Kellblog

Kellblog Predictions for 2016

Posted on December 31, 2015 | 6 comments

As the new year approaches, it’s time for another set of predictions, but before diving into my list for 2016, let’s review and assess the predictions I made for 2015.

Kellblog’s 2015 Predictions Review

The good times will continue to roll in Silicon Valley. I asserted that even if you felt a bubble, that it was more 1999 than 2001. While IPOs slowed on the year, private financing remained strong — traffic is up, rents are up and unemployment is down. Correct.
The IPO as down-round continues. Correct.
The curse of the mega-round strikes many companies and CEOs. While I can definitely name some companies where this has occurred, I can think of many more where I still think it’s coming but yet to happen. Partial / too early.
Cloud disruption continues. From startups to megavendors, the cloud and big data are almost all everyone talks about these days. Correct.
Privacy becomes a huge issue. While I think privacy continues to move to center stage, it hasn’t become as big as I thought it would, yet. Partial / too early.
Next-generation apps like Slack and Zenefits continue to explode. I’d say that despite some unicorn distortion that this call was right (and we’re happy to have signed on Slack as a Host Analytics customer in 2015 to boot). Correct.
IBM software rebounds. At the time I made this prediction IBM was in the middle of a large reorganization and I was speculating (and kinda hoping) that the result would be a more dynamic IBM software business. That was not to be. Incorrect.
Angel investing slows. I couldn’t find any hard figures here, but did find a great article on why Tucker Max quit angel investing. I’m going to give myself a partial here because I believe the bloom is coming off the angel investing rose. Partial.
The data scientist shortage continues. This one’s pretty easy. Correct.
The unification of planning becomes the top meme in EPM. This was a correct call and supported, in part, through our own launch of Modeling Cloud, a cloud-based, multi-dimensional modeling engine that helps tie enterprise models both to each other and the corporate plan. Correct.

So, let’s it call it 7.5 out of 10. Not bad, when you recall my favorite quote from Yogi Berra: “predictions are hard, especially about the future.”

Kellblog’s Top Predictions for 2016

Before diving into these predictions, please see the footnote for a reminder of the spirit in which they are offered.

1. The great reckoning begins. I view this as more good than bad because it will bring a return to commonsense business practices and values. The irrationality that came will bubble 2.0 will disperse. It took 7 years to get into this situation so expect it to take a few years to get out. Moreover, since most of the bubble is in illiquid securities held by illiquid partnerships, there’s not going to be any flash crash — it’s all going to proceed in slow motion, expect for those companies addicted to huge burn rates that will need to shape up quickly. Quality, well run businesses will continue attract funding and capital will be available for them. Overall, while there will be some turbulence, I think this will be more good than bad.

2. Silicon Valley cools off a bit. As a result of the previous prediction, Silicon Valley will calm a bit in 2016: it will get a bit easier to hire, traffic will modestly improve, and average burn rates will drop. You’ll see fewer corporate buses on 101. Rents will come down a bit, so I’d wait before signing a five-year lease on your next building.

3. Porter’s Five Forces comes back in style. I always feel that during bubbles the first thing to go is Porter five force analysis. What are there barriers to entry on a daily deal or on a check-in feature? What are the switching costs of going from Feedly to Flipboard? What are the substitutes for home-delivered meal service? In saner times, people take a hard look at these questions and don’t simply assume that every market is a greenfield market share grab and that market share itself constitutes a switching cost (as it does only in companies with real network effects).

porters-five-forces

4. Cyber-cash makes a rise. As the world becomes increasingly cashless (e.g., Sweden), governments will prosper as law enforcement and taxation bodies benefit, but citizens will increasingly start to sometimes want the anonymity of cash. (Recall with irony that anonymity helped make pornography the first “killer app” of the Internet. I suspect today’s closet porn fans would prefer the anonymity of cash in a bookshop to the permanent history they’d leave behind on Netflix or other sites — and this is not to mention the blackmailing that followed the data release in the Ashley Madison hack.) For these reasons and others, I think people will increasingly realize that in a world where everything is tracked by default, that the anonymity of some form of cyber-cash will sometimes be desired. Bitcoin currently fails the grade because people don’t want a floating (highly volatile) currency; they simply want an anonymous, digital form of cash.

