November 19, 2019
Building Vibrant Developer Communities
We brought together a group of developer relations experts to discuss how they've built and maintained vibrant, active developer communities...
In episode 21 of EnterpriseReady, Grant interviews Michel Feaster, CEO and Co-Founder of Usermind. They discuss Michel’s journey from gas station employee to enterprise software CEO, as well as the relationship between product, product marketing, and marketing.
About the Guests
Michel Feaster is the Co-Founder and CEO of Usermind, responsible for company vision, strategic direction, planning and execution. Michel’s enterprise software career spans almost 20 years with roles in sales, products, strategic marketing and general management.
Grant Miller: All right, Michel. Thank you so much for joining us.
Michel Feaster: My pleasure. Thanks for having me.
Grant: I think it'd be amazing to start off just with a little bit of your background and how you got into enterprise software.
Michel: Now it's 20 years in enterprise software, I think I
have a nontraditional background.
I dropped out of college, and believe it or not I was working at a gas station for a long time.
I ended up working my way up from doing overnights
at $4.50 an hour to managing a bunch of gas
stations, and I actually got my break into tech in
It was the boom when there was not enough
qualified people to support tech
companies in so many roles, services or sales,
When my partner at the time was selling inside software and called me up and said, "There's this opportunity at Compuware and it sounds just like you.
It's for a pre-sales role.
'Thinks quickly on your feet, likes problem
solving, good communicator.'
You should really apply."
So I ended up applying to this
job, and it was one of those programs, they called it the "Professional development
program," where they literally moved you to Michigan for three
months and taught me to code, and taught you to pitch, and
taught you to do SQL and build
It was my first time using e-mail, if that
dates me. That was a big pivot point in my
I got out of this program and worked for them, ended up
And of course, for people who don't know Compuware, it
was a billion dollar mainframe software company and they tried
The owner of it, by the way, owns the Detroit Red Wings.
They are a pretty famous entrepreneur in Michigan.
I joined them when they were pivoting into client server, and trying
to expand their offerings from just
mainframe into the emerging client server and web tech
I got this role doing pre-sales, and they gave me this boring product line, this testing product line.
Of course, you had to you had to pay them back for their investment.
I was supposed to work for 18 months at a dramatically reduced
salary, and I went into the field.
I moved back home to New England and started taking deals from the number one player in the market, which was a company called Mercury Interactive.
That led to me being headhunted by
Mercury, and eventually I joined Mercury, which
was this incredible opportunity in my
I ended up spending seven and a half years there, and the
first half of my time at Mercury--
That's an enterprise software company for people,
viewers or listeners who don't know.
Mercury was a company that started out doing testing
products, eventually went into monitoring, and was in the
space of competing with BMC and
HP and its operational management
So I joined those guys, and man what a what a
For the first four years I was in pre-sales, and I
got to go through this really incredible growth process where we went
from doing $75K deals with anybody
who talk to us, to by the end I was a strategic
pre-salesperson and I had the biggest accounts in the company.
Like Liberty Mutual and GE, and I don't know if you've been through that growth process in a software company, but as you go from selling transactional deals at $75K or $100K deals, to strategic accounts where you're trying to get seven figure deals, the way you sell changes.
A very big part of what I was doing as a pre-salesperson wasn't just technically proving software anymore, it was building these multiyear roadmaps.
I ended up for the customer, right around "What does a
partnership with Mercury look like?
A strategic relationship."
I ended up working really closely with the product organization on not just pulling them in to speak to my customers about roadmaps, but eventually influencing a couple acquisitions that Mercury did prior.
Most notably the Cantone acquisition, and
somewhere in there one of my mentors from corporate called me up and
said, "Why aren't you in product?
You have such a-- You've worked with these customers, you can
articulate where you think the company should be going, you have a vision
for how the strategy should look over the next couple
Of course, in my life, I don't know about you but quite often I say "No" before I say "Yes."
I actually originally said "No," and I remember I was sitting in my
house and I was reading a book about Hemingway one day,
and there's-- I'm on page four and Hemingway's
There's these couple lines of text where he
literally is 18 and gets in a boat and sails to France and has no money.
And I thought to myself, "If he can do that--"Right? Then he became Hemingway. So I was like, "What am I afraid of?"
And I called them back and said, "I'm going to want to take the job."
That was probably the next great real pivot point in my life.
I ended up moving to California, I was just
so lucky in retrospect.
The very first product management job I was given for two reasons.
One, I was given a product called Load
At the time it was a $250 million dollar business for Mercury. It was the biggest product line at the company.
We probably had 70% market share.
Our customers loved this product.
It was a load testing product, and it's still around.
HP owns it now, but it was just this dominant product in the
So imagine being given a product that works, that customers love, if you want to go meet with them and pull out requirements and interview them, everyone wants to talk to you.
I can't even imagine a more perfect first product
The other interesting thing at the time, at Mercury we didn't have a split between product management and product marketing.
When I joined, I was this mini-CEO.
I owned everything from talking to engineering about-- I'm
going to date myself, the MRD.
It was an on-prem software and our release cycles were like
nine months to a year.
I did everything from all the
inbound product management work, to I did all the sales training and all the sales
enablement, and all the product marketing content and form sheets and training.
Like I said, I think it was the most perfect first role you could possibly
It took me a bit of time, in pre-sales you get used to getting wins real fast. At the time, product management was an exercise in patience.
You might have a vision, but it takes a year to release
software. It may be another six months to know if your ideas were
So, that was a huge adjustment for me, but man it was my
I ended up just falling in love with going out,
interviewing customers and being a detective
and figuring out what people really wanted or
I got super lucky, spent another four years in the product loop at Mercury, all on the QA side.
We got acquired by HP Software, and of course that was a super sad day. But as so often happens, new doors opened for me.
My boss called me up and said, "There's this really
troubled business within HP's datacenter automation,
but it's in focus in the field.
It's a great career opportunity, you should go do
So I took over this business and I ended up building a business
case to buy a company called Opsware, which at the
time was $100 million dollar run rate.
