Author Archives: Dave Kellogg

My Thoughts on the SVB Meltdown

(Revised 8:56 am 3/19)

Looks like I picked the wrong week to be off-grid in Argentina.

When I came back on-grid last night, I quickly discovered that the world, or more precisely, my Silicon Valley business world, had basically exploded while I was flyfishing in Patagonia.

A few weeks ago there had been talk of a mass extinction event for startups in 2023.  It was about funding, not banking, and the prediction was for the second half of 2023.  But perhaps it had come early and for a different reason.

Instead of writing yet-another explainer article, I’ll do two things:

  • Provide links to the best explainer articles I’ve found thus far
  • Share some of my own views on the situation, reminding readers that I am go-to-market person and former CEO (and not a finance person or former CFO)

The Best Explainer Posts I’ve Found

My Personal Views on the Situation

I’ll quickly share my personal views on the situation here:

  • Almost every company I work with uses SVB.  They are the default startup bank in Silicon Valley.  Many keep all their cash there because it’s a fairly standard term of an associated venture debt loan.  If depositors lose their funds I believe large numbers of startups could fail, eliminating the thousands of jobs that they provide.  The Alderaan scenario.  I think it’s unlikely, but absolutely must be avoided.
  • Startup death is a natural part of the Silicon Valley ecosystem, the Darwinian process that produces the innovation that drives a large part of our economy.  Startup death is a natural part of the process — but it should result from a bad idea or a unworkable product.  Not from your bank failing.
  • There is a blame game with three primary parties involved:  VCs for provoking the bank run, the Fed for raising rates (which devalued SVB’s long bonds), and SVB for putting themselves in an weak position.  Who you blame seems to say more about you than the situation.  People who like SVB blame the Fed.  People who dislike VCs blame them.
  • Answering the question “what happens to us if rates go up?” seems absolutely core to the operation of a bank.  (Think:  it’s what we do here.)  SVB put themselves into a situation where the liquidity rumors couldn’t be easily dismissed.  Yes, VCs likely provoked the bank run, but SVB put themselves in a place where they couldn’t stop it and bungled communications on top of that.
  • You cannot overstate the interconnectedness around SVB.  I know startups with all their money there.  I know VCs who are unable to provide bridge loans to startups because all their working capital is also at SVB.  I’ve heard of founder/CEOs who have all their personal money there as well, so they are unable to even use their own funds to bail out their companies.  The single worst story I’ve heard is a startup who had all their money in SVB successfully arranged a loan to cover payroll and wired that money to their payroll provider … who then put it in SVB.  Additionally, startups often sell to other startups, so the web is intereconnected not just across investors, but companies and customers.
  • SVB’s depositors must be protected.  I’m not talking about bailing out SVB investors or management.  I’m talking about protecting depositors, thousands of startups, the jobs they provide today, and their potential to become world-leading tech companies  — the next Oracle, Cisco, or Salesforce might be killed off if we don’t.

Personally, while I’m not an expert in banking, I am uncharacteristically optimistic because SVB owns plenty of high-quality assets and, as mentioned above, those assets exceed deposits in value (though that is a function of valuation method as discussed in the Rubinstein article).

They are not sitting atop a pile of incredibly complex, thinly-traded derivatives (e.g., CDOs, CDO swaps).  They are sitting atop a pile of long government bonds.   This is not 2008.  SVB is not Lehman Brothers.  Because of this, I think there is a good chance that someone acquires them this weekend (or soon thereafter), finding opportunity in SVB’s wreckage and ending this industry-wide liquidity crunch.

Let’s hope so, at least.

Is a Dream a Lie if It Don’t Come True? Founders, Aspirations, and Company Potential

“Is a dream a lie if don’t come true?” — Bruce Springsteen

The River was one of my favorite songs in college and whenever I listen to the above line near the end, I start thinking about Silicon Valley.

Consider three entrepreneurs.

Founder 1:  Elizabeth Holmes.  Was she simple con artist who chose fraud over business failure or a broken visionary trying to walk in the footsteps of Steve Jobs?

Founder 2:  Adam Neumann who dressed up the equivalent of Regus as a tech company and successfully raised money at valuations up to $47B before flaming out on the approach to an IPO.  (Now seemingly running a similar play with Flow.)

Founder 3:  Joe, our friend at BigCo who quit his VP-level job to found a company, spent 10 years sweating it out, pivoted, recapped, and finally threw in the towel for a carve-out in a $30M sale that didn’t clear the preference stack.

Which are they?  Were they dreamers or liars?

To try and sort that out, I’d consider three questions:

  • Did they truly believe in the dream?  It’s hard to know what anyone truly believes [1] — and we need to separate visionaries from lunatics [2] — but in many situations you can develop a sense for whether someone is a true believer or a poser.  This one’s hard to assess, but important.
  • Were they lying about progress?  While there is a small gray zone of exaggeration, misunderstanding, and embellishment [3], for the most part this one is black and white.  Were the numbers real?  Was the demo faked?  Were the milestones hit?
  • Were they making big money before realizing the dream?  This didn’t used to be possible in Silicon Valley, but a side effect of the recent financing environment [4] was the rise of secondary sales that made it possible for founders to reap 10s to 100s of millions before a liquidity event that shared success more broadly across investors and employees.  Situations where a founder can make “done” (or “lifestyle changing”) money before realizing the dream can present the potential for conflicts of interest.

We ask a lot from founders.  And what we ask is often in diametric oppposition.   We ask founders to be:

  • Unreasonable, but reasonable.  Founding a startup against long odds is an inherently unreasonable thing to do.  But, aside from that, we want them to be reasonable people.
  • Optimistic, but realistic.  We want them to believe they can accomplish the nearly impossible, but be realisitic in setting goals and operating plan targets.
  • Big-picture, but detail-oriented.  We want them to create a disruptive, big-picture vision of the market, but be able recite SaaS metrics from memory.

This alone is a good reason to have both co-founders and a strong executive team.  While F Scott Fitzgerald said, “the test of a first-rate intelligence is the ability to hold two opposed ideas in the mind at the same time,” it’s hard to do so in every situation, all the time.

While plenty has been written about Holmes and Neumann inspired a TV series, nobody talks much about Joe.   And there are a lot of Joe’s out there.

Joe’s employees and investors are disappointed in him and might be mad at him.  After all, Joe probably said:

  • The company faced an amazing opportunity
  • The space had the potential to produce a public company
  • He believed they could beat BadCo and WorseCo to win the market

But what did we want Joe to say?   Yes, Joe needs to be careful.  He needs to speak precisely and make precise claims.  He needs to hedge his language and not make promises.  But Joe is not only allowed to be optimistic, it’s his job.

As I’ve often said,

“As CEO, even if you’re standing neck-deep in shit, you need to be looking at the stars.”

Put differently, while the CEO needs to be aware of the company’s situation and have credible plans to address it (i.e., the neck-deep part), they must always be focused on and believe in the potential of the company [5].  The day they can’t do that is when they should turn in their badge.

So, going back to Springsteen, “is a dream a lie if it don’t come true — or is it something worse?”

# # #

Notes

[1]  I’d argue it’s sometimes easier to know when they don’t — e.g., if they’re telling all their friends they’re running a scam.

[2] For example, the difference between trying to emulate Steve Jobs and thinking that you have been sent by God as the reincarnation of Steve Jobs.

[3] Larry Ellison reportedly once said, “sometimes I just get my verb tenses mixed up” when speaking about product capabilities vs. roadmap.

[4] E.g., higher valuations, longer time to liquidity, higher bar on IPOs.

[5] My take on what’s known as the Stockdale paradox.

The Key to Making Market Research Actionable, Part II

In the first part of this two-part series we discussed the importance of timing in ensuring that market research is actionable.  The moral was to time the arrival of research (e.g., win/loss reporting, NPS surveys, awareness and marketing funnel studies) with the cadence of your company’s quarterly and annual strategic goal setting process.

Research that arrives asynchronously gets read (if you’re lucky) and then forgotten.  Research that arrives synchronously becomes a homework assignment for the meeting and a session on the agenda.  That way, its findings are top-of-mind when you sit down to decide priorities and hammer out OKRs.

In part II, we’ll take a more strategic look at the question.  Ultimately, to make market research actionable, you need to ensure five things:

  • Good timing.  It must show up at a time when you’re ready to absorb and action it.  The subject of the first post in this series.
  • High relevance.  It needs to help answer your most important strategic questions.
  • Action-oriented framing.  Work to ask questions in a way that provides action-oriented answers.  You can ask, “do you have plans to move to the cloud in the next five years?” or you can ask, “do you plan to move to the cloud in the next year, and if so, what are your top three evaluation criteria?”
  • Time for consensus building.  You can’t just spit out the answer from a black box.  At each stage of the process, you want to have discussion and get buy in, so that when the end is reached people feel the process was valid and buy into the conclusions.
  • A qualitative component.  Quantitative answers what, but not why.  Qualitative can lead to understanding why.  Pair surveys with interviews for this reason.

Put differently, as my friends at Topline Strategy say, market research that gets turned into action is market research that was designed from the outset to be actionable.

Let’s drill into relevance and action-orientation a bit more.  To ensure you’re asking relevant questions, you should do two things.

First, create what I call the hypothesis file.  This is a file where you write down, over the course of the year, every time you hear an assertion or a hypothesis that you’d love to validate.  Examples:

  • The problem is nobody’s ever heard of us.  We’re just not seeing enough deals.
  • The issue is we’re not making the short list and that because we’re not seen as a leader.
  • We can’t sell use-case A because we’re seen as weak on features 1, 2, and 3.
  • The leakage point in our funnel is demo.  We lose too many deals there and that’s because of our UI.
  • We’re not speaking to the business buyer’s priorities.
  • Everyone’s tired of talking about (e.g.,) data culture, we need a new message.
  • If we just focused on BigCo replacements, we could do the numbers on that alone.

These are rarely offered as hypotheses.  They’re usually statements, often presented as self-evident facts.  You need to tune your ears to hear them and write them down.  Don’t fight every one in real time.  But be keenly aware that these are the foundations of your internal corporate mythology — and it’d sure be nice to know if they’re true or not.

When it’s time to do a market study, review the questions in the file and decide which ones you want answered.  Picking the hottest questions will guarantee that people will be champing at the bit to see the results.

Second, to ensure high relevance, try to identify the core strategic questions you’re facing, whether they appear in the hypothesis file or not.  Such as:

  • In which segment are we really most successful, not just at winning deals but renewing and expanding them?
  • If our market is transitioning to a platform, what are the key elements that must be included?  Where do customers want us to partner so they can buy best of breed?
  • How can we easily regain some product differentiation that matters to customers with our limited R&D capacity?
  • Is our Microsoft partnership a key asset on which to double down or a liability to unwind?
  • Will our category be entirely absorbed into a broader suite or can we sustain a moat of product differentiation to protect us?

By debating these questions, and which ones to include in the study, you again guarantee a high degree of interest in learning the answers once they are available.

To ensure you’re asking action-oriented questions, as you build the study, question-by-question you need to ask yourself, “what would I do differently based on the answer?”

  • For multiple choice, what would I do differently if the majority answer were A vs. C?
  • For rankings, what would I do knowing the top three ranked choices were 123?  Or the bottom, 789?
  • For progression questions, what would I do if I notice everyone was dropping out at stage 3 of our funnel?

Sometimes, the answer is ask more questions.  So build follow-up and drill-down questions into the survey.  Sometimes, the answer is you’re circling the wrong question.  Think:  we asked a lot of questions that will help us determine the TAM.  Say we conclude it’s $20B, then what?  I think we’re asking the wrong question, I don’t want to know what the TAM is, I won’t to know the velocity with which it’s coming to market.  Ultimately, how many deals will be happening in the space next year and how many of those do we need to participate in to make our numbers?

Simply put, you can research how big the TAM is, or you can research how much of it is coming to market next year.  The latter is a lot more actionable than the former.

So let’s wrap up.  If we want to ensure that our market research is actionable, then we need to:

  • Time its arrival
  • Study the right questions
  • Ask those questions in a way that provides action-oriented answers

The Key to Making Market Research Actionable

Ever heard any of these?

  • “Yes, we run a quarterly customer satisfaction (CSAT) and NPS survey, but I feel like we don’t really take any action based on it.”
  • “Win/loss reporting, well yes, we do it, but I don’t think it’s very effective and I don’t think we do much in response to the data.”
  • “Wow, we ran an amazing overall market study last year, but I can’t think of a single strategic change made as a result of the findings.”

What’s wrong here?  Why are we going to the trouble of doing good market research, but not making any use of it?

Sure, sometimes it’s politics or change resistance or the ostrich effect.

But more often than not, in my experience, the reason is simple:  timing.  Companies know they should do this research.  They know their peers do.  They know it’s a best pratice.  So they do it.

But it arrives asynchronously.  Like the win/loss report that arrives the week after the ops review.  Or the strategic analysis that arrives six months before the strategy offsite.  Or the CSAT survey that arrives arrives mid-quarter.

When these deliverables arrive asynchronously, people do their best to read them.  They share them on Slack or email and make a few interesting observations.  But then they forget them.  There is so much other data.  And so much else to do.

How do we fix this?  Simple.  Fix the timing.  If your company has a a quarterly business review (QBR) where next-quarter OKRs are discussed and assigned, then ensure the CSAT report arrives days before that meeting.  Better yet, make the CSAT report review a standing item on the QBR agenda.  Along with a review of the win/loss report.  Time the annual state-of-the-market study so it arrives a week or two before the strategy offsite.

Don’t fight your company’s cadence.  Instead, slide into it.  Design the surveys to be actionable and time them so they can be.  Work with your vendors to move to new timing.  Otherwise, you’re spending $50K to dump a report into a drawer (or a PDF into a shared drive).

Market research doesn’t improve with age.  Build it and time it appropriately, and you’ll find that it’s a lot more actionable than you might think.

For the second part of this series, see The Keys to Making Market Research Actionable, Part II.

(Expanded to two-part series on 2/14/23)

Appearance on the SaaS Revolution Podcast

SaaStock recently released an interview with me on their podcast, The SaaS Revolution Show.  The interview, conducted by SaaStock founder and CEO Alex Theuma, was notionally about the Balderton Founder’s Guide to B2B Sales that I published late last year.  While we ended up discussing that, we also covered a whole lot more, including:

  • My background as a CMO, CEO, and independent director
  • My work with Balderton as an EIR 
  • Which job I prefer, and why:  CEO or CMO
  • Why we made the Founder’s Guide to B2B Sales
  • Key takeaways from the guide
  • The transition from founder-led sales (FLS) to sales-led (SLS)
  • When to hire your first sales executive or leader
  • Why it’s important to define process (and metrics) early — before you need to
  • The Holy Grail of a repeatable sales process
  • Why salespeople are like airplanes (they only make money when they’re in the air)

If you’re interested in listening to the episode, you can find it here.

I’ll see you at SaaSstock USA in Austin this June where I’ll be talking about conversation intelligence, inspired by my work with Jiminny, a UK-based startup where I sit on the board.