Earlier this week, I released my third quarter report on European and Israeli VC trends. You might have noticed that I didn’t release one for the second quarter. The reason for that is simple: my first daughter was born April 18 and I got way behind on the data entry for the second quarter. The issue is, I enter all the data for these reports manually — reviewing every single deal across a pretty wide range of news sources. That is a ton of work, and I fell behind in the second quarter. So this quarter, I pretty much worked double-time on the data and got all caught up.
This got me to thinking about why I bother doing this in the first place. The bottom line is that I’m reading all these news stories and newsletters anyway — and trying to pay close attention to what is going on and which sorts of companies are getting financed — and why. It’s a way to force myself to really think about what is going on — and to try to notice both trends (and interesting outliers) that can drive my own investment decisions.
In the past, I’ve used a range of data sources to try to think about VC trends, and what I realized pretty quickly is that a lot of that data had to be cleaned if it was going to have value. And that cleansing process meant that I would have to maintain my own data anyway…the data available from paid sources just isn’t clean enough to use. A big part of that is that the paid data sources want their data to be “complete” — they don’t want any “venture investments” to be missing from their databases. So far, no one seems to have bothered to build a reliable algorithmic approach to this. In addition, I wanted the flexibility to categorize the investments into groups that made sense to me. Since I was gathering the data anyway, I realized it was a pretty short jump from there to releasing much of that data publicly — a nice way to give back to the community of which I am a part and hopefully to raise my own profile a bit in the process.
So what am I actually doing to clean the data? Basically, I’m going through every reported venture deal in Europe and Israel and eliminating the ones that don’t really belong:
- First of all, you’d be amazed how many companies based in Vienna, VA are mis-categorized as European.
- I try to eliminate companies that were founded way too long ago to count as “venture deals.” When a company that was founded in the 1990s raises money, it’s typically not a truly a venture round. It’s generally more of a buyout or a traditional private equity round, as opposed to a venture round.
- I try to eliminate rounds that are actually private equity take-outs disguised as investments. This can be tricky, but sometimes it’s pretty obvious what is going on.
- Many PIPEs and other investments into public companies get mistakenly listed as VC investments. I knock those out as well.
- In their exuberance to show a healthy local VC eco-system, many news outlets report every corporate investment into another company as a venture round. Many of these have a lot of the characteristics of a true venture round (i.e. they are done by a corporate venture arm, or alongside traditional VCs) but many do not. For example, when a corporate makes a strategic investment in a highly complementary product or service in a slow-growing industry, that’s generally not really a venture round. It’s just an investment.
- Because I’m focused on Western Europe and Israel, I limit my deal-tracking to the EU + Shengen countries. That means I don’t track Turkish deals, Russian deals, or Ukrainian deals.
- I also don’t track most medical (therapeutic/diagnostic/pharmaceutical) unless they are primarily software-driven or healthcare IT.
- Finally, for my own sanity, I don’t track deals under $500K.
Ultimately, the real value of this exercise for me is the value that I get from forcing myself to think about every single deal done in the ecosystem — if only for a few seconds. The by-product of that process, is the unique dataset that underlies the report I put together.
You can find the latest quarterly report here: European and Israeli Venture Data 3Q17.