The founders of that company were Marc Andreessen and Ben Horowitz, who are now famous, obviously, for founding Andreessen Horowitz the venture capital firm.
But at the time, they were just famous because Mark invented the browser. They were already famous, but a previous chapter.
So Ben came in and became my boss at
HP Software, and he's the reason I ended up in startups and ended up being a
He said to me, "You're this crazy product person. What are you doing working in a big company? You should just go build your own company."
I remember at the time, laughing in his face and saying, "I've
never been at a startup.
I don't even know what it would be like to work at a company that was 20
people. I joined Mercury at a couple hundred people and I stayed through
So when he left and it
was not public yet that they were doing Andreessen Horowitz, one of their very
first investments was a company called Apptio.
I knew the CEO, he had worked at Mercury.
My other mentor, my boss, the head of products at Mercury
was on the board.
So I went in and interviewed for a VP
of products role and I ended up joining that company.
I was employee 17.
I tell everybody that was my college education or my
It was incredible.
From when I joined $500K of
revenue to $70 million when I left, and
IPO'd, and has since been picked up by private equity.
But that was my first chance to join a company and build a category from scratch.
I ran, again, both product management and product marketing there.
That is my passion, the intersection between product and narrative.
It was just an incredible experience to get to take an early product with a very small number of customers and figure out how to tell a story about it.
My passion was creating categories, really
telling and carving out a different market space in the market
and getting analysts to agree and recognize that.
And then obviously, building and delivering the product to support it.
That was just an incredible experience, doing our
early first product advisory boards and then building a community
and seeing that be successful.
It was pretty clear at the time, I'd been there three and a
half years, and it was really clear the company was going to go public.
For people who haven't prepared for that, quite often what you're trying to figure out how to do is find what's called another chapter of the company.
Just "What's your next act that's going to propel growth
from 100 to 500, or from 100 to a
I think at that
time, the CEO and I just ended up with different points of view on what
would be next.
I'm not super good at
executing plans I don't support, so I ended up leaving and really
wasn't sure "Did I want to found, or did I want to do another product
Because I just genuinely love being a product leader, and I love interviewing customers and then going back and showing them something and watching their eyes light up that we found something amazing for them.
So, I probably did a couple months just to network and talk to
people, and if I was netting it out, I just didn't find anyone I really wanted to
I said to myself, "Why don't I want to work
for any of these people?"
And the answer was "Because I think I would have a better plan
That was my clue that I probably shouldn't go work for anyone and I should found my company.
When I realized that, that was when I started my journey toward
Usermind. At this point I'm thinking, "A lot of people who
found had worked on their idea for years."
Of course, I wasn't planning that and hadn't decided that, so I literally just sat down and made a list of things that I thought were changing in the world.
All of which, by the way, seem very obvious now.
The business will buy more software than IT going forward,
and it will all be SaaS and it'll all be subscription and everything will have an
API. Long story short, I thought
"There's going to be little IT teams that form in marketing and
sales, because it's actually quite hard to manage software even if you're not managing
So I literally just went out and did a bunch of interviews. I probably talked to 300 people over four months, because I thought there was a persona that was underserved, and I went through two ideas.
The third one ends up being Usermind.
My first idea was a little too consumer, I didn't think I was the right founder.
My second idea I was too late, there were already five or six companies
So when I happened on this idea for
Usermind, I thought "I think I'm going to be early.
I think it could be big.
I think it's a category.
I actually think it'll change the world positively.
I could do it for ten years."
So that's when I said, "This is the idea.
This is the company I'm going to go do."
I've been at that for six years and we've raised $50 million dollars, and it's been incredible to see both our early customers like our alpha beta customers, and then our early paying customers.
We recently just had our first very large seven figure deal, and it's just been super cool to see this idea that began on paper in my living room become real software, delivering real value for real customers.
Grant: That's amazing. Tell me just a bit more about what is Usermind, what are you guys doing, and what's the core thesis there?
Michel: Sure. I came at it bottoms up, so we would
describe ourselves today as being all about the customer
I'll explain what I mean.
When I was talking to these operations teams--
And mind you, I
went to talk to marketing operations and sales operations and HR
operations, and what became really clear is that each of those
teams has a lot of software and that's a lot of
software that stores customer data.
None of it's really connected in real time.
What became really clear to me is that the
more that the front office is transformed, the more tech
they buy, the more channels emerge to talk to
customers, actually the more fragmented the customer
I thought, "Wow."
By the way, I probably didn't realize it at the
time, but for people who don't--
When I started to do research
and customer experience, it became pretty clear that
almost every big enterprise company--
By that I mean banks, JP Morgan Chase or retailers. Every single big enterprise company is being disrupted by startups right now, and if you look at Uber "How did Uber disrupt taxis?" Its customer experience.
We could just go on a phone and ask for a cab, I don't even need to
We know exactly how long it'll take, and the car will be clean. But it was this digital disruption.
I thought, "There is this moment in time right now where every single big company is going to have to figure out how to deliver really great customer experience, or some startup's going to kill them."
So there's urgency, there's a business problem and people need to act now and now in the next 5-10 years or there's very negative consequences.
You always need that for a software company to
emerge, so how would you do
If you're a JP Morgan Chase or you're a big bank, or you're United Health Care Group or a big customer, you can't have a luxury of releasing a mobile app.
You can't just throw away your call center.
So I thought, "Someone will invent
integration software that essentially connects those systems
together, and deliver software that monitors
these customer journeys in real time and
intervenes to optimize customer experience."
Imagine a scenario where you're
starting to open a new bank account and today the bank
doesn't even know you're doing it, and you might get these timed emails.
What if you got a bunch of behavior driven communication?
If you didn't finish KYC, what if someone called you and asked you for that data?
What if we could predict that some of your payee data was wrong and change the experience? I thought "That's the future. These big companies can't rip and replace all these channels, they can't just throw it away and build a mobile app. They're going to need this glue to monitor these customer journeys and help optimize and improve them in real time, so they deliver better experiences."
