The 99th episode of Datacast is my conversation with Gary Hagmueller—the CEO of Arcion Labs.
Our wide-ranging conversation touches on his early career leading business development and bringing a company through an IPO; his responsibilities in C-suite roles at various startups; advice on fundraising, sales operations, and go-to-market; his current journey with Arcion Labs; lessons learned as an appointed CEO building a high-performance team, finding customers, usage-based pricing; and much more.
Please enjoy my conversation with Gary!
Listen to the show on (1) Spotify, (2) Apple Podcasts, (3) Google Podcasts, (4) TuneIn, (5) Stitcher, and (6) iHeartRadio
Key Takeaways
Here are the highlights from my conversation with Gary:
On His Education
My academic experience is a formative piece of what propelled me forward in my career. Arizona State was great for a combination of things. Number one: If you apply yourself and work hard at it, you have a lot of academic depth. Number two: Its social aspect was a hallmark of just anybody moving into a leadership role and being successful.
I then spent a few years between undergraduate and graduate school at USC — working in the consumer products industry. I was doing a combination of sales and marketing, which was not interesting enough to get me cognitively stimulated. I found out that I was always gravitating towards technical innovation instead. So when applying to graduate school, I only looked at programs that could help me move forward in that specific direction — which was telecom at the time. I managed to land a fellowship in USC’s MBA program.
My counsel for anyone approaching that decision is to do it in a focused way. Those at USC who were very mercenary and focused on a specific outcome did some amazing things. I married my classes, internships, and networking opportunities with my goal. It became super easy to jump into one of those networking conversations, landing a job at Verizon out of the MBA program.
On Leading Business Development at Verizon
I came out of graduate school with a curious combination of a guy who could sell things, was deeply analytical, and invested a lot of time in being up-to-speed on technology. I had this 3-legged stool, if you want to call that. That allowed me to talk to our partner organizations, generally large companies that were buying from Verizon on the enterprise side, and build business plans with which we could jointly go to market. I would have to go to my company leadership at Verizon’s regional offices and get them to fund the infrastructure of building these things.
The most valuable thing that came out of that experience was honing the ability to marry the needs of my customers and my partners. They often had very different needs and expectations than the executives who controlled the company’s purse strings. I learned the attenuation process of figuring out what these guys want and how to make a case that appeals to everybody in the timelines they need.
Another way to pass that was the ability to understand requirements and work through the influencing process. Remember that before 1996, the entire telecom industry was this regulated beast that nobody did anything outside the box. Here I came along with a group of people doing wild, crazy things that all these executives (who were close to retirement) had never seen before. A big chunk of what I did was talking to people who did not understand what I was asking them to do and getting them excited about that. Those influencing skills were instrumental in what came later in my career.
On Taking NorthPoint Communications Public
The experience at NorthPoint was off the charts and amazing on many different levels. We had grown enormously throughout my time there. When I joined, I was like employee number 200. When we went through the IPO a year later, we were 1,800. Our revenues have gone up 100x during that period. When you wind up going public, you must present yourself as being a bit more buttoned-up than a couple of guys in the garage trying to build some stuff. It was incredibly difficult to have the processes and capabilities to make you look like an adult firm.
The challenge we had was basically putting together programs that would give us the ability to project our metrics. Because our revenue was still growing so much thanks to expanding customer use cases, those metrics were still being formulated concurrently. It became difficult to give projections on what we would do and how to evaluate ourselves. We had deployed a very large amount of capital to build this network out, but the amount of usage for that was not quite substantial yet. So we had to help the Wall Street analysts devise a way to value us and predict what we would be in the future.
That became one of our biggest challenges: talking about our valuation metric. We dawned upon a data population past — how many people in a population could we serve with the infrastructure and capital that we had already deployed? That worked well since it wound up being a $6 billion IPO.
On Being a CFO
The CFO role is a steward of economics. As a highly analytical person, my corporate development experience at NorthPoint gave me a crisp understanding of corporate economics. That is how I wound up in CFO roles, to begin with.