5. The Internet of Things (IoT) starts its descent into what Gartner calls the Trough of Disillusionment. This is not to say that IoT is a bad thing in any way — it will transform many industries including agriculture, manufacturing, energy, healthcare, and transportation. It is simply to say that Silicon Valley follows a predictable hype cycle and that IoT hit the peak in 2015 and will move from the over-hyped yet very real phase and slide down to the trough of disillusionment. Drones are following along right behind.

6. Data science continues to rise as a profession. 23 schools now offer a master’s program in data science. As a hot new field, a formal degree won’t be required as long as you have the requisite chops, so many people will enter data science they way I entered computer science — with skills, but not a formal degree. See this post about a UC Berkeley data science drop-out who describes why he dropped the program and how he’s acquiring requisite knowledge through alternative means, including the Khan Academy. Galvanize (which acquired data-science bootcamp provider Zipfian Academy) has now graduated over 200 students. Apologies for covering this trend literally every year, but I continue to believe that “data science” is the new “plastics” for those who recall the scene from The Graduate.

the-graduate-plastics
7. SAP realizes it’s an complex, enterprise applications company. Over the past half decade, SAP has put a lot of energy into what I consider strategic distractions, like (1) entering the DBMS market via the Sybase acquisition, (2) putting a huge emphasis on their column-oriented, in-memory database, Hana, (3) running a product branding strategy that conflates Hana with cloud, and (4) running a corporate branding strategy that attempts to synonymize SAP with simple.

Some of these initiatives are interesting and featured advanced technology (e.g., Hana). Some of them are confusing (e.g., having Hana mean in-memory, column-oriented database and cloud platform at the same time). Some of them are downright silly. SAP. Simple. Really?

While I admire SAP for their execution commitment — SAP is clearly a company that knows how to put wood behind an arrow — I think their choice of strategies has been weak, in cases backwards looking (e.g., Hana as opposed to just using a NoSQL store), and out of touch with the reality of their products and their customers.

The world’s leader in enterprise software applications that deal with immense complexity should focus on building upon that strength. SAP’s customers bought enterprise applications to handle very complex problems. SAP should embrace this. The message should be: We Master the Complex, not Run Simple. I believe SAP will wake up to this in 2016.

Aside: see the Oracle ad below for the backfire potential inherent in messaging too far afield from your reality.

8. Oracle’s cloud strategy gets revealed: we’ll sell you any deployment model you like (regardless of whether we have it) as long as your yearly bill goes up. I saw a cartoon recently circulated on Twitter which depicted the org charts of various tech megavendors and, quite tellingly, depicted Oracle’s as this:

oracle-org-chart-300x195

Oracle is increasingly becoming a compliance company more than anything else. What’s more, despite their size and power, Oracle is not doing particularly well financially. Per a 12/17/15 research note from JMP,

Oracle has missed revenue estimates for four quarters in a row.
Oracle provided weak, below-expectations guidance on its most recent earnings call for EPS, cloud revenue, and total revenue.
“While the bull case is that the cloud business is accelerating dramatically, we remain concerned because the cloud represented only 7% of total revenue in F2Q16 and we worry the core database
and middleware business (which represents about half of Oracle’s revenue) will face increasing competition from Amazon Web Services.”

While Oracle’s cloud marketing has been strong, the reality is that cloud represents only 7% of Oracle’s total revenue and that is after Oracle has presumably done everything they can to “juice” it, for example, by bundling cloud into deals where, I’ve heard, customers don’t even necessarily know they’ve purchased it.

So while Oracle does a good job of bluffing cloud, the reality is that Oracle is very much trapped in the Innovator’s Dilemma, addicted to a huge stream of maintenance revenue which they are afraid to cannibalize, and denying customers one of the key benefits of cloud computing: lower total cost of ownership. That’s not to mention they are stuck with a bad hardware business (which again missed revenues) and are under attack by cloud application and platform vendors, new competitors like Amazon, and at their very core by next-generation NoSQL database systems. It almost makes you feel bad for Larry Ellison. Almost.

8. Accounting irregularities are discovered at one or more unicorns. In 2015 many people started to think of late-stage megarounds as “private IPOs.” In one sense that was the correct: the size of the rounds and the valuations were very much in line with previous IPO norms. However, there was one big difference: they were like private IPOs — but without all the scrutiny. Put differently, they were like an IPO, but without a few million dollars in extra accounting work and without more people pouring over the numbers. Bill Gurley did a great post on this: Investors Beware: Today’s $100M+ Late-Stage Private Rounds are Very Different from an IPO. I believe this lack of scrutiny, combined with some people’s hubris and an overall frothy environment, will lead to the discovery of one or more major accounting irregularity episodes at unicorn companies in 2016. Turns out the world was better off with a lower IPO bar after all.

9. Startup workers get disappointed on exits, resulting in lawsuits. Many startup employees work long hours predicated on making big money from a possible downstream IPO. This has been the model in Silicon Valley for a long time: give up the paycheck and the perks of a big company in exchange for sleeves-up work and a chance to make big money on stock options at a startup. However, two things have changed: (1) dilution has increased because companies are raising more capital than ever and (2) “vanity rounds” are being done that maximize valuation at the expense of terms that are bad for the common shareholder (e.g., ratchets, multiple liquidation preferences).

In extreme cases this can wipe out the value of the common stock. In other cases it can turn “house money” into “car money” upon what appears to be a successful exit. Bloomberg recently covered this in a story called Big IPO, Tiny Payout about Box and the New York Times in a story about Good Technology’s sale to BlackBerry, where the preferred stock ended up 7x more valuable than the common. When such large disparities occur between the common and the preferred, lawsuits are a likely result.

Many employees will find themselves wondering why they celebrated those unicorn rounds in the first place.

10. The first cloud EPM S-1 gets filed. I won’t say here who I think will file first, why they might do so, and what the pros and cons of filing first may be, but I will predict that in 2016 the first S-1 gets filed for a cloud EPM vendor. I have always believed that cloud EPM is a great category and one that will result in multiple IPOs — so I don’t believe the first filing will be the last. It will be fun to watch this trend and get a look at real numbers, as opposed to some of the hype that gets circulated.

11. Bonus: 2016 proves to be a great year for Host Analytics. Finally, I feel great about the future for Host Analytics and believe that 2016 will be a wonderful year for the company. We have strong products. We have amazing customers. We have built the best team in EPM. We have built a strong partner network. We have great core applications and exciting, powerful new capabilities in modeling. I believe we have, overall, the best, most complete offering in cloud EPM.

Thanks for your support in 2015 and I look forward to delivering a great 2016 for our customers, our partners, our investors, and our team.

# # #

Footnotes

[1] These predictions are offered in the spirit of fun and I have no liability to anyone acting or not acting on the content herein. I am not an oracle, soothsayer, or prophet and make no claim to be. Please enjoy these predictions, please let them provoke your thoughts, but do not use them as investing or business consulting advice. See my FAQ for additional disclaimers.

6 Comments

Posted in Bubbles, Enterprise, Predictions, Silicon Valley, VC

The Great Reckoning: Thoughts on the Deflation of Technology Bubble 2.0

Posted on December 26, 2015 | 12 comments

This post shares a collection of thoughts on what I’ve variously heard referred to as “the tightening,” “the unwinding,” “the unraveling,” or “the great reckoning” — the already-in-process but largely still-coming deflation of technology-oriented stock valuations, particularly in consumer-oriented companies and particularly in those that took large, late-stage private financings.

The Four Horsemen

Here are four key signs that trouble has already arrived:

The IPO as down-round. Square, Box, and Hortonworks all went public at valuations less than their last private financings.

The IPO as last resort. Box is the best example of this, and while I can’t find any articles, I have heard numerous stories of companies deciding to go public because they are unable to raise high-valuation, late-stage private money.

The markdowns. Fortune ran a series of articles on Fidelity and other mutual funds marking down companies like Snapchat (25%), Zenefits (48%), MongoDB (54%), or Dataminr (35%). A unique feature of Bubble 2.0 is publicly-traded mutual funds investing in private, VC-backed companies resulting in some CEOs feeling, “it’s like we went public without even knowing it.”