That's how we would describe what Usermind does.
The analysts, being analysts, call it terrible names.
Forrester describes what we do as "Journey orchestration," meaning we're a conductor conducting the orchestra channels to improve customer experience.
Gardiner calls what we do a "Customer engagement hub," a hub that
connects the systems together to optimize engagement and
But either way you slice it, the
concept is to move from essentially
siloed, channel-oriented interactions where you're on the website
and you call the call center and the call center person has no idea what you were
I give my example, I
bought a house in Comcast's-- I was calling them, I
wanted my subscription and I wanted HBO to watch fights
and it was super nice.
Then the guy comes to the house and sets it all up, and again he's super
nice. Then I go to watch my first fight and HBO is not set
It's a great example of where the touchpoints were great, but the journey didn't happen.
The funny thing is, nobody knew-- When I called them and said "I don't have my HBO subscription," actually no one needed to come to my house.
It was a software setting.
So, it just seems silly.
Why isn't there something like Usermind listening to each of those touchpoints and checking and saying, "On Michel's bill of lading she ordered HBO."
But checking the technician settings, HBO is not configured.
If there was something like Usermind listening in the background for
those signals and then intervening--
Potentially as simple as changing the setting or putting a pop
up to the guy at the house and saying, "Before you leave, you're missing-- You
haven't turned on HBO."
Either way, I would have had a massively better
experience than what I did.
To me, that's the future.
"How can you listen to these disconnected systems
in real time and put some intelligence around that to
intervene and improve customer outcomes?"
That's what we do. I feel like I've been very lucky to found the company in a moment where that's very strategic.
Many companies are really trying to figure out how to reinvent
themselves around customer experience.
They're trying to think about what they call "Customer journeys," and "How do you help a customer get to the outcome they want?"
I think it's -- Although I came at it bottoms up, I came at it thinking about it as a technology problem, it's turned out to probably be the biggest business problem I've found in my career with the biggest business impact.
If you can do it well, end customers are happier,
companies have more loyal customers and they make more money.
It's probably the most impactful problem I've ever worked on my career.
Grant: Who would the buyer be ? Is this a customer support--? Like, what org inside of an enterprise is buying Usermind?
Michel: I would say today that's our biggest challenge in the maturity of the market.
First of all, anytime you are running a category creation play or you're creating a brand new piece of software that somebody has to move budget to and there's a buyer for, it's harder in the beginning to get it off the ground.
What I would say is "We don't just sell to one persona.
We sell to three core personas."
In some cases, we sell to a centralized CX team, and
we call them "Empowered CX teams."
They started and they were early adopters of NPS, they probably own Medalia or Qualtrics, and they have journey maps in their thinking, and they're respected cross-functionally.
That's one of our buyers.
So, the second buyer would be in some cases
companies want to solve the problem of experience and they
create a digital team.
Our second target is a chief digital officer, where people have said "We're going to take Martek and the call center and some of these back office apps and put them all together, and empower someone to improve these experiences."
That's our second buyer.
It's not really a shared service, CX is like a shared service.
It's more like, "That buyer owns everything that's involved in delivering the experience."
Our third target is marketing. In some companies, like e-commerce companies or particularly in retail, quite often the marketing team is really the ones empowered to think about these experiences and journeys for better or for worse.
The benefit of what we're doing is it is a new idea and a new
market, and a new category.
You get to set your price point for the value you're delivering.
The challenge is that I don't think yet it's settled on what's the
best org structure to drive this, and so maybe five years
from now we'll see that like it's always CX or it's always a
But today our target can vary by vertical and by company.
Grant: That makes a lot of sense.
I want to go all the way back to the beginning of
where you started, because there's something you said that I thought was pretty
It's that you were working at a gas station and then you got a training program that basically put you on a trajectory in order to be where you are today.
I just want to talk a little bit about that, because I think that's pretty
amazing. It feels very
relevant in today's age.
Can you just provide a little detail
on how intensive was that?
Three months, six months?
It sounds like it was almost like an apprenticeship where you were then staying there for some amount of time afterwards. Is that right?
Michel: Yeah. No, that's exactly right.
It was called the "Professional development
These types of programs, I'm sure it's relevant now, they're quite often created when you have a scenario where more skills are needed in a market than exists.
I would say for developers today, a lot of these programs exist. Think about code schools or--
Grant: Lambda School.
Michel: Exactly, Lambda School. Right.
You want to find a set of skills that's highly valued and
scarce, and then quite often big companies--
Particularly it's a good entry point because a big company has the money and the ability.
A startup really probably wouldn't hire you because you have to have every hire be maximally productive.
But big companies, you don't have the luxury of time and money.
Compuware built this program, as I mentioned, called the PEP program.
They were taking everyone from fresh college graduates to people like me, putting you through a rigorous interview process, and then literally they gave you a house in Detroit, a rental car, rent paid.
You received a very minimal salary, and over the course of three months they put you through this rigorous training program.
Everything from simple coding to SQL, I think your graduating project was building applications and demoing it and making it work.
The idea was that they can produce consultants and technical salespeople who were deep enough to be effective. You weren't going to go and lead a project, you were going to go and work with more experienced people to be successful.
Then if you look at a pre-salesperson at that time, you're probably talking six figures.
The way that you had to commit to work for
them-- I think it's 18 months, but it might have been--
at a reduced salary, and that's how they recouped their
investment upfront in you.
But I look back on it, and it's completely Life-Changing.
The fact that I probably wouldn't have gotten the interview
if I didn't have college on my resume.
But it was this incredible opportunity, I certainly worked my butt off while I was there and tried to make the most of it.
But it completely set me on the path of being in tech
and frankly, in enterprise.
It changed the entire course of my life.
So, a very unusual opportunity, but I think
there's probably plenty of those today.
I think if people are looking to get into tech, it's a great way to get in.
Grant: Yeah. It's also just-- I think it's a really interesting and inspiring thing to want to be able to do someday, to be able to even maybe have that same program Usermind in five years, where you can bring up the next set of leaders to make the same journey.