Being a CFO in a startup is different from being a CFO in a public company. A public company CFO is pretty narrow in scope. In the CFO startup world, you get to own everything other than engineering — sales, marketing, customer success, etc. Job responsibilities in that role include accounting, finance, and some HR/facility work. During the Zuora days, there was a period where almost every function in the company reported to me as engineering leaders left or sales leaders transitioned.
I was also heavily involved in the outbound part of the fundraising process. In startups, it is often the CEO or another technical founder who does that. But I have the ability to paint the story and help people understand why they would want to put money into our startup. I ran around and raised a bunch of different rounds for both KnowNow and Zuora.
During the Zuora days, since the product was sold to the CFO, they often put me on the sales call. That turned into a pretty effective sales tactic, and I certainly enjoyed the heck out of it.
On Fundraising
Early on, when your startup is pre-Series B, you want an investor who knows the technology well. You do not want somebody who is peripheral to it. There is a lot of good, smart money out there, but you want people who understand the technology because they can be helpful in getting you things you want to do. If you get somebody more of a generalist, you wind up in a world where the advice they give you does not really fit the problem because they do not understand the problem deeply enough.
If you are in a leadership role of your startup, navigating the growth journey can be very confusing, especially if you are doing it for the first time. You will look around people on your board to figure out the direction. People who are generalists may pattern-match on things they have seen in the past. People with specific knowledge of how the technology works and perhaps how the market views the technology may offer a different path. One path might lead to a cliff, while the other might provide better input on your decisions.
This flips when you get to the later stage. Now you actually want people with a good enough understanding of the technology who can bridge you to the future stages. Future-stage investors are even less technically competent about what you truly do. They fit your technical area into their investment theses. It is an interesting dynamic that usually changes around the series C stage, where you want to get investors who can take you to the next level.
On Being in the C-Suite at Ayasdi
Ayasdi was a huge learning experience for me on many different levels. It was an ML/AI platform that combines unsupervised and supervised ML to let people build amazing applications.
I had no real background in ML, so I had to figure out ML from scratch. Over time, I began to realize that ML was slightly more advanced than things I had already done.
I also had to wrap my head around different ML techniques, which have different efficacy. I had to figure out what circumstance a technique would be useful.
I had to execute the go-to-market model. Back in 2012, nobody used the word AI. It was all Big Data Analytics. The big problem at the time was accessing data, so understanding where our customers’ source data lives was crucial.
We had a generalist model in the early days and tried to sell everything to everybody. I do not recommend that to anybody because it is super hard to build a repeatable sales model when everything is from scratch. Over time, we realized that certain people were willing to pay more than others who were not, so let us focus on the first group. We built a strategic matrix to identify the solutions and markets we should be going after. That made things a lot easier to figure out a repeatable sales model.
On a personal level, I was appointed by Ayasdi’s board to run the sales operation for a while. The interesting epiphany was that transitioning from the CFO role to the COO/CRO role led to a lot of new ground. Even though I had already been a strong partner with the CEO, I realized the changing motive and drive when I truly put on the hat. A CFO’s job is to ensure things do not spin out of control. A CRO’s job is to maximize the number of sales opportunities to drive revenue. Even though I have been a salesperson in the past, that was still a big learning curve.
On Enabling Sales Efficiency
Source: https://finovate.com/ayasdi-raises-55-million-in-series-c-funding/
The high level of thinking about this is figuring out your valuation drivers — recurring revenue, total revenue, usage-based pricing, etc. Go figure out what matters to your investors today and in the future. That becomes the central unifying effect that should percolate into the rest of the organization. When you have that, you can think about ways to incentivize your sales reps.
If you have a big land-and-expand operation, you will put a lot of money upfront to get new logos because you are confident that your reps can sell them new stuff over time. This works well when your sales and customer success departments are distinct.
If you have a unified sales and customer success structure, you need to figure out how to succeed over time. You do not want to put a large incentive upfront in these instances. You put incentives for building a strong relationship that grows over time. This practice is common in many cloud providers — they do not pay their sales reps necessarily on net new deals but instead on the amount of compute or widget.
Enabling sales efficiency is a study in organizational behavior — how to align the incentives that your sales and customer success have to make you successful as a company in the eyes of the investors and other people who care.
On Being A First-Time CEO
CEO is probably the loneliest job out there. You do not realize that until you actually are. You have to make decisions sometimes with a limited amount of information. The way I navigated that in the early days is similar to how I have tried to navigate my entire career: finding awesome people and surrounding myself with those who know what they are doing. I want people around me to have opinions and be resolute about pushing their opinions. Good things will end up happening.
The other thing is that your instincts really matter. If you have gotten yourself to a leadership position, you probably got there for a reason. When your gut tells you something, you probably need to look hard at it and move on that. Get as much research or data to be comfortable with what you need to do, but your guts are worth something.
On Driving Go-To-Market for ML/AI
From 2012 to 2015, the industry evolved from big data analytics to ML/AI. At CLARA, we just tried to figure out what would work. We had successes, but it was not the easiest thing in the world. We had to really know the use cases and get people to understand what we did. Then, we needed a big political sale to get everybody lined up and collect their data.
From 2015 to 2018, ML/AI became a defining category. Customer demand went up. People read about AI in some magazine articles or blog posts, but they were clueless about how it works. It was a tire-kicking demand to set up the 3rd phase, which was the vertically applied-selling model. Now customers understood the exact benefits of what they spent their money on.
After we sold Ayasdi, I specifically sought out vertically-applied AI companies as the next step, which led me to CLARA Analytics. It is hard to sell a generalist model, so focusing on a particular efficacy is better.
On Being an Entrepreneur-In-Residence at Redpoint Ventures
The experience at Redpoint was over the top. I have spent a lot of time pitching to VC firms. This was a super exciting opportunity for me to work with a partner over there whom I respect and look at enterprise infrastructure startups.
The genesis of the idea was basically going out and looking at ways to solve the problems that a lot of customers I had and the companies I worked in. I got to spend time working with brilliant people and thinking about what is missing from the world and how to position something to solve these problems. I also participated in many interesting opportunities with entrepreneurs whom Redpoint had funded and struck great relationships with. I was blessed to be in an advisory role with some of them.
At the end of the day, the real gist was putting together two parts of the tech industry that I have always been associated with: the operating side and the financing side. It was great to see how those things come together. Candidly, that experience gave me a purse as I think about growing future companies and fundraising down the road.
On Becoming the CEO of Arcion Labs
For the entirety of 2021, I spent a lot of time looking at various business models that solve problems in AI and BI. I have been working with Arcion (formerly Blitz) to determine the market opportunity. The interesting epiphany that happened here was:
The product works. As I had previously worked with tech companies, sometimes the product did not work that well or do what it was supposed to do. Arcion’s product worked well with giant companies.
There was no salesperson. They managed to sell multi-million dollar contracts to big companies with a couple of engineers.
Doing more due diligence, talking to customers/investors, and examining the marketplace helped me build confidence that: not only does Arcion has an interesting technology, but its market timing was also perfect.
Throughout my career, for the most part, I have been super lucky in that I have been able to fit from a culture and personality perspective with the founders of the businesses I would be associated with. Arcion’s founder, Rajkumar Sen, is an amazing technologist who understands the market explicitly. We figured out a division of labor between us to establish our strengths and address our overlaps. We had done an excellent job of fleshing out those details. Raj is also an amazing networker with a deep network in the industry. I have a good network too, but it is in a different area. Bringing those two together makes for an awesome combination. It has been a great partnership thus far, and I am looking forward to seeing how that continues to evolve.
On Data Mobility
It was difficult to build ML applications because most of the data you need live in enterprise data stores (Oracle, SAP, MySQL, etc.). Getting information out of those systems is super difficult for a few reasons:
If you buy from them, they are not incentivized to give you information since they have lock-in benefits.
Even if it is super easy to get those things, IT people do not want you in their production systems. The more you hammer these systems, the slower they get.
Even if you tolerate having performance issues on these systems, the IT people do not want you messing with the security profile of those things. That is how hacks happen.
Arcion technology provides an alternative way of getting at transactions in real time. It does so by tapping into the production system logs. All the inputs that go into a database also end up showing in the log at some point. But the big challenge is that the log data is unstructured, so you must figure out how to rebuild the database from the logs. Most people do not do this on their own since it is quite hard.
Arcion’s architecture is straightforward:
There is an extractor that understands the schema of the database attached to it. This extractor listens for when data changes as soon as it sees an entry posted in the log, depending on how you want to configure it.
It then takes the schema and normalizes it.
There is a centerpiece that deals with all the metadata components.
Finally, there is an applier module that writes the data into the target data system.
Arcion has a cloud-native architecture:
It is highly distributed and can deal with a high burst of workload.
It performs various types of normalization.
It is primarily designed to take things from legacy data stores and move them into the cloud.
On Finding The Technology Partners
Our partners are generally target systems where data goes into. They essentially wind up making more revenue and deeper customer relationships by having more data come into them. Every time you unlock the ability for somebody to get into a database, replicate the data, and move that data into a cloud system, the cloud utilization increases. At the same point, the customer experience for their customers also goes up.
For Arcion, partnerships entail approaching the cloud systems we can transfer into. Three factors go into the decision:
Business model alignment: They should have an active book of business.
Technical capabilities: They should focus on real-time databases and real-time applications.
Strategic roadmap: Maybe they do not have a large customer base at the moment, but they are building a seminal technology that drives a future thing.
On Hiring
Hire the best possible people you can find by figuring out how to tell the story of what the company is doing (so that it makes sense to them). You can mix and match depending on the needs for the actual roles, but the core story goes back to: What is the vision you are driving at? Why does that vision matter?
Another critical thing is ensuring that they are culturally aligned with the company. Every time I talk to somebody, I do 30–60–90 days of planning to get people on board. The more you help somebody get on-boarded in the first few weeks, the more successful they will be. The longer you wait to get them on-boarded, the harder you will make it. The more distracted you are as they come up to speed, the harder it is for them to be effective and drive value for the company. I genuinely think that winds up being a big decisive factor.
On Building A High-Performance Team
There must be a real bond and good cultural alignment between leaders of different departments (sales, marketing, engineering, etc.) because they will do the same for their teams. Before you know it, you will have an organization that embodies the culture of things.
Here is the philosophy that I have developed over the number of years:
I try to ensure I am not an asshole, and other people I bring on board are not either. You can pick it up on references.
I encourage everybody not to pick the first person who comes along, even if you are of an immediate need to hire someone. Pick the best possible person even if it takes a little longer because it will be a better outcome for everybody involved.
I look for people with an opinion. I try to ensure they do not have massive egos, but I want to bounce ideas off and get honest answers from them.
I also find in interviews that clarity of thinking (the ability to say what you mean or ask in a way that I understand) is crucial. A hard thing in startups is getting down to the most important one or two action items. The clearer you can make that ask, the less ambiguous and more effective everyone will be.
I am the sort of guy who likes to get into the weeds as needed. I do not view myself as a micro-manager, but I will volunteer if you need me to jump in the trenches and dig things out. This shows you are part of the crew and not aloof. It is also a solid bonding opportunity.
The final thing is being transparent — what you see is what you get. If you do that, sometimes, it can be painful. There will be instances where people ask hard questions, and you do not want to answer that right now. But if you provide the context around your thinking, people at least understand what thought went into it.
On Usage-Based Pricing
In the early 2000s, a lot of technology adopts the “fake it ’til you make it” model. Now, we are moving into a world where it is far easier to understand whether or not something does things that they need to do. We can separate marketing hype and product capability.
The concept of usage-based pricing is very powerful. If it is not useful, you will not use it. People used to say this about subscription services but frequently have to commit for a year or two. Usage-based service is much more crucible since you have to provide value immediately. That forces a big focus on the utility and customer experience you provide.
Show Notes
(01:45) Gary walked through his academic experience getting a Bachelor’s degree in Business Administration at Arizona State University and an MBA in Finance at USC — Marshall School of Business.
(04:52) Gary recalled the most valuable lesson from leading a business development team in the enterprise offerings group at Verizon.
(07:45) Gary recalled the challenges of bringing a company public during his time as the Director of Corporate Development at NorthPoint Communications.
(12:18) Gary shared his learnings while holding a COO role at Vinfolio — an innovator in the wine Industry.
(15:19) Gary talked about his responsibilities in the Chief Financial Officer roles at KnowNow and Zuora.
(19:06) Gary gave advice to founders seeking the right investors for their startups.
(23:51) Gary walked through the learning curves while serving as the CFO, CRO, and COO of enterprise AI pioneer Ayasdi.
(31:06) Gary shared his playbook on building a well-oiled sales operations machine.
(33:46) Gary shared his journey as a first-time CEO at CLARA Analytics.
(36:37) Gary talked about his proudest accomplishments while driving significant growth for CLARA.
(37:52) Gary discussed the go-to-market motions implemented at CLARA.
(41:07) Gary walked through his brief stint as an Entrepreneur-In-Residence at Redpoint Ventures, a top-tier VC firm focused on early-stage investing.
(44:14) Gary rationalized his decision to become the CEO of Arcion Labs in December 2021.
(49:39) Gary explained the high-level architectural design of Arcion’s data mobility platform.
(54:19) Gary discussed strategies for finding the right technology partners to collaborate with.
(57:42) Gary highlighted a few customer use cases of Arcion.
(01:01:48) Gary shared valuable hiring lessons to attract the right people who are excited about Arcion’s mission.
(01:04:28) Gary distilled lessons learned while building a high-performance team at Arcion.
(01:09:14) Gary described the benefits of adopting usage-based pricing in enterprise technology.
(01:11:41) Closing segment.
Gary’s Contact Info
Arcion’s Resources
“Dawn of the Data Mobility Era” (Feb 2022)
“Arcion lands $13M to help companies replicate data across platforms” (Venture Beat, Feb 2022)
Mentioned Content
Content
The Network Effects Bible (by James Currier of NFX)
Blog by Tomasz Tunguz of Redpoint Ventures
People
Gurjeet Singh (Co-Founder and CEO of Oma Robotics, Ex-CEO/Co-Founder of Ayasdi)
Satish Dharmaraj (Managing Director at Redpoint Ventures)
Notes
My conversation with Gary was recorded back in March 2022. Since then, many things have happened at Arcion. I’d recommend checking out:
The introduction of Arcion Cloud.
This article about data mobility on The New Stack.
This article about change data capture on Venture Beat.
This big product launch on Oracle log reader availability featured by VentureBeat
The article about the missing piece for the Modern Data Stack featured by Crunchbase
Arcion is launched with Databricks Partner Connect, featured by Datanami
Arcion is a proud sponsor of the Oracle Cloud World 2022 in Las Vegas, Oct 17–20. If any data professionals are attending the conference, they should stop by the Arcion booth to say hi!
About the show
Datacast features long-form, in-depth conversations with practitioners and researchers in the data community to walk through their professional journeys and unpack the lessons learned along the way. I invite guests coming from a wide range of career paths — from scientists and analysts to founders and investors — to analyze the case for using data in the real world and extract their mental models (“the WHY and the HOW”) behind their pursuits. Hopefully, these conversations can serve as valuable tools for early-stage data professionals as they navigate their own careers in the exciting data universe.
Datacast is produced and edited by James Le. Get in touch with feedback or guest suggestions by emailing khanhle.1013@gmail.com.
Subscribe by searching for Datacast wherever you get podcasts or click one of the links below:
If you’re new, see the podcast homepage for the most recent episodes to listen to, or browse the full guest list.
