The Official Blog of Cogmap, the Org Chart Wiki


Archive for May, 2007


The Time is Money Gap – The Penny Gap for Content Creation

Thursday, May 31st, 2007

So sometimes I wonder about our map quantities. We have a few thousand organization charts when Hoovers touts “18 million” and Zoominfo says they have “3.5 million”. My contention in a recent conversation was that in a consumer generated world instead of an automated world, the focus on what is actually important becomes more visible. The 80/20 rule applied to their automated service means most of their data is not valuable, whereas I only capture the 20% people want via self-selection. What people make available, people want. A lot of those millions of companies aren’t the companies anyone cares about. Whereas when someone posts an organization chart, all of the sudden you are talking about a business.

It would be great to look at the overlap in companies from a list service. The odds that they reach out to the long tail when building lists is by definition relatively unlikely.

Wikipedia has an interesting, but woefully inaccurate, entry entitled “List of US Companies” that only has 400 companies in it. Obviously that is wrong, but it certainly provides some directional information about the huge gap between the number of businesses that exist and the number of businesses that people want to know about.

Then I went and did some research to figure out if I am lying. Here is what I found courtesy of Biz Stats:

  • There are almost 20m business entities in the US that have no employees. Real estate agents, nannies, Mary Kay and Amway account for more than 2 million businesses.
  • Only 0.5% of companies (16,740) had more than 500 employees (The most likely companies to have org charts here) but they represented almost 50% of total US employment.

Here is the chart:

Number of employees

# of firms

# of employees

% of all
Employer firms

% of total

less than 5

























more than 500





All employer firms





So looking at this, how important is getting access to 3.5 million companies? I know if I were an entrepreneur or a sales person that I am really interested in maybe 20,000. Maybe. Probably more like 10,000.

The next question is obvious and hard to answer: What I would love to do is figure out a way to map this back to Cogmap’s database. We definitely have small companies. We also have a lot of maps of big companies that only have 20 or 30 people in them and/or are broken down at varying levels of granularity. It is easy and probably accurate to assume that if Hoovers has 16 million companies, they have most of the organizations you care about.

Capturing more metadata about a company would be great, but the trade-off is that setting the participation bar higher for map creation results in fewer maps. In fact, one might suggest that when you compare Cogmap to automated tools, you are seeing “the Penny Gap” for content creation. Time is money in this model. The difference between requiring no effort to harvest data and effort to harvest data is the difference between millions and thousands of records. The good news is the concept of the Penny Gap only implies that good customers are customers that generate revenue, whereas in Cogmap, good maps are maps other people want to consume and also tend to be the maps people post. Thus the Cogmap model creates a system that surfaces the most valuable maps more rapidly.

While the absolute amount of data follows the straight, linear curves described in the Penny Gap, the value creation of the initial wave of maps is much higher.

Most people tend to work for one of just a few thousand companies, so those are the most likely maps to be posted and the most likely maps to consume. The biggest company maps have high value to lots of people. This is basic network effects 101 stuff. The value of maps drops to zero fairly rapidly and those are both unlikely to be made available and unlikely to be of interest – a happy concurrence. That millionth organization has very low value to very few people (org charts in three person companies are a bit superfluous and it is unlikely that headhunters or sales people want to call on those companies) and hence a low likelihood of getting posted. It is nice for the automated systems that they can capture it and add another statistic to their database, but not necessary for anyone who needs these systems because the odds they use it are low.

Hmmm, I need a better name for my Gap. Post comments.

Why we didn’t name this “LinkedHow”

Wednesday, May 30th, 2007

Coming up with the name “Cogmap” was a surprisingly arduous, agonizing process.  For a name that just rolls off the tongue, getting here was surprisingly tricky.  That is probably true for any company name: When you start with a blank slate any name will do and no name sounds perfect.

Cogmap was “Org Chart Wiki” for the longest time.  Then it was CogMap.  Frankly, it still is CogMap in lots of places because it hasn’t been worth the time to run around fixing it, but I recently felt like intercapping is so Web 1.1.  Now it is Cogmap

When we picked the name, we distinctly headed in the “name the company what it does” school of thought.  In the big scheme of things we were building Cogmap because I was sick of thinking what a good idea it would be to build.  This was definitely a product or feature, not a full-blown company.  ATTENTION: I am not planning world takeover based on this idea.

