Talent #2

Yesterday I started this mini-series analyzing whether being a B2B as opposed to a B2C startup impacts retention and turnover. The first post of the series was mostly based on interviews I’ve had with some really knowledgeable people, so today’s post will focus more on the data I actually collected on the matter.

How I picked the startups

Using Pitchbook, I filtered for active startups that meet the following criteria:

  • Headquartered in the Bay Area, in order to control for potential cultural differences linked to the region where a startup is based. Rumor has it that turnover is greater in the Bay than in other places, simply because of the range of opportunities available for talented people. That’s a question for another time, but anecdotally it seems to hold.
  • Founded since 2013 to capture relatively recently formed startups. Companies that are more advanced in their path may or may not be different in talent retention.
  • Having received funding from top VC firms Sequoia Capital and Greylock Partners, as a signal for technology companies with high growth potential (as opposed to lifestyle businesses). Why Sequoia and Greylock? I don’t have a scientific explanation for this. Doug Leone (Sequoia) has been awesome about helping out the VCPE Club at Sloan during my time there and earlier this year Simon Rothman (Greylock) visited Sloan as well — I guess both firms were top-of-mind for me when I wrote this paper. Across the two firms, I’d already reached a reasonable sample size, so I didn’t expand the set to include others (blame the laze).

The result is a list of 47 startups. I further analyzed the business models of these startups and classified them as either B2B or B2C, also collecting information regarding their most recent financing rounds. Below you’ll find a summary of some cool (in a nerdy way, ofc) statistics on the set. Please note that the data collection was done in late April — early May so things could have definitely shifted since then. As an example, Sprig, which was part of the set, is no longer in business.

Cool general statistics on our set

28 startups (~60%) were in the B2B space, whereas the rest were classified as B2C. Interestingly, the Sequoia set was far more heavily split towards B2B, with 12 out of 15 in that space, whereas Greylock companies were evenly split between B2B and B2C.

Pitchbook included data on the valuation for the last financing round for 26 companies. The median valuation was around $46M, with a relatively large standard deviation of $276M. The largest valuation was registered by unicorn Carbon 3D, at $1.1B.

More data was available on the size of the money raised in the last financing round, namely for 42 companies. The median raise was $8.25M, again with a large standard deviation of $38M. Carbon 3D’s $181M raise tops the set.

There was no overlap between the two VC firms in this set. No startup received funding from both Sequoia and Greylock. Sequoia only funded 15 startups that met the criteria, whereas Greylock funded 32 (initially 35, but two companies were closed down whereas one was acquired by Cisco).

Cool talent statistics on our set

I then looked into the employment statistics of these 47 companies. Here, data was taken from two sources: Pitchbook data and LinkedIn data. While far from perfect (I’ll elaborate in the third post of the series why), LinkedIn data is probably more accurate and forms the basis of our analysis, because Pitchbook tends to update its employment data less frequently. The two data sources are rarely massively out of lockstep though.

The 47 companies had a median of 22 employees, ranging from just one employee to 1,153 for Doordash, an on demand meal delivery company. This again results in a large standard deviation of 178 employees in the set.

B2C companies tend to have fewer employees, but there is greater variation overall. The median employee count for B2C companies was 14, whereas for B2B companies it was 38.5. The standard deviations were somewhat different: 91 for B2B and 37 for B2C (excluding Doordash, whose abnormally large employee count skews the data).

There seems to be a clear distinction between the types of companies Sequoia invests in and those that Greylock prefers, when it comes to young Bay Area companies that these firms have funded. In addition to the remark above about a preference for B2B, Sequoia’s portfolio includes fewer companies (15 versus 32), suggesting concentration. However, Sequoia companies are larger in both employee count and funding, perhaps pointing to a preference for later stage financing. The median employee count is just 13.5 for Greylock companies, whereas it is 62 for Sequoia companies (excluding Doordash, a Sequoia investment, doesn’t really change this number). The median last funding round valuation for a Sequoia company is $189M, versus $30M for Greylock companies (where this data is available). The median last funding round amount raised is also substantially different: $24M versus $4.5M.

Please turn over your turnover

Finally, I searched LinkedIn to get a count of the employees that have left the 47 startups in our set. I then calculated the ratio between the number of people who have left and the number of people who work there, as an imperfect proxy for turnover. The median B2B startup had a turnover rate of 28.6%, whereas the median B2C startup had a turnover rate of 50%. A regression analysis suggests the same conclusions: B2C startups have a statistically significant likelihood of having a higher turnover ratio.

In a model with the turnover ratio as the dependent variable and a dummy for B2B and B2C, a consumer-oriented startup’s turnover ratio will be 108 percentage points higher than a B2B startup (significant at 5%) if we include startup company Relcy (which has an abnormally high turnover ratio of 12x) or 49 percentage points higher (significant at 10%) if we do not take into account Relcy.

The data therefore suggests that Bay Area-based, relatively young startups funded by top VCs are more likely to see higher turnover if they are in the B2C space — confirming Hypothesis 2 in the post published yesterday.

Check out the data below visualized in graph form: 11 companies active in the B2B space had a turnover rate between 25% and 50%. Only one B2B company had a turnover of 100%+, whereas a full four B2C companies had this insanely high turnover rate.

Turnover distribution in B2B versus B2C companies

Here’s the deal, though: I can pretty much caveat every statement in the above article, which I will do in the third and final post of this series!