Things to Avoid in Selecting an Executive-Level Job at a Software Startup

This is a sister post to my recent one, Career Decisions:  What to Look For in a Software Startup.  That piece is all about what to look for when considering taking a job at a software startup.  This piece is kind of the opposite:  what to look out for when considering an executive job at a software startup.

This post isn’t simply the inverse of the other and I didn’t approach writing it that way.   Instead, I started blank slate, thinking what are the warning signs that would make me think twice before taking an executive-level job at a software startup.

Before jumping into the list, let me remind you that no startup is perfect and that unless your name is Frank Slootman that you are unlikely to get a C-level offer from a startup that has all eight of the things I say to look for and none of the eight I say to avoid.  The rest of us, to varying degrees, all need to make intelligent trade-offs in facing what is effectively a Groucho Marx problem [1] in our career management.

That said, here’s my list of things to avoid in selecting an executive-level job at a startup:

1. Working for TBH, i.e., working for a boss who is to-be-hired. For example, if a company’s board is leading the search for a new CMO while the CEO slot is also open, the CMO would be working for TBH.  Don’t do this.  You have no idea who the new CEO will be, if you will like them, and whether their first act will be to fire you.  Ignore any promises that “you will be part of the process” in hiring the new boss; you may well find yourself interviewing them as you notice an offer letter sticking out of their backpack, suddenly realizing that you’re the interviewee, not the interviewer.  Read my post on this topic if you’re not convinced.

2. The immediate need to raise money.  Particularly for a CEO job, this is a red flag.  The problem is that unless you are a tier 1 rockstar, investors are not going to want to back the company simply because you’ve arrived.  Most investors will want you to have about a year in the seat before considering investing.  If you’re immediately dispatched to Sand Hill Road in search of capital, you’ll be out pitching the company poorly instead of learning the business and making plans to improve it.  Moreover, to state the obvious, joining a company that immediately needs to raise money means joining a company that’s in the midst of running out of cash.  That means either the company gets lucky and does so (often via an inside round [2]) or it doesn’t and your first quarter on the job will be focused on layoffs and restructuring instead of growth.  Think:  “I love you guys; call me back once you’ve done the round.”

3. Key internal customer TBHs.  For example, the VP of Sales is the VP of Marketing’s key internal customer, so Marketing VPs should avoid taking jobs where the VP of Sales is not in place.  Why?  As your key internal customer, the VP of Sales has a lot of power in both assessing your performance and determining your continued employment [3], so you really want to know if you get along and see eye-to-eye before signing up for a new job.  Moreover, even if you are work-compatible, some Sales VPs like “travel with” their favorite VP of Marketing.  Think:  “Mary’s great.  I just want to work with Joe like I have done at my last two companies.”  Bye Mary.

4. Strategic “traveling” violations.  “Pivot” is one of my favorite startup euphemisms. While many great startups have indeed succeeded on their second try, after a strategic pivot [4], some startups seem to want to make the pivot into an annual event.  Let’s remember that pivots mean strategic failure and the virtual write-down of any VC that went into funding the failed strategy.  While pivots can save a troubled company from continuing to execute a doomed strategy, they’re not something you want to do at all, let alone on a periodic basis.  In basketball, you get called for traveling if you (a) take more than two steps without dribbling or (b) move an established pivot foot.  I call startups for traveling when they (a) do two or more strategic pivots or (b) pivot to a new strategy that has nothing to do with the old one [5] (i.e., moving both feet).

5.  N^th-place Vendors (for all N>=3).  Most high-tech markets have increasing returns effects because customers like to reduce risk by buying from market leaders.  In the early 2000s, these normal increasing returns effects were compounded by network effects [6] in many markets.  Today, machine learning is compounding increasing returns yet again [7].  In short, it sucks to be third in Silicon Valley, it always has, and it’s likely to suck more in the future than it does now.

Therefore avoid working at vendors who are not #1 or #2 in their category.  If you’re considering a #N vendor, then it should be part of it moving to a focus strategy to become #1 at a product or vertical segment.  Don’t get sold the idea that a mega-vendor is going to acquire #4 after being rebuffed by the market leaders or to get a better price.  Mega-vendors greatly prefer to acquire market leaders and recent history has shown they are more than willing to pay up to do so.  Tuck-ins and acqui-hires still happen, but typically for very early-stage companies and not at great valuations.

6. Sick cultures and/or dishonest leaders.  Silicon Valley companies often make a big deal about “culture” but too often they conflate culture with ping pong tables, free lunch, and company parties.  Culture, to me, is the often unwritten code [8] of what the company values and how business gets done.  Alternatively, to paraphrase Henry Ford’s thoughts on quality, culture is what happens when no one is watching.  While many Silicon Valley leaders — going all the way back to HP — are “true believers” trying to build not only unique products but also create unique places to work, there are unfortunately charlatans in our midst.  Some leaders are disingenuous, others dysfunctional, and a few downright dishonest.  If you sense cultural sickness during your interview process, back-checking references, or reading Glassdoor [9], then I’d say tread carefully.

7. Low post-money valuations.  You’ll hear this argument a lot with Nth-place companies:  “well, the good news is we only got an $80M post-money valuation on our last round of $20M, whereas we heard LeaderCo was valued at $240M — so if you come here you’ll start making money off $80M, not $240M.”  At one level, it’s persuasive, especially if you think LeaderCo and NthCo are similar in many respects — “it’s like buying shares at 2/3rds off,” you might think.  But that thinking basically assumes the venture capital market mispriced LeaderCo.  You might justify that position by thinking “valuations are crazy right now” but if LeaderCo got a crazy valuation why didn’t NthCo get one too, raising in the same market?  While some people will try to market low valuations as opportunities, I now see them as problems.

Think not:  wow, what a great arbitrage play.  Think instead:  (a) what don’t I know [10] such that the market priced NthCo at 1/3rd the price of LeaderCo, and (b) what effects that will have on future financing — i.e., it’s likely LeaderCo will continue to have better access to capital going forward.  (Remember, the IPO class of 2018 raised a median of around $300M.)

In olden days, the rule was if the market leader went public at a valuation of $1B, then number two was worth about $500M, and number three $250M (4x, 2x, 1x).  Today, with companies going public later, more access to capital, and stronger increasing returns effects, I think it’s more like $4.5B, $1.5B, and $300M respectively (15x, 5x, 1x).  Given that, and increasing returns, maybe a “crazy” early valuation gap isn’t so crazy after all.

8. First-time, non-founder CEOs.  First-time, founder CEOs are the norm these days and VCs do a good job of helping surround them with a strong executive team and good advisors to avoid common mistakes.  Personally, I believe that companies should be run by their founders as long as they can, and maybe then some.  But when a founder needs to replaced, you get a massive signal from the market in looking at who the company is able to attract to run it.  Back in the day, if you were Splunk, you could attract Godfrey Sullivan.  Today, if you’re Snowflake, you can attract Frank Slootman.

My worry about companies run by first-time, non-founder CEOs [11] is less about the difficulty for the first-timer in transitioning to the CEO job — which is indeed non-trivial — and more about the signaling value about who would, and more importantly, who wouldn’t, take the job.  Experienced CEOS are not in short supply, so if a company can’t attract one, I go back to what don’t I know / what can’t I see that the pool of experienced CEOs does?

That’s not to say it never works — we did a fine job building a nice business at MarkLogic under one first-time, non-founder CEO that I know [11].  It is to say that hiring a non-founder, first-time CEO should prompt some questions about who was picked and why.  Sometimes there are great answers to those questions.  Sometimes, things feel a bit incongruous.

# # #

Notes

[1] Marx often quipped that he wouldn’t want to be a member of any club that admitted him, the rough equivalent to saying that you wouldn’t take a C-level job at any startup that would offer you one.

[2] As one VC friend so tersely put it:  “our job isn’t to put more money into a company, it’s to get other people to put more in at valuations higher than the one we invested at.”  (This somehow reminds me of the  General Patton quote:  “the object of war is not to die for your country, but to make the other bastard die for his.”)

[3] The number one “cause of death” for the VP of Marketing is the VP of Sales.

[4] I particularly like when those pivots are emergent, i.e., when the company is trying one thing, spots that another one is working, and then doubles down on the second thing.

[5] In the sense that they moved an established pivot foot by changing, e.g., both the target customer and the target product.  Changing your strategy to sell a different app to the same buyer, or the same app to a different buyer feels much more like a pivot to me.

[6] Everyone wants to be on the social network that their friends are on, so the more your friends pick network A over B, the more newcomers want to pick network A.  Back when there was competition in consumer social networks, entire high schools went either Facebook or MySpace, but virtually none went both.

[7] Where machine learning (ML) is an important part of the value proposition, you have even stronger increasing returns effects because having more customers, which means having more data, which means having better models, which means producing superior results.

[8] In cases there may be a very public written code about company culture.  But, to the extent the written culture is not the one lived, it’s nothing more than public relations or a statement of aspiration.

[9] While Glassdoor has many limitations, including that reviewers are not verified and that most reviewers are recently-terminated job-seekers (because the requirement to look for a job is to write a review), I still use it in researching companies.  My favorite dysfunctional pattern is a litany of detailed, fact-filled, seemingly sincere negative reviews, followed by a modest number of summary, high-level, HR-buzzwordy positive reviews followed by someone saying “I can’t believe management is feeding positive reviews to people in order to up our ratings.”

[10] An economist friend once taught me that when economists studied established practices in any field, e.g.,  the need for a second-serve (as opposed to just hitting two first serves) in professional tennis, they start out assuming the practice is correct, i.e., that the professionals really do know what they’re doing, and then see if the statistics justify the practice.  One might apply the same philosophy to valuations.

[11] Yes, I was one at MarkLogic.  In terms of signaling value, I was at least CMO of $1B company before starting and while I’d not been a CEO before, I did bring an unusual amount of database domain expertise (i.e., Ingres, Versant) to the party.