I read once about REM’s process for coming up with a name for an album.  They put a piece of paper on the wall of the studio while they are recording it and every time someone has an idea they write it on the paper.  Then they review the paper at the end and pick the name for the album.  The process we used for coming up with a name was similar.  We started a thread with all of the concepts and then refined it to all the words.  Having said that, Cogmap was never on the list.  In fact, the term “Cog” was never on the list.  I came up with the name while agonizing on the short-comings of the list and the first feedback we got was “you are calling people a ‘cog’?  That’s insulting!”  My response was, “Hey we are a start-up.  If people get offended then I don’t like them either.  People have to love a name that is a little tongue in cheek.  It’s time for everyone to get in on the joke.”

I loved the name because it was a little funny and because it had a nice symmetry.  2×3.

I am a big believer that a decision that everyone likes is not radical enough.  We don’t want tame in a name.

Anyway, here is the list of names we considered:

First round of ideas:

Our Charts
Who Works
People Map


Here are some funny comments on this last round of names from Mark Maloney over at No Inc:

“LinkedHow is funny…just wondering how they’d feel about it.

CogMap definitely seems like the winner to me.

I’d say that if we had to put two toe-to-toe that it would be CogMap and

I thought Chartiki sounded fruity at the time, but in retrospect, an equally good name that we could have built a dandy business around.  I am surprised no one put “Chartpedia” on the list.

Another reason I selected Cogmap is that it seemed ripe for conversion to a verb:  You can cogmap things.  Did you cogmap it?  I was cogmapping.

If enough people ask, I will post some of the logo candidates and discuss how we chose colors.

How to name your start-up

Tuesday, May 22nd, 2007

There are a couple of questions I ask every entrepreneur because I think you always get a good story out of it.  The founding story is always great.  I am always fascinated by how people come up with ideas.  Also, no founding is super smooth, but the bumps in the rose are how you take time to smell the roses.  The other story I always ask is how they came up with the name.  I think there are two schools of thought when it comes to naming businesses. 

The first school of thought is to name the company based on what it is.  This is the Microsoft,, Coca-Cola, Cogmap school of company names.  This is a great approach because when the name says what the company does, all you have to do is hear the company name and you kind of know what the company does.  Software, online advertising, cola. maps of cogs.  Done. 

The second school of thought is the random name strategy.  Yahoo!, Google, eBay, NetGravity, Engage, and Group Cortex are all examples of this approach.  The benefit of this approach is flexibility.  Most companies don’t end up doing what they started out to do, so picking a name you can keep using when you change businesses can be nice.  This also gives you a name you can kind of grow with.  Great counter-example here is GE.  General Electric used to be a name that told you what they did.  Now they do everything so the name is no longer the asset it was from a descriptive value objective.

Cogmap definitely is an example of the name it what it does school of thought, however it is kind of tongue in cheek.  I will talk about how we named it in my next post because someone might think that is kind of funny.

Microsoft succumbs to deal heat

Monday, May 21st, 2007

Everyone else is talking about the rabid pace of M&A activity in the ad serving sector, so I guess I want to weigh in also.

Here is a thought on each transaction:

  • $3.1 billion was a lot to pay for DoubleClick. Nice work by that team. The big focus on Project Wolf as the auction started probably worked wonders for that team. Google was the company that had the most to gain and the most ability to leverage value out of the platform, so it is not a super surprise that they were able to justify a higher valuation than anyone else.
  • Yahoo pays $840 million for Right Media and makes DoubleClick look reasonably priced. Everyone in the media talks about how Right Media “has projected revenues of $70 million in 2007”. Hey, it’s the second quarter. I assume they have a big hockey stick on that to justify the valuation, so the actual revenue so far is probably not a lot. Yahoo places a big valuation on a company whose upside has yet to be realized. For a bit more you could get a company with a similar product (Wolf) and a near monopoly on the ad serving business (with relationships everywhere).
  • Microsoft needs an adserver too and the second biggest one is still on the market! Go get it! The problem is that it is owned by this big professional services business. I haven’t seen anyone try to really break down aQuantive numbers, so I will take a quick shot based on Q1 results:
    • aQuantive is primarily a professional services firm. More than half of their revenue is professional services related. If you value that part of the business at the same multiple that Digitas is trading at right now (32x PE), it threw off profits of $10m in Q1 so let’s say that it is worth $1.4b. I was pretty generous there.

