Category Archives: Management

How Should CEOs Answer the Question, “What Keeps You Up at Night?”

I’ve always felt that “what keeps you up at night?” was a trick question for CEOs.

There’s one part of it I’m quite sure about.  There cannot be anything that you control that keeps you up at night.  Why?  Because you’re the CEO.  If something is keeping you up at night, well, do something about it.

Stress, as I like to say, is for VPs and CXOs.  They’re the ones that need to convince the boss about something.  They’re the ones worried about how something might look.  The CEO?  Well, you’re accountable for results.  You get to make or approve the decisions.

If you’re a founder/CEO then you shouldn’t be particularly worried about how things look to the board.  It’s your company.  You’ve got an invisibility cloak that your hired CEO counterparts lack, and which you should use when needed.  Think of founder privilege the way the kitschy classic Love Story described love:  it means never having to say you’re sorry.

For what it’s worth, and I won’t claim to have been God’s gift to CEOs, I lived by the control rule — that is, if I controlled it and it woke me up in the middle of the night, then I was going to do something about it.  That’s why one of the worst things I could say to one of my VPs was, “I woke up last night thinking about you.”  If that happened, and it sometimes did, then either our working relationship or their employment status was changing soon.

I put this in the same “listen to your gut” class as the I don’t want to talk to you anymore rule.  If you’re one of my VPs, then you’re running a key part of my company, then I should look forward to speaking with you each and every time.  If I don’t look forward to speaking with you, it’s a massive problem, and one I shouldn’t ignore.  After all, why wouldn’t I look forward to speaking with you?  Who don’t I like speaking to?  People who:

  • Don’t listen
  • Don’t follow through
  • Can’t keep up
  • Grinf-ck me
  • Can’t or won’t change
  • Are negative
  • Are mean

There are probably other classes, but the point is if I don’t want to talk to someone, it’s a huge signal and one I should dig into, not ignore.

Waking up in the middle of the night is an even bigger signal.  If you agree that CEOs should not wake up in the middle of the night over things they can control, then we can move onto the second category:  things they can’t control.  Should CEOs wake up in the middle of the night over them?

Again I say no.  Why?

Making bets is a big part of a CEO’s job.  Based on available information and working with the team, the CEO places a set of strategic bets on behalf of the company.  The company then needs to execute those strategies.  While the quality of that execution is under the CEO’s control (and should be high to remove execution as a source of noise in the strategy process), the outcome is not.

Why be stressed while the roulette wheel is spinning?  It’s a natural reaction, but does it change the outcome?  You’ve placed your chips already.  Does stressing out increase the odds of the ball landing on your square?  Does not stressing out decrease it?  No.  It changes nothing at the roulette table.

I’d argue that in business, unlike roulette, stressing out can effect the outcome.  A CEO who’s constantly under stress while the wheel is spinning — e.g., waking up in the middle of the night — is likely to perform worse, not better, as a result.

  • A tired CEO does not make great decisions
  • A haggard CEO does not inspire confidence
  • A grumpy CEO does not handle delicate situations well

I’m not trying to minimize the very real stress that comes with the CEO job.  I am, however, trying to provide a rational, contrarian, and hopefully fresh point of view that helps you better frame it.

In the end, there are two types of things that CEOs can potentially stress about:

  • Things they can control.  They shouldn’t stress over these because they should do something about them, instead.
  • Things they can’t control.  They shouldn’t stress over these because doing so will not change the outcome.  Worse yet, it may well change the outcome — for the worse — over the things they can control.

Ergo, CEOs should never stress about things.  QED.

As Warren Buffet said, “games are won by players who focus on the playing field — not by those whose eyes are glued to the scoreboard.”  Focus on what you can control and, as Bill Walsh says, the score will take care of itself.

Congratulations.  You’re the CEO.  You’ve got the best job in the world.  Enjoy every day.  And sleep well every night.