The denial. No bubble would be complete without strong community leaders arguing there is no bubble. Marc Andreessen seems to have taken point in this regard, and has argued repeatedly that we’re not in a technology bubble and his firm has built a great data-rich deck to support that argument.

The Unicorn Phenom

If those aren’t sufficient signs of bubbledom, consider that mainstream media like Vanity Fair were writing about unicorns and describing San Francisco as the “city by the froth” back in September.

It’s hard to talk about Bubble 2.0 without mentioning the public fascination with unicorns — private tech companies with valuations at $1B+. The Google search “technology unicorn” returns 1.6M hits, complete with two unicorn trackers, one from Fortune and the other from CBInsights. The inherent oxymoron that unicorns were so named because they were supposed to be exceptionally rare can only be lost in Silicon Valley. (“Look, there’s something rare but we’re so special, we’ve got 130 of them.”) My favorite post on the unicorn phenom comes from Mark Suster and is entitled: Why I Effing Hate Unicorns and the Culture They Breed.

As the bubble has started to deflate, we now hear terms like formercorns, onceacorns, unicorpses, or just plain old ponies (with birthday hats on) to describe the downfallen. Rumors of Gilt Groupe, once valued at $1.1B, possibly selling to The Hudson’s Bay Company for $250M stokes the fire.

What Lies Ahead?

While this time it’s different is often said and rarely true, I do believe we are in case when the unwinding will happen differently for two reasons: (1) the bubble is in illiquid assets (private company preferred shares) that don’t trade freely on any market and (2) the owners of these illiquid shares are themselves illiquid, typically structured as ten-year limited partnerships like most hedge, private equity growth/equity, or venture capital funds.

All this illiquidity suggests not a bubble bursting overnight but a steady deflation when it comes to asset prices. As one Wall Street analyst friend put it, “if it took 7 years to get into this situation, expect it to take at least 3.5 years to get out.”

Within companies, particularly those addicted to cheap cash and high burn, change will be more dramatic as management teams will quickly shift gears from maximizing growth to preserving cash, once and when they realize that the supply of cheap fuel is finite.

So what’s coming?

Management changes. As I wrote in The Curse of the Megaround, big rounds at $1B+ valuations come wrapped in high expectations (e.g., typically a 3x valuation increase in 3 years). Executives will be expected to deliver against those expectations, and those who do not may develop sudden urges to “spend more time with the family.” Some CEOs will discover that they are not in the same protected class as founders when these expectations go unmet.

Layoffs. Many unicorns are burning $10M or more each quarter. At a $10M quarterly burn, a company will need to layoff somewhere between 200 and 400 people to get to cashflow breakeven. Layoffs of this size can be highly destabilizing, particularly when the team was putting in long hours, predicated on the company’s unprecedented success and hypergrowth, all of which presumably lead to a great exit. Now that the exit looks less probable — and maybe not so great — enthusiasm for 70-hour weeks may vanish.

Lawsuits from common stockholders. Only recently has the valuation-obsessed media noticed that many of those super valuations were achieved via the use of special terms, such as ratchets or multiple liquidation preferences. For example, if a $100M company has a $300M preference stack and the last $100M went in with a 3x preference, then the common stock would be be worthless in a $500M sale of the company. In this case, an executive with a 0.5% nominal ownership stake discovers his effective ownership is 0.0% because the first $500M of the sale price (i.e., all of it) goes to the preferred shareholders. When people find they’re making either “no money” or “car money” when they expected “house money,” disappointment, anger, and lawsuits can result. This New York Times story about the sale of formercorn Good Technology provides a real example of what I’m talking about, complete with the lawsuits.

Focus will be the new fashion. Newly-hired replacement executive teams will credit the core technology of their businesses, but trash their predecessors for their lack of focus on core markets and products. Customers unlucky enough to be outside the new core business will be abandoned — so they should be careful to ask themselves and their vendors whether their application is central to the company’s business, even in a downturn or refocus scenario.