    • They have DrivePM, an advertising network. Let’s value that similar to ValueClick, which is doing well right now with all the press. (47x PE.) They had Q1 profits of $3.2m, so let’s say that business is worth $600m.

    • That leaves the adserver part of the business. It did $38m in revenue and $13m in profits. Let’s assign that a 50x PE and maybe it is worth $2.6 billion.

So then there is a $1.4 billion premium paid on the deal.  You could get there from here.  It is a stretch – Digitas’ PE is already inflated by the acquisition from Publicis, ValueClick’s PE in inflated by rumored acquisitions, these numbers are all big, so a premium on top of it is really big, but it’s possible.

If these numbers are kind of like the numbers they used, then Microsoft paid ~4-5x revenue for the professional services business.  That is aggressive.

Sounds like deal heat to me.

Subscribe to organization charts in your RSS reader

Friday, May 18th, 2007

Have you noticed that there are now RSS feeds on every map page?  That is right, using your RSS reader you can subscribe to maps and be always up-to-date on the latest changes.  Insanely great or just insane?

This is a great example of coding something because it is cool rather than coding something people will actually use.  Our next big thing is focused on email subscriptions, which will be a way to subscribe without having to geek out.


Marketing and Sales in a Web 2.0 World

Thursday, May 17th, 2007

When people talk about how building a Web 2.0 business has changed, much of the focus is on the business model and on the technology expenses. As I talked about in the last post, the business model change is huge. Technology expenses have changed a lot also. Cogmap runs on Media Temple’s grid server and that does not cost a lot. (Is there a place I should host it to make it run faster? Let me know.) It’s LAMP all the way and that is all free, so that’s good. My first Internet start-up had to pay more for bandwidth in a month then I will spend in a year on Cogmap. We had Oracle databases running on Sun boxes. That means you have to spend a lot of money on hardware and software to get going.

But I don’t think those are the biggest changes in how to run a Web 2.0 business.

Marketing and sales used to be expensive also. I tell everyone I know that I think it is one of the most widely under-reported phenomenons of Web 2.0.

Marketing a new web site is crazy cheap today. Think about how tough it used to be to educate a million people about your company. You had to advertise. You hired a PR firm. You did marketing. You got trade show booths. You dropped a million dollars on the launch. And that was perfectly justified because you had just spent five million dollars hiring editors.

Today you send one email to Michael Arrington and a million people come play with your toys. Pretty inexpensive! The growth of the blogging community and the ability to reach out to them to introduce new web services has made marketing Web 2.0 products cost almost nothing. People like or even Google (Ads by Goooooooooooogle – the biggest unpaid advertising campaign in the world) can essentially spend nothing on marketing. And let’s be clear: This is not viral. I don’t use because my friends do. In fact, I am not sure if my friends do or not. This is just really cheap marketing thanks to word of blog.

Sales is expensive too. In Web 1.0, after you hire your fifty editors, you hired a sales force. They would go close some launch advertisers and try to sell your inventory. Of course, the sell through rate is low, ROS inventory was inexpensive, the online advertising market was small, so budgets were tiny. The result was that ends didn’t frequently meet (certainly not early on) so people resorted to subscription models. As The Penny Gap discusses, this kills traffic.

Today, every impression can be monetized by Google and you can defer the cost of hiring a sales force until traffic volumes and advertiser demand justify the head. Google will happily let you outsource all that work to them and pay decent CPMs along the way. Don’t like Google? Dozens of companies will give you tools to let you liquidate your inventory in a few hours a month –, YPN, ValueClick, Blue Lithium, Casale, Right Media, etc.

Free marketing and free sales forces take a lot of friction out of the system.

Magazines lose money for years when they start-up as they pay editors, marketing and sales people while they wait for traffic and traction. Now people build Web 2.0 companies for nothing and don’t pay any of those people a dime.

The irony of much of this is clear: VCs and bloggers have discussed ad nauseum how funding requirements for start-ups have gone down as a result. But what is equally interesting is how they are no longer interested if all you have done is market to a million people, because everyone knows that is easy!

Cogmap and the Web 2.0 Business Model

Wednesday, May 16th, 2007

I wanted to expound some more on my ideas about Web 2.0 and how it has impacted our thought process:

Preface: When I say Web 2.0, I talk exclusively about business models. Some people, when they say Web 2.0, mean technology like Ajax. I love all that, but I don’t think eliminating page refreshes is what is really crazy about the Web 2.0 world.

So, in reference to business models, here is what I think makes Web 2.0 work and Web 1.0 a disaster: In Web 1.0, you raised a bunch of money and hired 50 editors to create content. These 50 editors had to get paid so becoming profitable took a lot of revenue.

Web 2.0 is about user-created content. If you can harness the public to provide the content they consume, then you have changed the economic model for creating value. A good Web 2.0 website provides a framework for extracting knowledge from visitors and framing it in a way that (hopefully) adds some value then reflecting it back to the community for consumption. The result is that a Web 2.0 effort is a (small|medium|large) technology effort and that is it. Web 1.0 ideas had the same LOE (usually more, that’s some other post) for technology plus editors. How this changes the consumption of VC money is documented in many many plances.

Gotta love MySpace. It looks like it was probably coded by half a dozen people and now content is created by 70 million unpaid editors every day. That is a financially reasonable content creation model.

We can really illustrate this by breaking down Cogmap and looking at the business list industry. One of the things that we thought a lot about in creating Cogmap – our competitive advantage – is that companies like Factiva, InfoUSA, and others generate lists by hiring armies of people to call into companies and generate organizational data. Not only is it expensive, the ability to maintain it is limited by the size of their small army. I have laid out their model in this picture:

Their Way (Web 1.0 in theory)

Web 1.0 Business Model

The result is the amount of data generated is directly correlated to the expense. To generate more data (both volume and constantly confirming accuracy) they have to add an editor. Scale is limited by staff.

The good news for them is that this model is not a straight professional services model. They can resell the data an infinite number of times, so revenue is not correlated with expense and they can make a lot of money. Unfortunately, the amount of money they can make is constrained by their willingness to invest in accurate data. The more money they want to make, the less they spend on good data. The less they spend on good data, the worse the long term prospects of the business. So they must constantly re-invest in data to justify the ever-growing customer base. The result is a good business with a high start-up cost, but not a great business.

In a Web 1.0 editorial world, it’s just a bad business. Generating content is directly correlated with page views. Conceptually, you get the same data consumption angle, but practically what happens is that you need more and more content to generate the page views so revenue becomes tied to content generation, capping the upside while creating a huge and expensive launch where content must be generated prior to pages being viewed.

So here is the Web2.0 approach:

The Cogmap Way (Web 2.0 in theory)

Web 2.0 Business Model

The Cogmap business model is Web 2.0-ish. We build technology, look to the community to contribute the data that other members of the community value and consume and our data increases (along with our ability to monetize data, theoretically) while our costs do not.

The next hurdle to make it uber-2.0 is to figure out a way to get the community to force non-community members to play. LinkedIn uses you to go recruit your friends. MySpace has people go recruit “friends” that are not on MySpace yet. The result is a strong centrifugal force that sucks people into these networks. Cogmap is a Web 2.0 tool, not a social network, so it will never generate the high volume, high frequency page views that these sites generate.

There are several more things that have changed in the market that drive company building and capital consumption, but I want to talk about those in a different post. Moral of this post: Hiring editors is expensive for start-ups. Try not to do that.

Who loves OpenCoffee?

Tuesday, May 15th, 2007

I love the whole OpenCoffee idea. Who doesn’t want to hang out with like-minded people and make some new friends?  I think congregating techies is one of the best things that can happen in the world.  I keep thinking about trying to start one in the Baltimore area. Leave a comment or ping me if you are interested.

Why doesn’t LinkedIn have APIs?

Monday, May 14th, 2007

Yes, I know that magical LinkedIn integration would be kind of cool.  LinkedIn shows social relationships, Cogmap show organization relationships.  There is probably synergy.  Having said that, if integration just means a field in our system where you paste a LinkedIn URL for a person, we aren’t going to do that because it is lame.

Post your ideas for how we should do it as comments.  LinkedIn can contact me directly if they want to do something.  :-)

Easter Egg: Try our escape key!

Saturday, May 12th, 2007

We have been up for six months and I bet here is something that no one knew: Anyone tried hitting the Escape key on a map page?  That’s an Easter Egg!  You will go nuts for sure.