That's really amazing. Cool, so then a couple of other questions about some of the things you talked about.
Grant: I'm guessing that a lot of our listeners have read Ben Horowitz's book, The Hard Things About Hard Things, and probably read his perspective on the Opsware acquisition by HP.
I'm really interested to hear your take and how
you perceived it inside of HP, and how
you built consensus around getting this acquisition to
I think that's an interesting take that we don't often hear. I 'd love your experience there.
Michel: Sure. A little bit of backdrop, one, I had completely fresh eyes.
You had a bunch of leaders coming in
from Mercury who weren't attached to previous
decisions, and sometimes for big outcomes that is
I think as a leader, ask yourself, "Where do I need fresh eyes?"
But when we were acquired, I didn't have a background in the
I came from the testing space.
I didn't know about data center automation or operations, so when I took over
the business I think I was just asking--
I was all
about first principles.
If I was describing to you the state of the business within HP, I would
say "First thing, HP had acquired three
separate companies that they put together to make this data center
For people who don't know Opsware, Opsware
was a company that built automation software for
discrete pieces of the data center.
They had a server automation product, and a network automation product, a storage automation product, and they had a run book technology that glued them all together.
The value proposition of something like Opsware was
"The infrastructure in the datacenter is exploding,
and you're virtualizing so it's only going to get infinitely more
Humans can't keep up, you need to automate patch management and deployments, and so forth."
Again, they were an idea, which is a great idea
but their timing was impeccable.
People were just-- With the rise of virtualization, it was just
impossible to hire enough admins.
That was what was driving the shift in the
market or the "Why," "Why was why was this market
That's always the first question when you're in a big company, at
least from my point of view, you have to ask yourself.
"Is this important as a market?
How big is it? How fast is it growing?"
From my point of view, it seemed that the market was now and it was going to grow extremely fast.
Without answering whether HP had to own it, it was
very clear that it was-- Or, that it could be.
Because if you're HP you can't be in a market that can only generate $50 million in revenue, you have to ask yourself "Can I get a $500 million dollar business here?"
If I can't, it's not worth doing.
Like Check, huge market and
very fast growing, being driven right now by these
technology explosions with no other alternative.
Really clear to me it was a great market to be in and you could build a big business there.
Then the second question you have to ask yourself is "How well positioned are we as
HP with our current set of offerings to win it?"
As I mentioned, the company had acquired a company called
Radia, Innovadyne and another product
called App IQ, and I'm forgetting what the third thing was
at this point.
But it was this patchwork quilt of stuff that
was put together, and it didn't really integrate, and they didn't really have
a great vision.
More damning from my point of view, I think we were sixth or seventh in market share.
When I looked at it, I said to myself "OK. 2 core things. One, where are we in the market? Is it early in the market and we can go from seventh to third? Or is it late in the market?"
At the time, it was pretty clear that the top three players were Opsware, Blade Logic and IBM.
If you just added up the market share of those three
vendors, and market share is always a
question, but it was pretty clear that they were north of 80%
of the market.
When I did that market analysis, I said
"This isn't an early market.
This is not a market where if you add everyone up, we have 20% of the
market. This is a market where the players have been
Then you have to ask yourself, or this is a thought process I walked the leadership team through, of "OK. If the top three players have basically dominant market share, which in my mind means the market position has been set, and the market is now an hypergrowth.
Ask yourself as HP, can we --? Do we have the capability to out-innovate HP or Blade Logic and Opsware, and somehow disrupt this market? Like, disrupt the position of market?"
My answer to the leadership team, what I said to them was "HP is not an innovation company, it's a channel company.
What are the odds that the market's going to have some weird
technology shift that's going to let us-- And by the way, we're going to be right about
Then we're going to do it for 18 months and we're going to be able to fundamentally change what I think is a settled market?"
It's a very low likelihood that that's going to happen, so then the question becomes "OK. If we agree that it's a very big market and you could make a ton of money in it, and leaders are already very clear and it's highly unlikely that HP is going to innovate our way through just engineering and product into winning that market," then you have to ask yourself as a company, "Do we have to win the market?"
Because if you have to win the market, then you really only have the choice of buying
the leader. Because you're not clearly going to build your
way to winning this market.
So you have to ask yourself, "Can we just walk away?"
That was what I proposed.
So option A was go buy a leader, and option B was shut that business down.
Because it doesn't make sense to put a ton of engineering on a market you can't win, and if you want to win it, you need to go buy a company.
That, when I did my first QBR, that was
how I summarized what I thought was the state of the market and the
state of the product.
I think then the question becomes, "Do
you have to be in this market?"
I would say two things about that, one, virtualization at that time is like containers today.
It is the biggest disruption to the way the data center is run in however many years. I think people felt like that was a tornado, and it was important to be there.
I think the other piece was HP had its
cash cow, by the way.
Its low runner was a product called HP OpenView.
If you don't know, HP OpenView is monitoring technology,
and it was monitoring technology for the datacenter.
I think the question I asked the leadership team was, "OK.
Let's imagine we walk away from this market.
What happens if monitoring and automation
converge? What happens if winner of
automation can then disrupt the winner of
Then the risk of not buying a
company, because you can't build your way there, the risk of not buying a company is you
might create risk in your most profitable business in the
If you really walk back through that math, it's pretty clear that if you think that it's not only a big opportunity, but if you don't act it might create risks to the most profitable business in that software company, I think it becomes very clear to the executives that they should do something.
So that was the first phase, creating this narrative around why and where the market was, etc.
Then our leaders went out to the sales guys and asked, "How often are you even asked about the software?"
And I think it was at a QBR that we talked to all the sales leaders and they were all, "We're having our ass handed to us," and "It's urgent."
We're seeing these opportunities, so I framed up this narrative and then we asked the leadership sales organization "Could you sell this? How often are you walking away from revenue?"
And the answer was, "We're a channel company and our channel is in all these deals."
Not only is there no way to
build it, but if we bought it everybody thought we could double or
triple or quadruple the business really fast.
Because our salespeople are selling to the same people who are buying this stuff, so that was the second piece of the puzzle.
Getting alignment around the fact that our channel could be
these opportunities, and if we bought the company probably there'd be a bunch of growth
that would come out of it.
That netted us out to
agreement that we wanted to go look at M&A,
and everybody's aligned on that.
Now the third thing is like, "What are you buying?
Are you buying Opsware? Are you buying Blade Logic?"
For people who don't know, Blade Logic was not a suite
It was only server automation, and it had partnerships for these other things like network automation, storage automation, and Opsware was a suite.
They had done some acquisitions and they had this broader
Unlike-- Just to net it all out, because I could talk in detail about all this.
Grant: This is incredible. This is such great detail. I love this.
Michel: I think our theory was basically that the
winner was a suite vendor, the winner was going to have the
automation capability for the whole data center, not just server.
So if you believe that that's where the market was going, and that's cool now I look back on it and there's a Gartner MQ and they had a data center automation wave, and the market was agreeing that it wasn't one product.
It was going to be a suite that would win, and if I was netting it out Blade Logic had a better product than Opsware.
Where the customer wanted to just buy one product, blade
logic was winning, and where the customer wanted to buy a suite
Opsware was winning.
We went through this whole evaluation and talked to both companies, and of course we had CorpDev So it wasn't like I was leading any of that.
I was providing supporting commentary, I was on all the calls and giving my feedback on the product.
But at the end of the day,
HP was not going to be able to do five small acquisitions and
As a company, you have to
know your own strengths and weaknesses.
If you really look at it, were we more likely to succeed with one acquisition of all the products already put together that we just put in the channel and win?
Even if some of them weren't as good as competing products, one suite or one acquisition, less integration risks, a bigger story to sell.
I think that was the right move for HP.
It's a better acquisition for HP, and so we went through this
debate. It's funny, I'll never forget it.
We were in a room all talking about
who's voting for what acquisition, and people had voted
Opsware and people had voted Blade.
I was the final vote and I voted Opsware, and that's the one we
I didn't do the negotiation, and obviously that ended up being a pretty big $1.6 billion deal.
I did sit in a sweaty deal room writing up the press releases and eating disgusting food while they were all negotiating, but it was a pretty incredible experience.
Grant: That's amazing.
And look, I think the really interesting part
of these deals in my opinion, is not the negotiation.
It's not the terms.
it's that case that you built.
perspective of "How do you build this case?"
I think you just did such a marvelous job of
walking through the logic and the different considerations.
So, thank you for that. It was really amazing.
Grant: Let's just keep going on this thread, because I think it's super interesting.
Once you acquired the company, once you acquired Opsware, integration and integrating companies you're acquiring is generally a fairly complex task.
Can you talk just a bit about your perspective on what allowed you to integrate them so well, and to really use your channel to sell through and to make that successful?
Michel: Yeah. Look, I think one, it helps that Ben came in, Ben Horowitz, and took over HP software and was in charge.
You need a lot of exec and recover to make an integration work, and we
had probably one of the best guys who could be out there doing it and
I had a great relationship with Ben from day
one, so I feel like I had all the support I needed.
Because look, big companies are irrational.
That's one of the central issues of integration, is that the standalone company is optimized to do what it does and everything about it is optimized to do what it does.
Then you try to integrate into the borg, and the
borg has a lot of irrational things that make sense for a big
company, but for example, don't make sense for little
I'll give the examples. When we were doing integration and doing the integration of Opsware, I don't even remember what source control system HP had.
But HP had this mandate that all development in HP has to
be in source control system "Blah."
Of course, Opsware didn't use that source control system, and frankly actually the code base Opsware was written on wasn't supported in this configuration management system that HP has chosen.
It sounds silly, but the number of calls I was on
to actually get a waiver--
Because look, if you're HP there's huge benefits at scale. Remember, this is when Hurd was running HP.
The benefit of simplification is actually quite
gigantic at the scale of a company like HP, but it's
irrational to spend $1.6
billion dollars and waste even a moment of your engineering time
porting your source control and rewriting the software into a
different code base.
So one challenge of integrations at big companies is these weird, irrational things that just don't make any sense, and so you have to just wade through those and you need enough organizational capital and support to deal with that stuff.
If you don't do it well, it can really derail things you're doing.
I would say there was a lot of those, although I would say as the person doing the work it was actually quite eye-opening.
I learned a ton. Having been in pre-sales on product, it was my first exposure to every single piece of business because you have to think about source control, and think about license management, and redo all the reps comp plans and partners.
Honestly, it was an incredible learning opportunity for
me on all the pieces of a company and how to think about
getting them to mesh.
There was a bunch of weird irrationalities, but I would say the flip side, something we did well and we didn't do it perfectly, is you have two phases.
You want to put that product in the channel as fast as you can and train all those reps, and of course HP had the benefit that we were selling to the data center team already.
It wasn't like the rep had to look learn a new person to go
talk to, but we turned the Opsware sales team into an
overlay and I think we did a really good job of
We did a ton of training as fast as
we could, and of course in that first year BMC bought
It created a lot of competitiveness and organizational
alignment around that, and we wanted to seize the momentum and
really sell as much of it as we could to
seize market share.
I think those are things we did well, and maybe we did them-- I think partially we did them well because I ended up taking over the Opsware product team and they were really good at it.
Opsware had a great sales management team, but I think my background as a hybrid product manager/product marketer, I understood how critical sales enablement was to actually maximize return on the deal.
I spent that first year literally just doing training and flying around the deals, and I think we grabbed a ton of market share from Blade Logic and from BMC in the process.
I think that was super well done.
I would say another thing is in order to make an acquisition work well, actually retention is one of the most important things. If you have a ton of attrition, it's actually quite challenging because in enterprise software the IP is people. How you sell is IP, and how you build your software and what it does is IP.
More than in any other business, people are very much the IP of M&A.
I was very proud that part of it was that Ben took over the team,
part of that was that we did a lot of work in the Mercury culture
which meshed with Opsware pretty well.
I think we've retained a ton of people for the full two years, and that was probably a factor in success.
Then you just make a bunch of mistakes.
I point back to, there was one decision I wish we had done better when we were making all the Opsware guys overlays there was a question about compensation.
We gave the Opsware guys pretty generous
comp plans to just go try to sell Opsware into
There was an internal conversation about "Should we pay the HP reps extra?" Because as you know, if you're in a bigger company the comp plan dictates your focus.
Reps are very good at gaming the comp plan, and if you change the comp plan you can change rep behavior in a pretty powerful way.
Probably my one regret is that we were arguing for the HP broader sales organization to have very specific steps in that first year to maximize their attention and really grab market share.
Do a land grab, and think about the size of HP's channel.
It was shot down because it's a pretty big
cost investment, and the company was such a bizarre
This was one of the weird things about being in HP when Hurd was there is that the company was super focused on costs.
I think in software companies, all profit goes to the victor.
You don't optimize for costs, you actually try to lose money to gain market share.
Because once you've won the market it's super profitable for a long
period of time.
I think that was an area
where our software leadership understood that, but the broader
leadership of HP didn't buy into it.
So that's just a great example of where we didn't win a battle, and I think we probably could have even stolen more market share and been more successful if we'd been able to do that.
Grant: It's really interesting.
I think when you talk about the
incentives really driving bigger companies, it's so true.
I think there's a special skill of designing those comp plans and understanding what you're really trying to strategically achieve through them as well, so it's a very important area.
One thing that you said too, was
you mentioned in passing that you were
doing product, but there wasn't really a division between product
and product marketing.
I think when I look back, and I talk about this sometimes, which is you read Ben Horowitz's piece about good product manager and bad product manager, it feels like historically product was actually a subset of marketing.
Then eventually, I think now we're in a
world where there's a pretty clear line between
product and marketing, and product
marketing ends up being this org that sits between the two.
But I love your perspective on how that role has
evolved, how you think it should be structured and what you
think makes it or defines really
good product, versus really good product marketing, versus
really good marketing?
How do you bucket those?
Michel: I think I would start by saying part of the reason people are so confused about this, in my mind, is that the consumer and enterprise worlds are really different.
In the world of a consumer product, whether that product is Uber or whether it's Nike shoes, you're just selling me a thing.
Of course, I'm not going to minimize the role of marketing, but
the number of touches and the amount of
content to do a deal is massively
I go to the Nike Web site and I search about the shoes and I spin the shoe around and I change some colors, I'm not minimizing how much influencers are involved and how key the community and brand are, but you're really selling just to me.
There's not a lot of other people involved in that decision.
Nike is selling shoes to Michel, or
maybe Michel's mom.
The number of vectors, or things intellectually that have to be evaluated to make that decision, are smaller.
The price point, the investment, therefore the risk of getting the decision wrong is much lower.
Even though it feels painful, it's a $100 pair of shoes.
And by the way, you can return them.
So now let's talk about enterprise, and I'll explain why I think product marketing is so
Enterprise is totally different.
By the way, in that world you have product people and your marketers.
In enterprise, it's the opposite. You're selling me a very high value item, or even something like Box is thousands of dollars a month.
Which none of us would just choose to go do that.
Like, who would just sign up--?
Even a car payment, it's a pretty serious decision for most
So you're talking about the investment, which is much higher, and an enterprise software that could be millions dollars a year.
That's thing one, so you better get the decision right. Thin two, quite often lots of people are involved.
By the way, "Lots" could be one to two people or it could be 10 or 20 depending on the complexity of your enterprise software.
Then the argument to be made, of why to choose option A versus option B is actually often quite nuanced.
Meaning, why do you choose for example,
Salesforce Marketing Cloud vs.
I bet if you went to the website you probably couldn't figure it
You have to look at it and get
details, and there's just a lot of nuances that might affect your
company and help you make your decision.
If that's true, if that's your sales motion and that's the kind of
thing you're selling, the product marketing is an essential
It is of equal importance
to marketing, and frankly I would say my second belief,
which I think is in the minority of the world right now, I think the
companies that win in enterprise are the ones where there's
very little daylight between product management and product
Because at the end of the day, think about when you when you sell. If my job as a product marketer is to explain "Why Usermind?"
And deliver all the collateral to explain that and differentiate more competition, how do I those product strategy if I don't understand the market landscape?
The best product people understand the competition, understand the landscape, understand what parts of pitch work.
Then potentially, by the way, build features just for the
I think sometimes we say "A product manager's job is to go talk to customers."
And, yes. Certainly your job is to make customers happy, but your
North Star job of product management and product marketing is to win the
Win the market, that's your job, to build a profitable business for the company.
So I have a pretty strong point of view that those two things should be the same function, they should have a single head.
That is controversial, most marketers want to own product marketing because they think it's more important that the narrative be aligned to the marketing motion. Whereas to me, if you really want to dominate a market you need to make sure that the product marketing is intimately feeding into the product management function and vice versa. On the companies that do it well, they create markets and dominate them.
When I was at Apptio, I ran both product management and product marketing
and it was extremely effective.
But I think the work structure is, I think first of all most people don't understand these roles.
They don't understand why they exist, and as a result you end up with weird work structures.
Product management reporting into engineering, or product marketing reporting into
I think it's because these are roles that didn't
exist 30 years ago.
I think many executives don't fully understand what these functions do, and I think it's actually a third thing.
Which is every other function in the company lends itself
to measurement very easily.
If you think about it, marketing can be measured.
Leads and events and pipeline generation, and engineering can be measured by feature velocity, and services can be measured by time to value.
How do you know if your product managers or product marketers are good?
You actually can't measure that quarter of a quarter, I don't think.
I think this is the other challenge with creating an independent
function, is it just more comfortable to just measure
velocity or just measure leads?
Put product management, engineering and product marketing, because it's really tough to know.
By the way, the people who are really good, it's 10x better than the
people who aren't good.
But either way, you can't look at a number
and measure the product market or measure the product management.
What are you going to measure? How many decks they produced, or--?
"They built 10 Jira stories, but are they the right 10 Jira
Because there's a big difference between the right 10 features and wrong 10
I think this is part of why there's so much
angst around "What is product?
Where should it report?
What's the saying?"
Because if you really make it strategic and build it as its own function, it's actually really hard for the CEO to know if it's good unless they've done it.
Grant: That's amazing.
In your perspective, you would have a chief product officer, a CTO, and a CMO all be three peers running their perspective organizations, and you'd have product marketing fall under the chief product officer?
Michel: I think in an ideal world.
By the way, I'll tell you at my company it doesn't look like that.
My head of product, my product marketing right now is in
marketing, but I also run the company.
I think some of it is the skill set you
I think my head of product is amazing and
he's better at inbound, so some of it is also the people that you
But look, I've been doing this for 20 years.
I don't have a hard time telling whether my product or my product marketing
team is good.
But I also, if you had a sales-oriented CEO or an engineering-oriented CEO, I don't know how I would train them to know.
Grant: Yeah. It's interesting, so your perspective is because you are a product and product marketing leader, then that becomes a central skillset of the companies almost because of that.
To have another person who's doing that at the same level that you are, it just becomes a little bit like you need more assistance on sales.
Because you're not a sales-focused CEO, you're a product marketing
So you're saying almost like it's based on the skill set of the top leadership, and then supplementing their base from there?
Michel: Yeah, I think definitely the kind of founder you are changes that dynamic in a leadership team.
They say the hardest function to have is always the one the CEO used to do.
I think in our company, it's the opposite. I am actually, even though I'm involved in product decisions, I don't spend a lot of time on product.
Because I can spend a very little time with the room
and assess his plan and give him feedback, and do that
Whereas before working in the field, I don't have 20 years of sales experience, so it's going to take a lot more time for me to probably achieve the same outcome because that's not my superpower.
Grant: Got it. You think the challenge there is giving you the context that you need in order to help make effective decisions in some of these other orgs?
Michel: Yeah. Look, my head of sales has been doing selling in the enterprise for as long as I was doing product, and they can tell whether a deal is real really fast.
He's got a spidey sense and he's developed this however we want to describe it, it's expertise that is somewhat tangible and somewhat intuitional, and I don't have all that.
Grant: And that's Derek, right? Who I worked with for a few years when we were both at LivePerson.
Michel: Yeah. He's incredible.
Grant: You have a great sales leader there, that's for sure.
He always-- Interestingly, my perspective and part of the reason I
always liked him is because he understood the value
of product and product leaders to come in and spend
time with customers and explain things
in a deeper way than just handwaving
and trying to push the deal through on a relationship basis.
He really liked that deep value explanation.
Michel: He's a great guy. He's an enterprise guy, and enterprise guys understand that they need product and they need product marketing.
If they do their job, they can sell pretty high price point deals.
But ultimately the product better work and we deliver on our promise.
He's an incredible head of sales. I feel very lucky.
Grant: OK, I want to touch back on a couple of the things you said.
One thing that I think is really interesting is the ideation phase.
A lot of times, to your point, the way that I think most people would talk about founding a company is they would say "If you just want to start a company to start a company, you're doing it wrong. You need to have some crazy idea that you've been obsessing about for the last five years, and it's the only thing you could think about, and then you go do it."
And that's actually not the experience that I have either.
I wanted to start another company, I started one previously and then I
was ready to do the next.
My same co-founder and I actually spent a couple of months just writing down one pagers on different ideas and prototyping different things and investigating different markets in areas that we wanted to build for.
So, I'd love to hear a little more
detail about how you trace down those different
It sounds like you had three that you really liked, and what made you exclude the other ones other than just being too consumer too late?
What are the details in that?
Michel: If you remember back to I
said, "My high level hypotheses," because I
always think maybe first thing I would say is I
believe that the thing that causes software companies to
get created is change, or as we like to
call it, technology disruption.
Only when the change is big enough from the status quo to the new reality, is there a reason for somebody to buy something new or do something different.
In some cases, that that's just a revolutionary new technology. Containers are so much better than VMs that we all have to do it.
Or, "VMs are so much better than hardware that we all have to
But actually that
discontinuous innovation where the new thing is just a new
paradigm is pretty rare in software.
Most software companies-- Take Opsware, the VM made it impossible for humans to keep up with the amount of change in the datacenter.
So in order to deal with that reality-- And by the way, there weren't enough administrators.
Even if you could hire enough people, there weren't enough of them.
Inevitably there was enough pain that people were going to buy something to change the status quo, and so the central premise always to me of building a company is like, "What is the changes that are going to create an opportunity for a new thing, a new way of working to be inevitable?"
That's my first statement.
My second statement is I have a secondary belief, which may or may not be true,
but I believe it.
It's that a $100 million dollar
company, and I mean exit like you got a couple million in
revenue and somebody buys you.
Maybe you had one change. But if you really want to build a public company or a real material software company, almost always there's multiple disruptions in that or to that persona or that market at the same time.
The more things changing on that person, what that means is the more pressure there is to act now and the more and the bigger the change that has to happen, so the value of your software is greater.
I was looking, that was what I started with, is "How can I find a person or a set of personas who are the subject of multiple changes?"
I was looking to try to find a category, so that's why I wrote down that list of changes and decided I was going to focus on these enterprise personas.
So actually, I didn't quite start with operations. I started with just end users.
I'm like, "What's the impact on all of us that there's all these different systems that are not integrated?"
My technique was really simple.
I sat down, and I like Day of the Life interviews.
I think, by the way, the worst product people ask their
customers "What feature they want," and build a feature the
I think that's the opposite of good
Good product management is interviewing the customer and they tell you they want something, and finding out why they want it, and then collecting all these problems and then going back to your engineering team and thinking of a better solution to the set of problems that you found than what the customer imagined.
Because the customer's feature is an imagined solution to a problem they have, so first you have to figure out "Is this a ubiquitous problem? Should I spend any time solving it?"
And then second, "What's the best solution?"
Really, the answer to neither question is the customer the best
one to answer.
So when I say "That's what product managers and marketers should be thinking about and doing," and the way I like to do that is ask about a day in the life.
Ask who you report to, ask what you do, ask what you like and what you don't like.
I like to ask a lot of very fact-based but not super leading questions and see the patterns in a bunch of interviews, so I literally just wrote down a bunch of questionnaires.
They were things like, I don't know, "When do you open your laptop
first in the morning? How many software applications do you use?
Tell me about what words you use then, and which ones do you
love and what ones do you hate?
What do you wish you could do?
If there was some magical thing somebody could build for you, what would that be?"
Originally my hypothesis was
like, "Salesforce sucks.
Somebody should reinvent the desktop, and I'm going to go do
My first round of interviews were probably like 40 salespeople, and I've been in enterprise so long that it was pretty easy.
I just pinged people I work with and said "Can you give me five minutes, or half an hour?"
And people did. So my first idea was "Somebody
should invent a new UI for the CRM
that basically connects not just to Salesforce, but email
My conclusion was reps do two
things, account research and deal
There's not one system that's good for either
I could build an app that basically for research is connected to Yahoo! and Discover.org and LinkedIn, and all this stuff.
Then for deal management, it's like Salesforce and email and
Probably you could build a really killer
experience that transformed the way of working,
but no customer asked me for that.
That was me listening to their description of what they did and saying, "That's stupid that you have all these different systems, you don't like them, and you have to context switch."
The end of my interviews, that was my synthesis. I thought, "Awesome."
It's right, by the way.
Somebody should do this, and people are, and they should just--
My theory was you could build a new UI on Salesforce and give it away freemium, and eventually disrupt the entire CRM market and build a new back end.
Because actually Salesforce doesn't really do what reps need to do, it's a forecasting tool.
So, that's ideal one.
Then I thought to myself, I am not a consumer person.
Number one, I don't know much about how to build a beautiful, sexy
My real reason for not doing it is salespeople are laggers,
not really early adopters of technology.
So it's a terrible persona. The idea is right, the data is right, probably there is a whole market around these role-based desktops.
Somebody should do it for every industry, but not really-- I'm not the right founder for that company.
I went back to my interviews and there was the next thing, which was there was a thread.
A lot of them worked at SaaS companies and so they had a lot of usage data,
and they were constantly being surprised at software that they'd sold wasn't being
So I said, "Who would give you that data?"
And they said, "The operations team."
So I went and interviewed 40 operations people and it became really clear to me that customer data was going to be incredibly strategic for our next 20 years.
I thought about Mixpanel, KissMetrics, Segment, somebody should really build a collection set of capabilities that allow people to collect and optimize customer data.
Then I looked at the market and I thought, there's already all these companies.
Mixpanel, Hypergrowth, and KissMetrics had been around, and Segment and Google.
I thought, "I'm not really interested in being the tenth vendor in the space.
I'm right, great idea, somebody-- People are going to make a lot of money doing this,
but it's not what I want to do."
So then I went back to the operations team and I started to talk to them about, "OK. What if you had all that data at a single place or could connect it together, what would you want to do?"
And the answer was that what people really wanted to do was not--
every day they'd come in and download all this data from these different
systems, and expunge into a big Excel spreadsheet in order
to generate their, as an example, customer onboarding
Where are my customers, where are they dropping out, where are the
Then they would have a hypothesis
around how to improve that, and they'd have to go back to each individual tool
separately and make some changes.
Every day they were doing this.
I'm like, "This is really stupid.
Someone should build a technology that connects some of those
systems and optimize these critical--"
And I didn't think
about them as customer journeys, I was thinking about those workflows.
Like, lead to customers workflow, and customer to renewals
I thought "This seems really dumb that there's
no overwatch system, there's no monitoring system, moderations
customers looking at their context across all these
Then, the future is AI. Why do we need human beings to intervene?
We should be able to build rules and models that
basically auto intervene and optimize the
So, that was my third
idea. I thought, "I don't think anyone else is really doing it
and I think it could be big.
I think it's enterprise, so I'm a good founder to do
That's how I netted out on Usermind.
Grant: I love that. Michel, I know you have to go very soon, but I want to give you one last chance here.
Because I know Tableau is one of your customers and they are a
software SaaS vendor.
I'd love to hear your pitch on if there's a company out there that's a SaaS or software company, that's either an application vendor, how should they think about using Usermind?
Just give a pitch for however long makes the most sense for you.
Minute, two minutes, whatever. Whatever you can do.
Michel: Yeah. Look, if you're in a SaaS company your business is a subscription business.
Your KAC is probably high and you make all your money on renewals and lifetime value.
So here's the deal, if you're thinking about how to get customers renewed your customer onboarding journey is super strategic.
Probably you don't really have any visibility today
to all the steps across all the teams and systems and in
real time know where that customer is.
Just first of all, how much better would your company be if you had that
Then secondly, how much more lifetime value and renewals and cross-sell /upsell can you generate if you had that visibility and could experiment with different interventions to improve the customer's experience?
From my point of view, you're either a SaaS company and you have a
churn problem, you should be talking to us.
Or you have a renewal problem or you have a loyalty project, you should be talking to us.
Or if you've gone public and you want to lower the cost of
supporting these customers, orchestration and automation is a huge
value prop for that.
You've got a recipe, but you don't want to hire a thousand more people. You don't need to, robots can do that work.
You've either got a challenge, a churn issue or some renewal problem, and you need transparency and you want to try to get your way out of that hole, we can help.
Or Tableau being the case where they'd gotten so big that they just keep adding people to optimize these experiences, so they wanted to think about orchestration automation.
If you are in either of those two buckets, you should be thinking about orchestration technologies, and hopefully Usermind's.