EPM: Now More Than Ever

The theme of my presentation at past spring’s Host Analytics World was that EPM is needed in fair, foul, or uncertain weather.  While EPM is used differently in fair and foul weather scenarios, it is a critical navigational instrument to help pilot the business.

For example, in tougher times:

  • You’re constantly re-forecasting
  • You’re doing expense reduction modeling
  • You might do a zero-based budget (particularly popular among recently PE-acquired firms)
  • You’re likely to try and reduce capex (unless you see a quick rebound)
  • You’re probably making P&L, budget, and spend authority more centralized in order to keep tighter reins on the company.

In better times:

  • You model and compare new growth opportunities
  • You often build trended budgets more than bottom-up budgets
  • You adopt rolling forecasts
  • You increase capital investment and build for the future
  • You do more strategic initiatives planning
  • You decentralize P&L responsibility

These (and others) are all capabilities of a complete EPM suite.  The point is that you use that suite differently depending on the state of the business and the economy.

Well, now with the surprise election of our 45th President, Donald Trump, we can be certain of one thing:  uncertain times.

  • Will massive investments in infrastructure (including but not limited to, The Wall) happen and what effect will that have on economic growth and interest rates?
  • Will Trump deliver the promise 4% GDP growth that he’s promised or will the economy grow slower?
  • Will promised deregulation happen and if so will it accelerate economic growth?  What effects will deregulation have on key industries like financial services, energy, and raw materials?
  • What, as a result of this and foreign policies, will be the price of a barrel of oil in one year?  What effect will that have on key industries such as transportation?
  • Will Trump spark a trade war, increasing the price imports and reducing the purchasing power of low and middle-income consumers?  What effect might a trade war have on GDP growth?
  • What impact will all this have on financial markets and the cost and availability of capital?

I don’t pretend to know the answers to these questions.  I do know, however, that there is uncertainty about all of these questions– and dozens of others — that will directly impact businesses in their performance and planning.

If you cannot predict the future, you should at least be able to respond to it in agile way.

If your company takes 6 months to make a budget that gets changed once a year, you will be very exposed to surprise changes.  If you run on rolling forecasts, you will be far more agile.  If you have good EPM tools you will able to automate tasks like reporting, consolidation, and forecasting in order to free up time for the now much more important tasks of scenario planning and modeling.

Again, if you can’t know whether oil will be $40, $50, or $70 — you can at least have modeling out all three scenarios in advance so you can react quickly when it moves.

I’ve always been a big believer in planning and EPM.  And, in this uncertain environment, companies need EPM now more than ever.

A Key Lesson Marketers Can Learn from Donald Trump

While we won’t go into my views on the election here, I will say that all marketers and solution sellers can learn one “yuge” lesson from Donald Trump:  understanding your audience and talking to them in their terms will take you a long, long way.

I’ve always said that solution selling entails getting the customer to conclude three things:

  1. They understand my problem.
  2. They can solve my problem.
  3. I want to work with them.

I put this in reverse form (i.e., calling the company “they”) as a reminder that these are not assertions — they are conclusions.  These are three conclusions that we want the customer to draw.  Asserting them is probably one of the worst ways to get customers to conclude them.  So how might we get a customer to conclude these things?

They Understand My Problem

How might we lead someone to conclude that your organization understands their problem?

  • Hire people who have had the customer’s job and walked in their footsteps.
  • Speak to the customer in their own language about the problem.
  • Active listen to the customer, playing back what they are telling you about the problem.
  • Complete their sentences, saying “and I bet you saw this problem next.”

The ultimate goal is to get the customer to think “Holy Cow, these people might understand my problem even better than I do.”

They Can Solve My Problem

They are several ways to get someone to conclude you can solve their problem

  • Talking about similar reference customers — where similar is defined in the mind of the buyer — whose problems you have solved.
  • Bringing in staff who have worked on solving those very problems.  Telling Pearson, “oh, when we were over at McGraw-Hill we worked on the XYZ system.”
  • Filling in requirements documents but beware that these are often, dare I say “rigged,” by the vendors who got in first as they attempt to set their differentiators on the agenda.
  • Performing a prototype or proof of concept (POC) that shows how key requirements are met using your solution.

I Want To Work With Them

How do you get someone to conclude you they want to work with you?

  • Execute the basics:  show up on time, be prepared, do your homework, communicate status.  (I’m stunned how many people screw up these things and still expect to win.)
  • Be reliable.  Say what you’ll do and do what you say.  Customers want to know they can count on you.  Don’t surprise them.
  • Be personal, build relationships, get to know people, and make them understand you want their business and care about their success.

Back To Trump

Now I have always believed that the first of these tests was the most important:  getting someone to believe you understand their problem.  But Trump has taken my belief in this to a whole new level.

By driving hard on two fronts:

  • A huge dose of “I understand your problem” — with his speeches aimed at a segment of the public who feels unacknowledged and misunderstood, he energizes crowds largely by simply active listening their problems back to them.
  • With a small dose of “I want to work with him” — the whole political outsider, straight-talking guy image.

He has been able to “get the order” from a large number of Americans without providing much detail at all about the second — and one would think rationally very important — point:  the “I can solve your problems” point.  Put differently, I’d say he put nearly 100% of his eggs in the “I understand your problem” basket and virtually none in the “I can solve it” basket (i.e., a huge amount of what and a stunning lack of how when it comes to policy).

This is all more proof that by simply demonstrating that you understand the customer’s problem and by being someone the customer wants to work with, that you can get the order without actually convincing them that you can solve the problem.

In most corporate sales cycles people incorrectly assume all the importance is on the second point — can they solve the problem?  In reality, salespeople and marketers should put emphasis on all three points and on leading the customer to conclude, in this order, that:

  • They understand my problem
  • I want to work with them
  • They can solve my problem

[Reposted and slightly revised post election.]

How to Manage Your First Sales VP at a Startup

One of the hardest hires — and one of the hardest jobs — is to be the first VP of sales at a startup.  Why?

  • There is no history / experience
  • Nobody knows what works and what doesn’t work
  • The company may not have a well defined strategy so it’s hard to make a go-to-market strategy that maps to it
  • Any strategy you choose is somewhat complex because it needs to leave room for experimentation
  • If things don’t work the strong default tendency is to blame the VP of sales and sales execution, and not strategy or product.  (Your second VP of sales gets to blame product or strategy — but never your first.)

It’s a tough job, no doubt.  But it’s also tough for a founder or new CEO to manage the first sales VP.

  • The people who sign up for this high-risk duty are often cocksure and difficult to manage
  • They tend to dismiss questions with experienced-based answers (i.e., well we did thing X at company Y and it worked) that make everything sound easy.
  • They tend to smokescreen issues with such dismissals in order to give themselves maximum flexibility.
  • Most founders know little about sales; they’ve typically never worked in sales and it’s not taught in (business) school.

I think the best thing a founder can do to manage this is to conceptually separate two things:

  • How well the sales VP implements the sales model agreed to with the CEO and the board.
  • Whether that model works.

For example, if your team agrees that it wants to focus on Defense as its beachhead market, but still opportunistically experiment horizontally, then you might agree with the sales VP to build a model that creates a focused team on Department of Defense (DoD) and covers the rest of the country horizontally with a enterprise/corporate split.  More specifically, you might decide to:

  • Create a team of 3 quota carrying reps (QCRs) selling to the DoD who each have 10+ years experience selling to the DoD, ideally holding top secret clearances, supported by 2 sales consultants (SCs) and 2 business development reps (BDRs) with the entire team located in a Regus office in McLean, VA and everyone living with a one-hour commute of that office.
  • Hire 2 enterprise QCRs, one for the East and one for the West, the former in McLean and the latter in SF, each calling only on $1B+ revenue companies, each supported by 1 local SC, and 2 BDRs, where the BDRs are located at corporate (in SF).  Each enterprise QCR must have 10+ years experience selling software in the company’s category.
  • Hire 2 corporate reps in SF, each sharing 1 SC, and supported by 2 BDRs calling on sub $1B revenue companies.  Each corporate rep must have 5+ years experience selling software in the category.

In addition, you would create specific hiring profiles for each role ideally expressed with perhaps 5-10 must-have and 3-5 nice-to-have criteria.

Two key questions:

  • Do we know if this is going to work?  No, of course not.  It’s a startup.  We have no customers, data, or history.  We’ve taken our best guess based on understanding the market and the customers.  But we can’t possibly know if this is going to work.
  • Can we tell if the sales VP is executing it?  Yes.  And you can hold him/her accountable for so doing.  That’s the point.

At far too many startups, the problem is not decomposed in this manner, the specifics are not spelled out, and here’s what happens instead.  The sales VP says:

The plan?  Yes, let me tell you the plan.  I’m going to put boots down in several NFL cities, real sales athletes mind you, the best.  People I’ve worked with who made $500K, $750K, or even $1M in commissions back at Siebel or Salesforce or Oracle.  The best.  We’re going to support those athletes with the best SCs we can find, and we’re going to create an inside sales and SDR team that is bar none, world-class.  We’re going to set standard quotas and ramps and knock this sonofabitch out of the park.  I’ve done this before, I’m matching the patterns, trust me, this is going to be great.

Translation:  we’re going to hire somewhere between 4 and 8 salespeople who I have worked with in the past and who were successful in other companies regardless of whether they have expertise in our space, the skills required in our space, are located where out strategy indicates they should be.  Oh, and since I know a great pharma rep, we’re going to make pharma a territory  and even though he moved to Denver after living in New Jersey, we’ll just fly him out when we need to.  Oh, and the SDRs, I know a great one in Boise and one in Austin.  Yes, and the inside reps, Joe, Joey, Joey-The-Hacksaw was a killer back in the day and even though he’s always on his bass boat and living in Michigan now, we’re going to hire him even though technically speaking our inside reps are supposed to be in SF.

This, as they say in England, is a “dog’s breakfast” of  a sales model.  And when it doesn’t work — and the question is when, not if — what has the company learned?  Precisely and absolutely zero.

If you’re a true optimist, you might say we’ve learned that a bunch of random decisions to hire old cronies scattered across the country with no regard for strategy, models, or hiring profiles, doesn’t work.  But wait a minute — you knew that already; you didn’t need to spend $10M in VC to find out.  (See my post, If We Can’t Have Repeatable Success Can We At Least Have Repeatable Failure?)

By making the model clear — and quite specific as in my example above — you can not only flush out any disagreements in advance, but you can also hold the sales VP accountable for building the model they say they are going to build.  With a squishy model, as my other example shows, you can never actually know because it’s so vague you can’t tell.

This approach actually benefits both sides

  • The CEO benefits because he/she doesn’t get pushed around into agreeing to a vague model that he/she doesn’t understand.  By focusing on specifics the CEO gets to think through the proposed model and decide whether he/she likes it.
  • The Sales VP benefits as well.  While he/she loses some flexibility because hiring can’t be totally opportunistic, on the flip side, if the Sales VP implements the agreed-to model and it doesn’t work, he/she is not totally alone and to blame.  It’s “we failed,” not “you failed.”  Which might lead to a second chance for the sales VP to implement a new model.

Another United Odyssey

Despite being (or perhaps, because of being) a 2 million mile lifetime flyer on United, I generally do my best to try and avoid them.  But once in while, routes being routes, I decided to give them a chance and see if things have improved.

Silly me.

Today’s odyssey begins in Hartford where what looked like a fairly empty flight (from the check-in seatmap) ends up an overbooked mess at 6:30 AM.  The weather in Chicago chips in for some fun and after boarding 25% of the plane, they stop, and ask us to de-board citing a ground stoppage in Chicago.

Being one of the first on the plane I am one of the last off and face this line for a cup of coffee at Dunkins. (You can see the sign way down at the end.)

dunkins

We’re not off to a good start.  About 30 minutes later,  and before I could finish my hard-earned cup of Joe, we board again.  I am constantly using my United app to decide on the likelihood of making my connection.  But the whole time we’re boarding it shows us arriving at 845 and my flight leaving at 910.  Until they close the door.  Now it says 930 arrival, which quickly turned into 950.  Unless my connection is delayed, too — I am toast.  (And the app has let me know is that my flight had already landed from Hawaii hours before and ergo wasn’t going to get badly outbound delayed, due a late in-bound at least.)

So, now, I’m screwed.  If they had just been honest and either told me (or given me the information to conclude) that I had 0% of making my connection, I would have gotten off the plane.  I needed to back on the East Coast later that week and knowing that Chicago is in trouble and San Francisco is also having bad weather, I knew I would be toast if I showed up in Chicago connection-less.

Part of the problem here is misinformation — when my outbound showed no delay, United’s staff response was “they always show no delay right up until they delay it” — which gave me some hope that the outbound would be subsequently delayed and kept me on the plane.

Well, connection-less in Chicago is exactly what happened, after my outbound was delayed twice so I managed to miss it, just two gates down, by about 10 minutes.  Thanks for waiting.

Then I look at the customer service line, which was literally as long as a football field.  (You can’t even see the end of this one.)

united

I don’t think I’ve seen a line that long (and that’s before even entering the roped back-and-forth real line on the right) in a decade.

I slip into the United Club (which for some reason I still pay for) and it resembles a refugee camp at this point.  I fire up my web browser to look at rebookings.  I quickly realize — and this is possibly a good thing — that the United mobile app has way more functionality in this situation than the website which — and this is a bad thing — is basically useless and tells me I need to call and rebook.

I use the United mobile app.

The trick is every flight is full / standby, but one.  So I choose that one, for a whopping 10 hour delay in arrival.  Since I had ponied up for a First Class seat, I figured it would only show me flights with First available.  Nope.  They put me into an economy middle seat on a First Class fare without saying anything, proposing a refund, nothing.  Not a peep.

This continues United’s tradition of basically ignoring whether you paid extra for Economy Plus (where they are so careful to price each seat) or even First Class when things go awry.  All that’s out the window.

I figure First is small so maybe it’s unavoidable.  And then I see that they have been upgrading people into First on earlier flights that I am standby on instead of giving me the ticket that I paid for.  I look at my place on the “complimentary First Class upgrade” line for the flight they have me on standby — and I’m number 36.  But wait, I’m not asking for a complimentary upgrade.  I paid nearly $1000 for a one-way ticket that I needed to buy on short notice.

No one cares.  United Twitter support which promptly offers to help when I first reported my troubles goes silent when I raise this issue.  You feel a mentality of, “whatever, it’s a crisis — you should be happy with any seat regardless of whether we chose to charge you a premium when you bought it.”  I beg to differ.  If you want me to not care at crisis time, don’t care at purchase time.

Then I notice that the 777 is configured 3-4-3 in coach.  The 777 was designed to 2-5-2 but in order to squeeze in an extra seat per row, ever customer-centric United has decided to go 3-4-3 on some models, and of course this is one.

The hits just keep coming.  Everything they do to reduce my experience they have done.

I call my wife and she says to try another airline.  “No, that won’t work,” I say.  “The weather’s the problem and I’m sure if seats were available on other airlines, United would be routing people to them — instead of making them 12 hours late.  I’m sure that every flight on every airline is toast.”

Amazingly, I still give too much credit to United and too little to my wife. About 4 clicks later, I have a bulkhead seat on American.

I arrive 6 hours late having traveled 16 hours door-to-door on an East Coast flight and having needed to purchase a pricey second ticket on American.

The moral is to avoid two things at all costs:  one-stop travel (I should have driven the extra hour to Boston) and United.

Aligned to Achieve: A B2B Marketing Classic

Tracy Eiler and Andrea Austin’s Aligned to Achieve came out today and it’s a great book on an important and all too often overlooked topic:  how to align sales and marketing.

I’m adding it to my modern SaaS executive must-read book list, which is now:

So, what do I like about Aligned to Achieve?

The book puts a dead moose issue squarely on the table:  sales and marketing are not aligned in too many organizations.  The book does a great job of showing some examples of what misalignment looks like.  My favorites were the one where the sales VP wouldn’t shake the new CMO’s hand (“you’ll be gone soon, no need to get to know you”) and the one where sales waived off marketing from touching any opportunities once they got in the pipeline.  Ouch.  #TrustFail.

Aligned to Achieve makes great statements like this one:  “We believe that pipeline is absolutely the most important metric for sales and marketing alignment, and that’s a major cultural shift for most companies.”  Boom, nothing more to say about that.

The book includes fun charts like the one below.  I’ve always loved tension-surveys where you ask two sides for a view on the same issue and show the gap – and this gap’s a doozy.

sm gap

Aligned to Achieve includes the word “transparency” twenty times.  Transparency is required in the culture, in collaboration, in definitions, in planning, in the reasons for plans, in process and metrics, in data, in assessing results, in engaging customers, and in objectives and performance against them.  Communication is the lubricant in the sales/marketing relationship and transparency the key ingredient.

The book includes a nice chapter on the leadership traits required to work in the aligned environment:  collaborative, transparent, analytical, tech savvy, customer focused, and inspirational.  Having been a CMO fifteen years ago, I’d say that transparent, analytical, and tech savvy and now more important than ever before.

Aligned to Achieve includes a derivative of my favorite mantra (marketing exists to make sales easier) in the form of:

Sales can’t do it alone and marketing exists to make sales easier

The back half of that mantra (which I borrowed from CTP co-founder Chris Greendale) served me well in my combined 12 years as a CMO.  I love the insertion of the front half, which is now more true than ever:  sales has never been more codependent with marketing.

The book includes a fun, practical suggestion to have a bi-monthly “smarketing” meeting which brings sales and marketing together to discuss:

  • The rolling six-week marketing campaign calendar
  • Detailed review of the most recently completed campaigns
  • Update on immediately pending campaigns
  • Bigger picture items (e.g., upcoming events that impact sales and/or marketing)
  • Open discussion and brainstorming to cover challenges and process hiccups

Such meetings are a great idea.

Back in the day when Tracy and I worked together at Business Objects, I always loved Tracy’s habit of “crashing” meetings.  She was so committed to sales and marketing alignment – even back then – that if sales were having an important meeting, invited or not, she’d just show up.  (It always reminded me of the Woody Allen quote, 80% of success is showing up.)  In her aligned organization today, the CEO makes sure she doesn’t have to do that, but by hook or by crook the sales/marketing discussion must happen.

Aligned to Achieve has a nice discussion of the good old sales velocity model which, like my Four Levers of SaaS, is a good way to think about and simplify a business and the levers that drive it.

Unsurprisingly, for a book co-authored by the CMO of a company that sells market data and insights, Aligned to Achieve includes a healthy chapter on the importance of data, including a marketing-adapted version of the DIKW pyramid featuring data, insights, and connections as the three layers.  The nice part is that the chapter remains objective and factual – it doesn’t devolve into an infomercial by any means.

The book moves on to discuss the CIO’s role in a sales/marketing-aligned organization and provides a chapter reviewing the results of a survey of 1000 sales and marketing professionals on alignment, uncovering common sources of misalignment and some of the practices used by sales/marketing alignment leaders.

Aligned to Achieve ends with a series of 7 alignment-related predictions which I won’t scoop here.  I will say that #4 (“academia catches up”) and #6 (“account-based everything is a top priority”) are my two favorites.

Congratulations to my long-time friend and colleague Tracy Eiler on co-authoring the book and to her colleague Andrea Austin.

The Era of Consumer Deception:  Why Do We Tolerate Such Price Opacity?

I was wondering the other day why Southwest would spend millions of dollars to remind people that Bags Fly Free.  I’d argue there are two reasons:

  • It generally supports their friendly and transparent, low fees brand strategy
  • It reminds customers that a $500 fare on United might actually cost you more than a $550 on Southwest if you’ve got a few bags

Price have become so opaque over the past few decades that not only are consumers routinely surprised when they receive a bill, but companies now feel compelled to spend millions to remind them that quoted prices are often apples/oranges comparisons.

It’s not hard to find examples of price opacity:

  • Mortgages with variable rate structures people don’t understand and which exposes them to massive increase in payments (i.e., the 2008 crisis).
  • Bank accounts that have no monthly fee, but are laden with subtle and not-inexpensive fees that seem to silently sneak back in as terms are quietly changed.
  • Numerous airline refundability tiers, change fee policies, per-seat premium economy seat fees, and baggage fees that make true price comparison next to impossible.
  • Rental car policies like Hertz’s usurious $10/gallon refueling fee or the maze of overpriced and often unneeded insurance options that can double the price of a rental
  • Teaser rates for many services, including cellular and Internet, that bear no resemblance to the actual monthly fees

Most, but not all, of the time I manage to sidestep these problems because I’m sophisticated and can figure them out (when I take the time), because I carry balances that preclude most of the sneaky banking fees, and because I fly a lot and get exempt from some of the change fees and seat fees.

But just the other day, while I was in the midst of congratulating myself for avoiding the Hertz $10/gallon refueling fee, I looked on the receipt and saw a per-mile fee that nearly doubled the cost of my rental — when was the last time a rental car didn’t have unlimited miles?

It’s a cat-and-mouse game and companies keep getting better at playing it.

Now you could argue that this opacity is a company’s way of fighting back against price competition, and particularly the price transparency and comparability that the Internet brought.  In an era of price comparison engines that scour the Internet for the best deal, why not sneak in some fees that give you an edge?

You can argue, as people often do when it comes to the airlines, that we’ve done it to ourselves – our consumer behavior has trained the companies towards these strategies.  And that may be true, but we need to accept that these strategies are often fundamentally dishonest.

I realized this as my kids got older and I had to explain how rental cars work (which I still don’t know that well apparently), how airfares work (self-insure against cancellation by throwing out a ticket every now and then as opposed to getting gouged on refundable fares – or just fly SouthWest), how credit cards work (that’s a long one), how mortgages work, and on and on.

It’s what in Texas they call a boiled frog problem. It’s happened so slowly and incrementally that we’ve just gotten accustomed to it and the people most hurt by the practices tend to be at the bottom of the socioeconomic ladder (e.g., payday loans) and have the least voice.

And this society of deception already extends well beyond consumer pricing.  Contests and prizes are another huge area, like fake $1M TV show prizes (e.g., America’s Got Talent) that are actually a 40-year annuity worth more like $300K, fake unwinnable TV contests like American Ninja Warrior (which has only been completed twice) which are made harder every year so nobody wins the fake 40-year $1M annuity, or even state lotteries (which started the annuity deception) which typically pay out over 20 years, slashing prize values by about half.

But where we’ve ended up is not acceptable.  Ironically, after the Internet brought a brief wave of price transparency, we have ended with potentially more opacity than we had before as fees and terms and packaging get ever more complex.  We’re eroding consumer trust by living in an era of manufactured confusion and price deception.

You may not think this is a big deal, but I’d argue it’s like Malcom Gladwell’s broken window theory.  If we tolerate constant small deceptions in our lives, we open the doors to the big ones.

Thoughts on the Coupa S-1

It’s been a while since I dove into an S-1 and while I almost never get all the way through, here we go with another quick romp through a recent S-1.

Coupa, a ten-year old company that sells cloud-based spend management software and who pitches “value as a service” (ugh) recently filed its S-1.  Before diving in, I wonder if I should mention the potential irony in a company that sells “spend management” software running with a 40% operating loss.

But, remember the average SaaS business, per research from my friends at JMP, has negative operating margins at IPO time:  the median is -21% and the mean -36%.   So cheap jabs aside, Coupa is running a bit on the high side and, more importantly, in a time where I thought the markets were demanding better profitability than in the past.  That’s the interesting part.  From an operating margin perspective, Coupa is looking like a typical IPO in a market that was supposedly setting a higher bar.

Coupa’s most recent $80M private round put it in unicorn status (i.e., meaning that it was raised at a $1B+ valuation).

Estimating the shares outstanding after the offering is frankly quite confusing (e.g., share counts in the summary P&L seem to not include conversion of the preferred) and after spending 20 minutes trying to figure this out, I think there will be something like 180M shares outstanding after the offering.

Normally that might suggest a reverse split prior to IPO (as Talend recently did, an eight-for-one) but since I can’t find any evidence to suggest that, I’ll have to assume that Coupa and its bankers are bullish on valuation.  Otherwise, if I’ve got the right share count, any valuation less than about $1.8B will put them in single-digit stock price territory (which is the condition companies do reverse splits to avoid).

Highlights from the first pages of the S-1:

  • They connect 460 organizations (customers) with over 2M suppliers, globally
  • They estimate they have saved their customers $8B to date, on a cumulative basis
  • Fiscal year (FY) ends 1/31
  • FY15 revenues of $50.8M, FY16 of $83.7M, 65% YoY growth
  • FY15 net losses of $27.3M, FY16 of $46.2M, 68% YoY growth
  • 1H16 revenues of $34.5M, 1H17 of $60.3M, 75% YoY growth – accelerating

Now, let’s look at the income statement, which I’ve cleaned up and color-highlighted.

coupa1

Income statement comments:

  • Approximately 90%/10% mix of subscription to services, generally good
  • $83.7M revenues in last full FY is appropriate IPO scale by recent historical standards
  • 75% accelerating YoY growth in 1H17 over 1H16 is pretty strong
  • Subscription gross margins running 77% to 80%, pretty standard
  • Services gross margins of -89% in FY16 and -59% in 1H17 are horrific.  Happily it’s only 10% of the business.
  • Overall gross margins run around 60%, which strikes me as a bit low, but according to my JMP data, is roughly on target
  • 1H17 R&D of 25% of revenues, at the mean
  • 1H17 S&M of 58% of revenues, 7% above the mean
  • 1H17 G&A of 17%, 4% below the mean – but after running at a shocking 32% in 1H16
  • 1H17 total opex of 100% of revenues, about 3% above the mean
  • 1H17 operating margin of -39%, about 3% below the mean

They also present a non-GAAP operating loss which I can’t easily benchmark. They define it as:  operating loss before stock-based compensation, litigation-related costs and amortization of acquired intangible assets.  There was about a $12.9M delta between GAAP and non-GAAP operating income in FY16, which reduces to only $3.8M in 1H17.

Back to highlights from the S-1 body copy:

  • They typically do 3-year contracts
  • They say “we rely heavily on Amazon Web Services (AWS)” as a risk in the risk factors
  • 29% of revenues from international in 1H17
  • They had a “material weakness” in their FY14 audit, unusual
  • 25 pages of risk factors in total, normal
  • They’ve raised $165M in venture capital, and have $80M in cash
  • Almost $7M in litigation costs in FY15
  • They claim an estimated LTV/CAC that exceeds 6.0 in each of the past 3 years
  • They do an interesting analysis of their 2013 customer cohort concluding that its contribution margin was -249% in FY13, but 75% in FY14-16. (Page 53.)
  • Average ARR/customer up from $138K in FY15 to $183K in FY16

Here are the quarterly numbers; things look pretty consistent except for 2QF15, where among other things, they had $7.5M in stock-based compensation expense.

coupa2

More highlights

  • Operating cash burn of about $4M/quarter (page 67)
  • The CEO made $660K in cash comp in FY16, $320K and $340K bonus
  • The EVP of sales made $497K in cash comp in F16, $250K base and $247K commissions

Let’s take a closer look at the unicorn round:

  • It raised $80M at $4.18/share (page 119)
  • The beneficial ownership analysis (page 121) is based on 162.8M shares outstanding as of 7/31/16, but I believe excludes 61M shares associated with granted and un-granted stock options (page 43)
  • 162.8M + 61M = 223.8M shares on a fully diluted basis x $4.18/share = $935M
  • Not a unicorn you cry! ($935M < $1B)
  • But remember these claims are usually based on post-money valuation
  • $935M + 80M = $1.015B
  • So the math appears to hold up, but it’s also pretty clear Coupa was holding out for a valuation that squeaked them into the club

# # #

Be sure to read my disclaimers.