# # #

Notes

  • To reiterate, none of this is to trivialize the stress that comes with the CEO job nor to suggest that CEOs shouldn’t work hard.  It is to say that I believe they will be happier and more effective if they find a way to sleep well — as most senior executives do.
  • To look at this from an outcomes perspective, while I was pleased with the operational results at both companies I ran, I was not particularly pleased with the outcomes.  Did I work hard and obsess about things?  Yes, in general.  If I worried more and slept less do I think it would have improved my outcomes?  No.  Were some of the worst decisions I made in part due to being worried and stressed about things?  Yes.  Did I in general sleep well?  Yes.  I have always naturally focused on running the plays well believed that the score will take of itself.  In my experience, sometimes it does, but sometimes it doesn’t.
  • In writing this post, I found a few anecdotal, fun, and one arguably ironic article on success and sleep.
  • This Bill Walsh quote seems to undermine my argument.  “If you’re up at 3 A.M. every night talking into a tape recorder and writing notes on scraps of paper, have a knot in your stomach and a rash on your skin, are losing sleep and losing touch with your wife and kids, have no appetite or sense of humor, and feel that everything might turn out wrong, then you’re probably doing the job.”  That said, he’d use this as an opener to speeches which were largely about focusing on what you can control.
  • Walsh’s other quote on sleep was more proactive:  “If you want to sleep at night before the game, have your first 25 plays established in your own mind the night before that. You can walk into the stadium and you can start the game without that stress factor. You will start the game and you will remind yourself that you are looking at certain things because a pattern has been set up.”
  • Finally, while we’re on Bill Walsh quotes, I can’t but help to mention my favorite which I wrote about here“There are three types of players:  those who need a kick in the butt, those who need a hug, and those who need to be left alone.”

Traditional B2B Sales is Dead, Long Live the UCE?

In the land of disruption, there’s always something dying and something lining up to replace it, so we’re pretty used to hearing things like “on-premise is dead, long live SaaS.”  Sometimes, they’re right.  Despite the 2008-era views of our resident luddite below, SaaS really did kill on-premises.

Sometimes they’re wrong.  Despite years of hearing, “the data warehouse is dead, long live the data lake,” the data warehouse is doing just fine, thanks.  Snowflake can tell you 60 billion reasons why.

Sometimes, they’re both right and wrong.  Data lakes are doing pretty well, too.  Not everything is zero sum.

You don’t hear this just about technologies, but business models, too.

When the Internet eliminated sellers’ monopoly power over information, I heard, “traditional B2B sales is dead, long live facilitating buying processes.”  This was right and wrong.  B2B sales wasn’t dead, it just changed.  When buyers can get more information themselves and advance further without needing sellers, reframing sales as facilitating buying is a good idea.

When product-led growth (PLG) became the rage, you started to hear it again:  “traditional B2B sales is dead, long live PLG.”  While companies like Atlassian really did dispense with traditional B2B sales, other companies — like Zendesk, Slack, and Twilio — showed the power of blending the two models.  Heck, even Atlassian eventually blended them.

I think of PLG as embracing the continuation of a trend already started by the Internet.  In phase one, buyers no longer needed sellers to get basic product information.  (It’s almost hard to believe, but back in the day, if you wanted even a white paper let alone a demo, you had to talk to a seller.)  In phase two, buyers no longer needed sellers to get hands-on product trial.  It’s the same transformation, just applied to the next two phases down the funnel.

While some companies consider trials customers (and ergo need to count them in churn), I think most enterprise startups should consider trials leads, and the ones who do the right things with the product become leads worthy of passing to sales.  Because they’re qualified by product usage and not marketing actions, they’re called PQLs instead of MQLs.  (Ask my friends at Correlated, or any of the new PLG CRMs, to learn more.)

The other day an old friend of mine, now a highfalutin GM at a big-name software company, forwarded me this article, Traditional B2B Sales and Marketing are Becoming Obsolete.  So, anticipating the content, I donned my “it’s PLG and enterprise, not PLG or enterprise,” gloves and got ready to fight.

But I was surprised.  Instead of saying, “B2B sales is dead, long live PLG,” the article threw me a curveball:

“B2B sales is dead, long live the unified commercial engine (UCE).”

Huh.  The what?

Who wrote this, I think?  Ah, it’s some guy from Gartner.  Before I can add, “and they should stick to IT prognostication,” I see that “some guy” is Brent Adamson, coauthor of The Challenger Sale, one of my top five favorite sales books.

Darn.  Now I have to read this eight-page article and figure out what I think.  The rest of this post is the result.

The Article:  Summary and Analysis
The article argues that it’s no longer enough to try and integrate (in the sense of align) sales and marketing, we should instead unify them.  That’s because buyers have more access to information (including hands-on trials), buyers have access to that information via multiple channels (e.g., vendor websites, review sites), and buyers don’t want to interact with salespeople (which is not exactly new, though he argues that younger people want to interact with sellers even less than older ones).

Sales is thus fighting for relevancy in the buying process and seeking to regain customer access.  The linear model is dead, long live the unified model.  In short:

Helping today’s B2B buyers buy isn’t a sales challenge, nearly so much as an information challenge (or, alternatively, an information opportunity).

He begins with motherhood and apple pie:

The companies that best provide customers the information they most urgently seek, specifically through the channels they most clearly prefer, are in a far better position to drive commercial success in today’s rapidly evolving digital commercial landscape.

He moves into rhetoric to amp things up:

While once a relatively accurate proxy for the underlying buying behavior it was meant to approximate, the serial commercial engine is hopelessly out of date — and dangerously out of sync — with how today’s B2B buyers buy.

With a requisite Gartner dash of profundity:

Today’s buyers are not only channel agnostic in terms of behavior, they’re digitally dominant in terms of preference.

I think that means people like to research shit online before buying it.  Got it.  Stipulated.

I always say that any good sales pitch is 80% tee-up and 20% knockdown.  Now, on the receiving end of such a pitch, I need to advise some caution in that approach.  At some point people want to hear your solution; I’m on page 6 of what’s barely 8 pages and still waiting.  It’s always easier to agree on the problem than the solution (e.g., child poverty, wealth inequality, climate change).  It’s why the 80/20 formula works — you get people agreeing with you, sounding smart, heads nodding, and then you shift to a credible solution that drives your agenda.

But you can’t wait too long to shift to the solution (so I should probably revise my rule to 60/40).  And you should introduce the solution from first principles, not via a case study (which you can always present later, as proof). And if you’re going to introduce the solution via a case study anyway, it shouldn’t be a 1300-person company based in Calgary that I’ve never heard of.

Yet, here I am, about to learn how SMART Technologies found the answer to this pervasive problem by “rebuilding it from the ground up.”  But first, I need to learn about SMART Technologies.  I am now at page 6.75 of an 8.25 page article and still not heard the solution.

The answer:  completely dismantle traditional sales, marketing, success, and service altogether and reconfigure them into a unified commercial engine (UCE).

I’m now thinking:

  • Can you partially dismantle something?
  • Can you completely dismantle something without it being altogether?
  • Where did success and service sneak into things?  While I’d certainly, almost definitionally, want to put all customer-facing teams into a unified engine, how is it that success and service are totally omitted from the argument’s tee-up?

You create a UCE by:

Careful mapping of customers’ buying journeys across a range of predictable “jobs to be done” as part of a typical educational technology purchase.

Never one to miss a gratuitous Clayton Christensen reference, I have to observe that while I am big believer in his work and the jobs-to-be-done framework, I think this is something of a misapplication.  Christensen’s point was about innovation — if you think of products as hired instead of bought, and hired to do specific jobs, then you will anchor yourself in the customer’s point of view when contemplating new products and features.  Think:  not how can we make this milkshake tastier, but how can we make this milkshake more effective when it’s hired as a one-handed commuter breakfast.  What we’re talking about at SMART is simply mapping customer journeys.

When you do that careful mapping, this happens (or, at least, this is what happened at SMART):

Through that initiative the team identified five common buying jobs (Learn, Buy, Order/Install, Adopt, Support) and established an internal team specifically deployed to support each one, reassigning nearly every member of legacy marketing, sales, service, and success staff as a result. In all, over 250 team members received new job designations as part of the process.

You can’t do a re-org these days without creating a center of excellence, so SMART created three:

SMART created three centers of excellence, where they consolidated otherwise duplicative efforts across traditional functional boundaries, one for data and analytics, and one for customer insights and positioning, and one for creative and digital experience.

Those, by the way, sound like a good idea.  I like centralized, specialized support teams, particularly in areas where we’re trying to present one face to the customer.

And then, the re-organization:

Finally, the team then deployed their staff in geographically aligned “pods,” where each pod contains members supporting each of the respective five buying jobs. So, the pod for the southeast United States, for example, is made up of combination of individuals tasked with supporting the entire range of customer jobs from Learn to Support across all relevant digital and in-person channels (including third-party distribution).

In short, run your regions in the USA more like you run countries in Europe.

It’s neither a bad idea nor some insanely different approach.  It does create the need, however, for sophisticated regional leaders who are capable of aligning on both dimensions of the matrix.  Concretely:  is the French country marketing manager part of the French team or the marketing team?  Answer:  it’s a trick question.  The answer is both and they need to learn which way to look, when, as they face managerial decisions — e.g., look to the CMO for questions on messaging and positioning, look to the French country manager when prioritizing campaigns and investments.

I’m going to ignore the end of the article where the VPs of sales and marketing proudly introduce themselves “the former heads” of their respective departments, because they both seem to still work at the company and do something, though the article doesn’t say what their new titles and jobs are.  I’ll assume, hype and semantics aside, that they’ve implemented some sort of functional vs. pod matrix.  As one does with countries in Europe.

Before wrapping up, let me challenge some of the more detailed points in the tee-up.

  • Yes, the machine is by default linear.  But that’s just the first pass.
  • Contacts that don’t make it MQL or SQL get put into nurture and nurture is not linear.  Nurture is a popcorn machine where we dump kernels in, expose them to heat, and over time and in a pretty random order, the kernels pop into recycled MQLs.  I’ve run companies where half of all MQLs are recycled.
  • There’s the question of whether we should nurture people or accounts, as we would in account-based marketing.  Nurturing accounts is definitely not linear, it’s like having one popcorn machine per account.
  • There is no 11th Commandment where God said that nurture shalt be digital only.  While a lot of nurture is automated digital, marketers should remember that a nurture track, broadly defined, can also involve live events (e.g., C-level dinners), dimensional marketing (e.g., mailing a coffee-table book or a Moleskine), and live interactions  (e.g., SDR or AE outreach).  Nurture doesn’t definitionally mean a sequence of emails, nor should it.
  • The part of the linear handoff I detest is when sales “waives off” marketing once an opportunity is in play.  This happens less frequently than it used to, but it reveals a deep lack of trust that should be fixed by destroying walls, not erecting them.

I’ll conclude by saying I think the article misses the most important point in organizational design.  When it comes down the game on the field, who calls the plays and the audibles?  Sure, we have a playbook, and we all know the play we’re supposed to be running.  But things have changed.  There’s a new participant in the meeting.  They mentioned a new competitor who we didn’t know was in the deal.

With a group of talented people, they’ll usually be several different and vocal opinions expressed on how to proceed.  The AE may want to reschedule the meeting.  The SC wants to proceed with the demo.  The consultant thinks we shouldn’t be in the deal in the first place.  The sales manager thinks we can win it because the champion has our back.  What do we do?  Who calls the plays and the audibles that modify them?

In my mind, the person with the quota wins.  As the old joke goes about breakfast:  the chicken is participating, but the pig is committed.  In a world where accountability for results legitimizes decision-making authority, it’s not enough to have a pod/commune and say we can all work it out.  Sometimes we can’t.  And often we can’t fast enough.

Whether you call that person the regional pod leader or the country manager, the role needs to exist and they need to take lead on deal strategy. Everything else is a supporting resource.  Which is why I think marketing alignment with sales is enough.  Yes, we need to collaborate and yes we need smart managers to work in the functional/regional matrix.  Yes, in the case of marketing, we need field marketing to ensure ground-level local alignment.

But do we need to reorganize everything into regional pods?  No.  We just need to work together and be aware that buyers have more information, options, and control than ever before.

# # #

End note
The irony is I ran a company in a pod structure, MarkLogic, where we had vertical pods (aka, business units) where we didn’t have sellers, SCs, and consultants.  We had Federal sellers, Federal SCs, and Federal consultants — all working for a VP/GM of Federal.  Ditto for information & media.  But we did it not in the name of “traditional B2B sales is dead because buyers have more information,” but in the name of a vertical go-to-market strategy where we wanted specialization and alignment.  Pods can work.  It’s all about strategy first and organizational design to support strategy.

Appearance on the Precursive Podcast: The Role of Services in Today’s SaaS Market

A few weeks back, I sat down with Jonathan Corrie, cofounder and CEO of Precursive — a Salesforce-native professional services (PS) delivery cloud that provides PS automation, task, and resource management — to discuss one of my favorite topics, the role of professional services in today’s SaaS businesses.

Jonathan released the 48-minute podcast today, available on both Apple and Spotify.

Topics we discussed included:

  • The Hippocratic oath and executive compensation plans (do no harm).
  • How to frame the sales / services working relationship (i.e., no chucking deals over the fence).
  • Why to put an andon cord in place to stop zero-odds-of-success deals early in the sales process.
  • How to package services, including the risks of tshirt-sized QuickStart packages.
  • How to market methodology instead of packages to convince customers of what matters:  success.
  • The myth of services cannibalization of ARR.  (This drives me crazy.)
  • The alternatives test:  would a customer pay someone else to be successful with your software?
  • Selling mistake-avoidance to IT vs. selling success to line-of-business executives.
  • How and why to bridge “air gaps” between functions (e.g., sales, customer success, services).
  • How to position the sales to CSM “handoff” as à la prochaine and not adieu.
  • The perils of checklist-driven onboarding approaches.
  • The beauty of defining organizational roles with self-introductions (e.g., “my name is Dave and my job is to get your renewal”).
  • The three types of CSMs — the best friend, the seller, and the consultant — and how to blend them and build career paths within the organization.
  • Top professional services metrics.  Caring about (versus maximizing) services margin via compensation plan gates.
  • The loose coupling between NPS and renewal.

Thanks again to Jonathan for having me, and the episode is available here.

Everything I’ve Learned About Recruiting and Interviewing

The other day a founder asked me about interviewing because a candidate had described me as “a great interviewer,” and she wanted to know why. (And for that matter, so did I.)

Emboldened by this seeming endorsement, I dashed off what turned into a lengthy email on interviewing and recruiting, a topic about which I am passionate not because I think I am good it, but because I think I am not.  I find interviewing and recruiting difficult, have made plenty of mistakes over the years, and the consequences of those mistakes are invariably painful.  The wise manager approaches recruiting as a great opportunity to strengthen the organization, but does so with some degree of humility, if not trepidation.

Thoughts on The Recruiting Process
Let’s start by sharing some things I’ve learned over the years on the recruiting process, before we dive specifically into interviewing.

  • Know what you’re looking for.  Most troubles begin here because people fail to ponder and debate what they are actually looking for, so you do the equivalent of walking into Costco without a shopping list.  For example, for a seller, do you require software applications, platform, or data & analytics experience?  What size deals?  To line of business, IT, or both?  For a CFO, do you require a accounting or finance background?  A veteran or an up-and-comer?  A CF-No or a CF-Go style?  You should know the answers to these questions; keep yourself honest by documenting them in a must-have / nice-to-have document.
  • Remember it’s a mutual sales process.  Unless you’re blessed to be at the hottest company in town, always remember that recruiting is a mutual sales process.  That means you need to be selling and filtering at the same time.  Particularly at the end of the process, interviewers should be told whether they should be primarily in “sell mode” or “filter mode.”  As it turns out, the person who said I was a “great interviewer” was a late-stage candidate who saw me in sell mode.  (And, yes, we succeeded with the hire!)  But who knows what they’d have thought of me in filter mode?
  • Follow some methodology or book.  I’m not particularly religious about which one, but I think a common framework helps to ensure completeness and improve communication during the recruiting process.   My private equity friends at ParkerGale, who do a great job of methodology selection, swear by Lou Alder so I’ll plug Hire With Your Head here.  ParkerGale has their own hiring playbook available as well.
  • Use work test samples.  While I’m not big into puzzles with prisoners and lightbulbs, I am a huge believer in having candidates do anything that approximates the work they’ll be doing if they take the job.  Have a product marketing manager give a presentation.  Ask a seller to role-play a sales call.  Have an engineer write pseudo-code to generate the Fibonacci sequence (to see if they understand recursion).  My all-time favorite was giving two FP&A directors the same three-tab spreadsheet with instructions “fix it,” “answer it,” and “model it” to test their attention to detail, problem solving, and modeling abilities.  The two were neck-and-neck on paper and in the interviews, but the exercise revealed a massive difference between them.  (We hired the one whose work stood out and were happy we did.)
  • Check references.  While I suppose the standard process of checking candidate-supplied references is still de rigeur, my favorite reference checks are backchannel and framed not in a binary hire-or-not light, but instead in the light of:  if I were to hire them, what strengths and weaknesses should I expect to see and how should I work with them to get the best results?  This framing tends to produce a better conversation.
  • Consider a try-and-buy.  One way to remove enormous risk from the recruiting process is a try-and-buy:  hire the person as a contractor or consultant, try working together for 3 to 6 months, and if both sides are happy at the end of that period, then convert the candidate to regular employment.  This works for some positions better than others — e.g., fractional CFOs and rent-a-CMOs already exist, whereas fractional CROs and CPOs (product) generally do not.  This works for some situations better than others:  it won’t work when recruiting a veteran CMO out of an existing job, but it can work nicely when considering a between-jobs, up-and-coming VP of Finance for their first CFO role.  Be open, be creative.  I’ve made some great hires this way — and avoided some train wrecks.

Thoughts on the Interview
When it comes specifically to interviewing, here’s what I’ve learned.

  • After chit-chat, ask for a N-minute life story with an emphasis on the why, not the what (i.e., why did you major in X, take first job Y, or move to job Z, as opposed to what you did in each).  For math types, I call this the first derivative of your resume.  I like to time-bound it, typically to 5 or 10 minutes, to see if the candidate has the ability to manage time and summarize accordingly.  I like the first derivative because it provides more information:  I already (largely) know what a PMM or VP of Finance does at a software company.  I’d much prefer to hear why someone chose to work (or stop work) at company X.  Moreover, if I want to understand accomplishments or duties, I can ask that separately, not as part of the life story.
  • After hearing “tough, but fair” for the 100th time, I decided to never ask for philosophies of any type, ever again.  Instead, think about situations that are encountered on the job and ask for relevant stories:  tell me about a time your fired someone, tell me about a time you launched a product, tell me about a time you ran the planning and budgeting process.  The experts call this behavioral interviewing, and it works.
  • Drill, baby, drill.   While I first learned this technique as a way to catch liars and exaggerators (who are frequently ensnared by the details), drill-down questions make fantastic follow-ups to behavioral “tell me about a time” questions.  Example:  tell me about a time you ran a budgeting process?  Drill-downs:  what year was it, in what month did you start, what was the rough total expense budget, how did you define the process, how many budget owners were there, how many iterations did you go through, how did you agree on the sales plan, did salesops have their own model, who made the churn plan, did they properly handle multi-year deals, who was the hardest exec to get on target, what were their objections, how did you handle them, when did the board finally approve it, how many iterations did that take, what were the initial objections, what would you do differently?  I’ve literally started down this path and had people say, “uh, I didn’t actually run the process in that job, but I was part of it” — an important distinction.  Whether to catch embellishment or to better understand candidates, drill-down questions work.  It’s more effective to go ten feet deep on one situation than one foot deep across ten.
  • Consider a panel interview.  I’ve become a huge fan of properly conducted panel interviews.  But first, what a panel interview is not:  it’s not randomly throwing 2-3 interviewers into a room with a candidate with no structure or preparation.  That’s called a romp, and it’s usually a negative experience for everyone.  What I’ve seen work is the following:  after a screening process that results in three candidates who meet all must-have criteria, you appoint a lead interviewer to create 5 behavioral questions (based on expected job duties in the first 12 to 18 months), share those questions with the candidate in advance, and then run a 90-minute live interview with a panel of 3-5 members who largely listen and ask follow-up questions only.  You create a scoring rubric, have all interviewers complete it, and then conduct a live discussion to compare the candidates.  This is FIRE.  In theory any of three candidates can do the job, so you’re focused on picking the best one for the company and situation.  The panelists listen intently because they’re not worried about running the interview, the remaining time, or their next question.  All candidates are asked the same questions.  And then you debrief via a live discussion which, as much as I love technology, is far higher bandwidth than any collaboration mechanism.  And you avoid groupthink because the rubric has been completed in advance.  Fire.  I thank ParkerGale for teaching this technique to me; they have a Private Equity Funcast episode on how they approach hiring here.

A CEO’s Guide to Marketing: My SaaStr 2021 Official Video

I’d previously posted a video of my SaaStr 2021 presentation, A CEO’s Guide to Marketing, but it was a bit of hack (a link into the stage stream) done favoring time-to-market over production values.  In this post, I’m embedding the official SaaStr 2021 video of that presentation, which has improved production values.

Here’s the video:

And here are the slides: