Insights on Web funding from Crunchbase

In my CFO & Business Development responsibilities at pearltrees, I'm interested in reliable Web specific funding data but it's hard to find. Usually, Web start-ups are included in broader categories.  Actually, I haven't found any comprehensive funding analysis focusing on Web only. Discussing about this with Nicolas Cynober - a semantic Web expert and R&D engineer at pearltrees - we decided to find a remedy: we extracted the entire CrunchBase database of funding rounds. Zooming on investments in Web start-ups exclusively (since 2004 to end of October 2008, series A to F), we ended-up with a database of 914 start-ups, 1307 investment rounds by 817 investors worth nearly $14bn. While CrunchBase may have some pitfalls (e.g., fewer deals prior to 2006, US bias), it is one of the most comprehensive publicly available database on Web start-ups funding. You'll find below some "Web specific" insights such as the Web VC ranking or overall data on investment stage, deal size and implied valuations (the tables below and detailed investments by VC are in Excel here Download Tables Crunchbase).

VC Web investment ranking

As table 1 below shows, the usual suspects are at the top of the ranking with DFJ, Sequoia and Accel trusting the first 3 places. These 3 funds invested in more than 40 Web start-ups with a total of 127 individual start-ups (w/o double counting for companies with several VCs) representing 14% of all funded companies. They are followed by Benchmark, Index, Intel Capital, First Round Capital and DAG Ventures who each invested in >20 companies. In total, these 8 VCs invested in 215 Web start-ups (24% of total). Another 36 VCs invested in more than 10 Web start-ups. In total, these 44 VCs invested in 490 Web companies (54% of total). Finally, another 770 investors complement the picture with investments in ~760 start-ups (83% of total financed companies - note that companies having several VCs, the percentage sum up to more than 100%).

Web investment stage

As tables 2 and 3 illustrate, series A funding represented on average 57% of investments rounds (39% in terms of amount invested). Series B, C and D & above represented respectively 31%, 9% and 3% (35%, 19% and 7% in terms of amount invested). Overtime, investments tended to move towards later stage rounds reflecting the advancement of the investment cycle. In 2008 (year to date), series A represented only 44% of funding rounds and 27% of dollars invested.

Web round size and implied pre-money valuations

The median round size was $~5m for series A, $9.5m for series B, $15m for series C and $20m for series D & above. There is however a large variance as table 4 illustrates. For example, there is a ratio of nearly 1 to 3 between third and first quartile series A size ($2.5m vs. $7m) or above 4 between 80th percentile and 20th percentile ($2m vs. $ 8.3). Series B tend to be in the range of $6-14m and series C $10-25m.

There is one step between round size and valuation... unfortunately this data is not available. However, by doing some rough assumptions (series A tend to go for 25-50% of capital, series B for 20-30%, series C & above for 10-25%), we can infer some estimates of pre-money valuations that should be directionally correct. Table 5 shows the resulting rough estimates (the implied step-ups tend to be in line with market data): taking the median estimates, series A could typically go for $5-15m, series B $20-40m and series C $45-135m. With this estimation "technique", bigger round size mechanically come up with higher figures. A recent study by SVB argues that larger round size have on average higher valuation levels.  Hence, one could infer that "very hot" Web start-ups may go for >$25m pre-money valuation for series A or >$60m for series B (20th percentile high estimates).

That's it. I'll try to extract some further insights from this database for future posts so keep... posted.

Comments

[mark]

Thanks for pulling all this together. Very interesting..

[Healy Jones]

Wallen - cool info; thanks for putting it together. One of my key take-aways is that we need to get our portfolio companies into Crunchbase!

[Wallen's]

@ mark - welcome!
@ Healy - yes indeed... to improve the quality of this "ranking", the best thing is to input missing data directly in CrunchBase (the beauty of user-generated content).
I'll extract the database by year end to make an updated version.

[michael arrington]

wow. great job putting this all together.

[Wallen's]

Thanks, Michael. Would not have been possible without the CrunchBase API

There is more to come and I'll try to make it a regular review.

[Jussi Laakkonen]

Excellent analysis!

I've published similar analysis on VC deals in the online games space ($1,7 billion in 2007-2008) recently at:

Part 1:
http://bit.ly/CfY1

Part 2:
http://bit.ly/ztn

[Maud]

Great job!

Although Crunchbase does not for sure present either complete or updated list of deals.

When start-ups identify top VCs by ......
- size of the fund
- amount and number of investments

.....Among the VC community the success matrix differs radically: out of X investments, how many have been bought out or gone IPO?

In this regard, DFJ has had a poor record these past two years, having done a lot of investment in "cool" technologies with less "cool" business models.

[Wallen's]

@ Jussi, thanks. Your posts on the sub-segment of online games space is terrific!

[Wallen's]

@ Maud, fair enough... As an entrepreneur, I'm more interested in this perspective than VCs return multiples (in any case this data is even harder to get...). After all I'm writing for entrepreneurs.
The question is whether the # of deals or the # of start-ups invested in is the relevant measure. I think it depends to which question you want to answer. The former is a measure of intensity of activity whereas the latter is an indicator of the "attractiveness" of the web sector to a VC. I was more interested in the latter but for future post I'll maybe switch to # of deals.
Actually, what would the readers of this post prefer? A VC Web investment ranking based on # of deals or # of start-up invested in?

[Jussi Laakkonen]

Wallen, thanks for the kind words!

It'll be really interesting to see how the investment trends will move in the next 6 months (especially as we are also planning to raise some money =)).

[Nic Brisbourne]

Thanks for some great data

[Cherif Habib]

Julien - great post! The SVB study is also quite interesting. Thanks for sharing.

[fred wilson]

we've made 24 investments (actually 26 but two are in stealth and not yet announced) since 2004 and Crunchbase only has 11 of them?

http://www.unionsquareventures.com/portfolio.html

that seems wrong since we announce every single one of them on our blog and they are generally pretty well covered in the tech blog world.

[Wallen's]

@ Fred, thanks for your comment. As I mentioned in my intro, Crunchbase has indeed some pitfalls: the database has impartial coverage of 2004-2005, 2006 is better but 2007-2008 is much more comprehensive.
Specifically for USV, 5 of your investments were filtered out given the scope of the post (i.e., Web only and series A upwards):
* 4 of your investments were classified in another category than "Web" in Crunchbase (10gen - software, BugLabs - hardware, Pinch Media - mobile and Zynga - others).
* Zynga and Zemanta were classified as seed investments
That leaves 8 USV investments (probably prior to 2007) not included in Crunchbase.

This makes me think that I should do a post based on 2007 data only (and 2008 once the year is finished) to improve relevancy.