for a subscription ends up being critical that tells you something else.  My most recent effort was a service to have people play games to organize their photos.  As we looked at the numbers, we realized that even though Mom’s were the ones that cared the most about getting the problem of organizing photos done, that the percentage of Mom’s that played casual games, while pretty big, was a heck of a lot smaller then the Mom’s who had kids that played games.  So it became about figuring out how to tweak the system so the kids could rock the games, but the mom’s would get the results and the chance to lay down a credit card.  Understanding those knobs in the business models and how a slight tweak here or there makes a big difference in the bottom line gives you a very, very clear set of priorities for development.

From the investor’s standpoint, what they want to see is not a Pollyanna look at the numbers showing them money raining from the sky (although, if they get a whiff of a real possibility, it doesn’t hurt.). In one of those rare honest moments, most of the people who dole out the money don’t believe the numbers either.  It is a shared lie we all agree to believe. What they want to see is your ability to look at the numbers with some intelligence and understanding.  They want to know that you know your market better then they do, and they want to find out if you’re gonna try to feed them BS.   As they look at that TAM, they will drill to the point of pain and challenge like hell.  What they are really doing is seeing if you know yours stuff and if they can trust you.  Having said that, being too conservative doesn't gain any points either.  They will either think you don't want to swing for the fences, or that you are trying to do a small business, not a potential big hit, and you are bound to hit the bricks.

Part 2-Market sizing as a word problem:

The first step to this is to turn what we think our addressable market is into a math problem. Remember, for most new markets, we’re talking about places where there IS NO GOOD DIRECT MARKET DATA!   If there were good data, it would be because it is an established industry and what the hell are you doing there trying to do new stuff unless you are Captain Disruptive.  This drives internal entrepreneurs crazy because most big companies want to have the market proven before they will dance.   So most of what we try and do is good bank shots, sort of like shooting around something to see the outline of it appear. 

I had a perfect example of this occur back when I was Mobile Editor guy for Game Developer Mag.  At the time, all the game developers wanted to know how many games were being sold for cell phones.  This was data that the game platform guys gave away freely as a way to convince the game developers to come play in their world.  So off I went to gather the data for the next issue…to find nothing.  For the first week, I just thought I was lame. After the second, my spider sense started tingling, and by the third I figured out what is now obvious, that the only folks who knew the market data for real where the cell phone carriers, and they were used to being the kings and keeping their dam data to themselves. So off I went into reporter mode.  I was able to find:

•  Analyst reports that talked about what market share the different carriers 
• Press releases with total % of sales as games vs ringtones vs apps
• The average sales prices were, but I was missing amount for the sales

Almost a year later, at a talk to brag about how well they were doing, the President of QUALCOMM talked about the total sales amount that Brew had come through their hands. 

So the math problem ended up as

Q(Qualcomm % of market) = Known

G(Percentage of overall sales that were games)= Known

A(average price per cell game)= Known

PP(number of games sold per person)=Known

QS(Qualcomm sales)=Known [Finally]

T(total game sales)=(QS/Q)*G

N(Number of games sold)=T/A

C (Potential customers)=N/PP

Later, when I had access to Adobe’s corp. subscriptions and was working on trying to get them to do a mobile version of Photoshop, I was able to find some Analyst speculation numbers that ended up matching up pretty close with what I’d been able to dope out.  What was interesting there was that the mid-level management folks who held the purse strings at that point refused to let the project go from prototype forward until I could show them data on apps that had sold for Cell phones.  I was unable to convince them that the fact that people were willing to do micro payments for leisure apps on cell phones was as close as we were going to be able to come until someone else had turned the market from being entrepreneurial into established and that they were going to need to pinch their nose and make a jump based on the market we could see echo located from the game app purchases.   This is not a unique failing that I encountered, and is in fact, one of the core observations of Christensen’s Innovators’ series of books. (Doesn’t make it any less frustrating, and if any of those guys are reading this…cough, cough, App Store…)

Part 3: TAM Data for Free:

Now here is one of those areas I have to admit to an unfair advantage.  Spending all those years as press-boy or analyst boy gave me a bunch of skills to sniffing out what was going on in a market that require not a single bucko. When you are looking for that source data, there are a bunch of cool places and techniques to find it.  Some are direct, some are bank shots.  These consist of public databases and research projects, press releases, news stories (often pulled from those releases), industry associations, trade shows, competitor’s sites/public documents and talks. And finally, a great bankshot is using some of the Adsense tracking methods as a way to gauge interest.

Direct Sources of data:  There are a number of interesting direct sources of public data these range from the Pew Internet Life to the CIA. 

Public data:

• Pew Internet Life
• CIA Factbook
• Netmarketshare-user shares for Internet technologies
• Market share for browsers, operating systems and search engines
• Compete.com
• Comescore.com
• Hitwise.com

Press releases/news stores: Plain old net searching can often lead to press releases that have great data.  These often consist of headline grabbing bits from one of their reports that they are looking to get known and purchased.  Sometimes they will be projects that were sponsored by a particular company in conjunction with an analyst group designed to bring some credibility to their industry or their particular story. 

Industry Associations/Trade Shows

For any given industry, there are a number of associations that are trying their best to promote that industry.  Most of them have a strong market intelligence component.  Searching through both their press releases and blogs can often bear good stuff.  For my games-that-sort-photos company, we used both the photo industry (PMA) and casual games (CGA) industry associations.  These also usually have related tradeshows linked from the site, which can be treasure troves in their own right.  Looking through past trade shows will give a good idea of the key players in that industry and looking through past conference lists for speaker names leads to a brand new source for market data..talk prezos.

Talks:

Often key members of a company are the ones tagged to go and do prezo’s at industry events.  They want to stand out and bring something valuable to those talks and will frequently bring data that they would otherwise keep close to the vest as an attractor to a given event.  The key data that I got about Qualcomm’s sales figures came from a keynote given by their president. The problem is that most of that good data is usally in the deck and leaves with the presenter, so if you weren’t there and fast of typing finger or digital snapshots, you’re out of luck.  But there is a new source for this data, that only rarely shows up in standard web searches.  These are the slide sharing sites. I’ve gotten great stuff from here, but it requires going directly to the site and doing the searching directly on their sites.  Two of the best are Slideshare.net and Scribd.com, which is more of a document sharing site then presentation, but still a great source. Try going to either and putting in “casual games market-share” as a search to see examples of what kinds of goodies shake from the tree.

Here's an example of a prezo on how to build a market-share analysis from SD Forum:  

How To Conduct a Market Analysis - SD Forum's Crafting A Fundable Roadmap For Your Startup

View more presentations from Steve Tennant.

Peers:

There are two ways to go about this.  One is to tap a bud on the shoulder who still works in a big company and see what he/she can fine.  This violates the terms of their companies agreements with the big analyst firms, so of course, never happens.  But if it did, it would be a good way to know if putting money down for a bit of data would be worth it.  A much better source these days is the Questions section of Linkedin.  You can get a lot of great responses from the community there, but it does come at a cost.  Two actually.  The first is you will get a number of people hopping out of the woodwork to sell you the data you need.  The second is that if you aren’t careful, you can tip your hand to a lot of smart folks.  Both of these are solved if you only ask the question of your network, or if you are a bit clever about the way you phrase your questions. (I’ve always found that it is impossible to hide anything directly, but you can hide thinks in a cloud of irrelevant stuff. We code-named on of my first products, SuperCard as Gershwin to create a false notion that we were working on music software.  It actually created enough alternate possibilities that it actually worked for us to keep the secret for a while.)

Indirect Shots:

            Adsense is a great source of market data if you know how to use it.  With some work, you can use it to gauge how many real people are looking for the words that indicate interest in your product or service.  Here is a good step by step on how to dope that out.

Lessons Learned: Using AdWords to assess demand for your new online service, step-by-step

A couple of quick tips…when you find a good source, in addition to making a note of the source and link location, do a PDF of the reference.  Things disappear on the net and there will come a day when you need to prove your sources.