Attention to customer success. Investors are going to focus back on customer success in assessing the real lifetime value of a customer or contract. People will remember that the operative word in SaaS is not software, but service, and that customers don’t pay for services that aren’t delivering. Companies that emphasized TCV over ARR will be shown to have been swimming naked when the tide goes out, and much of that TCV is proven theoretical as opposed to collectible.

Attention to switching costs. There is a tendency in Silicon Valley to assume all markets have high switching costs. While this is certainly true in many categories (e.g., DBMS, ERP), investors are going to start to question just how hard it is to move from one service to another when companies are investing heavily in customer acquisition on potentially invalid assumptions about long-term relationships and high pricing power.

Despite considerable turmoil some great companies will be born from the wreckage. And overall, it will be a great period for Silicon Valley with a convergence to the mean around basics like focus, customer success, and sustainable business models. The real beauty of the system is not that it never goes out of kilter, but that it always returns to it, and that great companies continue to be produced both by, and in cases despite, the ever-evolving Silicon Valley process.

# # #

Footnotes

This post was inadvertently published on 12/23/15 with an incomplete ending and various notes-to-self at the bottom. While I realized my mistake immediately (hitting PUBLISH instead of SAVE) and did my best to pull back the post (e.g., deleted the post and the auto-generated tweet to it, created a draft with a new name/URL), as the movie Sex Tape portrays, once something gets out in the cloud, it can be hard to get it back.

12 Comments

Posted in Bubbles, IPO, Silicon Valley, Startups, VC

Theoretical TCV: A Necessary New SaaS Metric?

Posted on December 5, 2015 | 2 comments

The more I hear about SaaS companies talking up big total contract value (TCV) figures, the more I worry about The Tightening, and the more I think we should create a new SaaS metric: TTCV = theoretical total contract value.

TTCV = PCV + NPCV

Prepaid contract value (PCV) is the prepaid portion and NPCV is the non-prepaid portion of the subscription in multi-year SaaS agreements. We could then calculate your corporate hype ratio (CHR) with TTCV/ARR, the amount by which you overstate ARR by talking about TTCV.

I make the suggestion tongue-in-cheek, but do so to make real point.

I am not against multi-year SaaS contracts. I am not against prepaid SaaS contracts. In high-consideration enterprise SaaS categories (e.g., EPM), buyers have spent months in thorough evaluations validating that the software can do the job. Thus, it can make good sense for both buyer and seller to enter into a multi-year agreement. The seller can shield contracts from annual churn risk and the buyer can get a modest discount for the contractual commitment to renew (e.g., shielding from annual prices increases) or a bigger discount for that plus a prepayment.

But it’s all about degree. A three-year prepaid contract often makes sense. But, for example, an eight-year agreement with two-years prepaid (8/2) often doesn’t. Particularly if the seller is a startup and not well established. Why?

Let’s pretend the 8/2 deal was written by an established leader like Salesforce. In that case:

There is a very high likelihood the software will work.
If there are problems, Salesforce has major resources to put behind making it work.
If the customer is nevertheless unhappy, Salesforce will presumably not be a legal lightweight and enforce the payment provisions of the contract.

Now, let’s pretend that 8/2 deal was written by a wannacorn, a SaaS vendor who raised a lot of money, made big promises in so doing, and is way out over its skis in terms of commitments.

There is a lower likelihood the software will work, particularly if working means building a custom application, as opposed to simply customizing an off-the-shelf app.
If there are problems, the wannacorn has far fewer available resources to help drive success — particularly if they are spread thin already.
If the customer is unhappy they are much less likely to pay because they will be far more willing to say “sue me” to a high-burn startup than to an established leader.

So while that 8/2 deal might be a reasonable piece of business for an established leader, it looks quite different from the perspective of the startup: three-fourths of its value may well end up noncollectable — and ergo theoretical. That’s why startups should neither make those deals (because they are offering something for an effectively fictitious commitment) nor talk them up (because large portions of the value may never be realized).

Yet many do. And somehow — at least before The Tightening — some investors seem to buy the hype. Remember the corporate hype ratio: TTCV / ARR.

2 Comments

Posted in SaaS, Startups, VC

Tagged ARR, TCV

The Best Work Parable

Posted on November 21, 2015 | 12 comments

I can’t remember when I first heard this great parable, and despite Googling around couldn’t find it online [see footnote], so I thought I’d take a moment to re-tell this pointed story here.

One day an employee is asked to write a proposal for a new business idea and submits it to his manager.

Employee: “Did you get a chance to read my proposal yet? What did you think of it?”

Manager: “You know, I need to ask you one question about that proposal — was it really your best work?”

Employee (reluctantly): “No … , in fact, it was not. I can think of several things I could have done better.”

Manager: “Great, so please do those things and resubmit it to me.”

The employee then does additional work on the proposal and resubmits it to the manager.

Employee: “Hi, did you review my revised proposal? What did you think?”

Manager: “Well, I need to ask you one question about that proposal”

Employee: “Sure”

Manager: “Does the revised proposal represent your best work?”

Employee (reluctantly): “Well, no, while I think it’s much better than the first version, I still have several ideas for how to improve it.”

Manager: “OK, so I’d like to ask you to implement those ideas and then resubmit the proposal to me.”

The employee then revises the proposal again and submits it for the third time to the manager.

Employee: “Did you get a chance to review my proposal? What did you think?”

Manager: “Does this third proposal represent your best work?”

Employee: “Yes.”

Manager: “Great, so now I’ll read it.”

If you’re playing the role of employee, do you submit your best work on the first go? If not, why not? Why do you want your management reviewing low-quality work?

If you’re playing the manager, are your employees getting you to do their jobs for them by having you correct/revise their work into the desired form? How can you set the bar so you get their best work on the first go?

[Footnote: while I couldn’t find this story via Googling several readers were kind enough to inform me that it appears to have been originally told about Henry Kissinger. See here.]

12 Comments

Posted in Leadership, Management

What Marketing Costs Should be Included in CAC Calculations?

Posted on November 18, 2015 | 5 comments

Dear Kellblog:

I’m working on my CAC calculations and I’m trying to determine if I should include all marketing costs or just my direct demand generation costs? I’ve talked to many of my CMO peers and can’t get a consistent answer to the question?

Thanks / Bewildered CMO

Dear Bewildered CMO:

My gut reaction is that you should include all marketing costs. Don’t try to argue that PR and product marketing don’t work on customer acquisition. Don’t try to argue that people aren’t programs and try to exclude the cost of your demandgen team.

Why? Three reasons:

Demandgen people and programs dollars should be fungible. PR and product marketing better be doing things that help acquire customers., even if indirectly.

Playing counting games can hurt your credibility. VCs aren’t just trying to compare metrics, they’re trying to get to know you by seeing how you think about and/or calculate them. I’d think you were a weasel if I found you excluding these costs without really good reason.

To the extent that people try to compare these things between private and public companies, remember that there is no way to split marketing apart (or split customer success from sales) with public companies which should suggest that by default you include things.

Best / Kellblog

For fun, let’s go quickly look at some sources for CAC definitions and see what we find regarding this issue:

Kellblog defines the CAC as:

S&M, by default, needs to include all S&M costs, so you can’t cut anything out.

(Side note: to the extent you amortize commissions, I would prefer to say cash sales expense as opposed to GAAP sales expense, because the latter will hide some costs — but that has nothing to do with marketing.)

The 2015 Pacific Crest Private SaaS Company Survey defines the CAC as:

How much do you spend on a fully-loaded sales & marketing cost basis to acquire $1 of new ACV from a new customer.

This seems to close one door (i.e., you better include IT and facilities allocations to your sales costs — as GAAP would require anyway), but open another because it defines the CAC not in terms of total new ACV, but new ACV from new customers. So if, for example, you had installed base upsell marketing programs, then I would not count those costs in the CAC calculation because they are not marketing costs spent to win new ARR from new customers. Is PR? Is product marketing? It’s a slippery slope. I’m not in love with this definition for that reason. You could never do it for public companies.

David Skok defines the CAC as:

Note that while Skok is calculating a cost to acquire a new customer as opposed to $1 of new ARR, his definition is clear when it comes to splitting marketing costs: include all S&M costs.

Bessemer prefers talking about a CAC payback period and defines it as: