Five years is a long time to do anything. That's longer than a four-year undergraduate program (how about that analysis). I was fortunate to spend these past several years doing some marketing & growth work at DigitalOcean, one of the largest cloud providers in the world (probably bigger than most realize).
At work, I learned a ton. At home, I got married, had a daughter and moved out to the suburbs. That's the long way of saying a lot of my life looks different since I first joined DO (excluding the craziness that is the COVID world). All these things are closely intertwined in my brain and leaves me feeling grateful and sentimental about most of my time at DigitalOcean. It wasn't perfect, but it was a hell of journey.
While under quarantine the last few weeks I spent time untangling an array of observations and thoughts I developed over the years. Nothing too heavy. Mainly opinions about marketing to technical audiences, SaaS go-to-market, interacting with humans, tools and more.
"...this poor innocent creature had very little time to come to terms with its identity as a whale before it then had to come to terms with not being a whale any more. This is a complete record of its thoughts from the moment it began its life till the moment it ended it." - Douglas Adams
I've found it's always a struggle to really understand customers if I can't relate to what they do professionally. And I'm not an engineer, yet work in highly technical businesses. Soo....I attempted to address this in a few ways during my time at DO:
(1) Learn enough about writing code and tooling by following tutorials and stitching things together.
Hey I'm rolling my own (Ghost) blog! Not that managed shit (cool)! This is an ongoing journey and a real rabbit hole.
(2) Develop friendships with people who will explain things to you and not think you're too much of an idiot
This is the "explain everything like I'm five" category. I spent hours riffing with our Product Managers and Engineers discussing ideas, asking naive questions, whiteboarding architecture and reviewing Wardley maps (shout out to Brooke, Byrd, Andy and others who were fond of lame-ideaz.). I'd like to imagine that I was able to hang by "asking the right questions". In turn, I think I was able help them clarify their understanding by having to explain these concepts to a schmo like me.
(3) Talk to a lot of customers
Early on, our CMO and co-founder, Mitch, gave me an arbitrary goal to talk to 30 customers that quarter and build some personas.
Up to that point, we kept our persona simple: we serve the software development community. The formula is pretty straightforward - we serve developers. Developers need hosting. And we sell servers/software to host it for them.
However, understanding customers in broad strokes is like looking at a mosaic: easy to see the big picture, but hard to understand the details (btw the opposite is also true). It's the difference of listening to Jimi vs HEARING Jimi
As a result, I spent a lot of time digging into customer use cases, backgrounds and how they got to now. Without direct domain experience myself, gaining enough exposure to the nuances helped me better see the mosaic. Put a different way: to do marketing well, you need to hear Jimi.
A few things I began discovering during my early blitz on customer interviews over hangouts, Meetups on our 2nd floor, and so on:
- What the cloud really is (there is no cloud, it's just someone else's computer)
- Who our customers really are
- Why they care about us
- Why would any developer not choose Amazon, Microsoft or Google for a project at work
Go to market
It's important to develop an opinion. That's obvious. And this matters at all levels in a business, but specifically as you move up the ranks. The more senior you get, the more you are getting paid for your (ability to develop an) opinion and execution of it. I think this is one reason (but definitely not the only) that average tenure for marketing leadership is 12-18 months.
Content is crap. I read this a few months after I first joined DO and it immediately resonated. DigitalOcean didn't write content - we wrote technical tutorials and articles to genuinely help people. And they had this magical side effect of being the gateway into our product over time. To quote the author: 'We never call anything that’s good “content”' (case in point).
A little bit of dogma goes a long way. DigitalOcean's flywheel is driven through our Community. It's pretty simple: spread love + offer value and we will be rewarded with their business in the future. Easy to say, hard to execute. Our Community team had an extreme commitment on open source, thoroughly explaining topics like you are five and not sacrificing these principles at the expense of quantity. Created by Etel in our early days, the torch was carried by Tim, Brian, Lisa and a host of other terrific individuals. At times I would butt heads with our Community team, but I always recognized their steadfast approach was what made us great.
Horizontal solutions are hard. I've worked in a Marketplace business (a job board) and an enterprise B2B company (application monitoring company). While these products had their own unique challenges - the use cases were fairly understood. Cloud primitives on the other hand - compute, storage, networking and DBs - can be used in an infinite number of ways (I think this speaks to the expression of creativity through code). With so many areas to potentially focus, choosing where you spend your time is really difficult - any one investment in content/use-cases always felt too niche.
Differentiation through humanity. Good customer support, honest and thoughtful answers on Hackernews, and human-beings behind all marketing emails make a difference. This isn't easy, especially at the scale of a company like DO. It's the classic sum-of-all-its-parts. Sometimes referred to as Brand.
There is a secret sauce at DO. Like most things, it's just a nuanced combination of ingredients and execution. Often imitated, rarely duplicated.
Marketing to developers
Devtools are punk bands
Going to local punk and ska shows as a teenager and occasionally hanging out in the forums was my first exposure to being part of an online community (well, except for those early days of the web where you'd be huddled around a desktop with 4 friends in a random AOL chatroom). HackerNews threads around developer tooling looks a lot like these communities. Similar themes like:
- 'They're selling out' (they only care about enterprise)
- One-upmanship on finding the band 2 years ago before anyone thought they were cool (how many times have you heard someone say they loved X conference until it got too big?).
Do what you will with this anecdote, but to me it feels important. And if you think it's not, maybe you're not Hearing Jimi. (Or maybe you just think ska is stupid, in which case I'll be writing more on this topic soon)
Distribution (of shirts) matters.
Sporting your favorite band or festival tshirt isn't all that different from reppin' your favorite devtool or conference shirt/laptop sticker.
Someone should write a case study on T-shirt investment and returns on capital. Seemingly all the recent waves of enterprise software have had massive winners in distributing tshirts & stickers. Look at the early days of Splunk, New Relic and most recently Datadog (and they happen to do some monitoring/logging too). And this Github store! People want to wear the communities that they support. DigitalOcean is on this list too: last year over 50k developers successfully completed the Hacktoberfest challenge and received a tshirt.
A few other thoughts
- Offer value before asking for anything in return. What an obvious statement. This is true in literally all aspects of life. But it seemingly really matters when there's a lot of competition (and where code is open source). There's no coincidence as to why OSS is the distribution channel for most developer focused companies
- Word of mouth is a powerful thing. Humans (in this case, developers) like to do things and tell others about it. Internalize this but don't play into it too hard. It’s fragile.
- Shelf space matters (fantastic read)
- Nights and weekends is where it all happens: low-risk items like personal apps and side projects is where you'll often be evaluated first. Not a novel statement. But DigitalOcean is proof this can be a massively successful strategy if you are patient.
Tooling & Process
I bought a lot of software at DigitalOcean. Like all companies, we always wrestled with building or buying solutions to help solve problems. On the marketing front, we typically landed on outsourcing the "undifferentiated heavy lifting" (hat tip to Andy Hats who loved to describe things this way and knows this space better than anyone), and therefore we were early adopters for many low/no-code tools. This led to heavy usage of tools like Segment, Tray.io Retool, Clay.run and Zapier. A few random thoughts on tooling:
Generally: tools are cool, but so so rarely solve anything without a clear understanding of the problem and/or desired end state. (Yes, for both building or buying)
Specifically: project management tools don't solve your process problems. Use a google doc and think about how you want to organize your world and then find a tool. And most of these tools suck for the collective group of users - not because their PMs don't know what they're doing, but because it's hard for 5 people to agree on how to organize their minds. (Or you need strict enforcement from management - which sucks, but is probably necessary)
Go-to-Market teams have fantastic options that allow them own their domain and become less reliant on product/development. Learn the basics of how APIs work (<– you should listen to this guy) and you'll discover a step-function improvement in your ability to solve problems without needing eng-time. More importantly, most developers want to work on new problems, not ones that have been solved time and again.
There are always opposing but related stories playing out in software-land between 'best tool for the job' and 'horizontal solutions'. The 2020 version is:
- The rise of vertical/specialized tools per function / use case. For example, Catalyst for Success Teams and Outreach for outbound SDRs vs just using Salesforce's suite of products. These are enabled by customer data being made available more easily via tools like Segment (which also is enabled by a massive shift to all services being delivered via APIs)
- Low/no-code tools: they look a lot like point-solutions and platforms at the same time. They provide a canvas like Docs and Sheets, with the extensibility (lego blocks) of Salesforce or JIRA. So you can do a lot if want to because the 'rails' are quite flexible, but they also just look like prettier note taking or spreadsheet apps when getting started.
Challenge your thinking
My team threw me a 'roast and toast' as a sendoff. It was one of the more speechless evenings I've had, which is fairly rare. (Thank you Michelle, Fung, Andy, Gabby, John, Hollie, Tim, Daniel and the rest of my team over the years). We spent some time discussing the diverse makeup of our team:
- The variety of places in the world we grew up, family backgrounds, and so on.
- Learning styles and educational backgrounds, including a combination of MBAs, comp-sci majors and English majors.
- And what we do with free time: writers, hackers, volunteers, activists, musicians, golfers, partiers, high and low brow-TV watchers, etc.
When everyone was clear on objectives, the combination of these experiences and interests led to very interesting outcomes The collective diversity of opinions and ideas of this group challenged and shaped my thinking over the years. If we're all just remixing, stealing and combining ideas, then I owe a lot to my team and peers.
Changing gears...I had never met anybody quite like the Uretsky's (Ben and Moisey founded DigitalOcean). They questioned everything. Often, for the right reasons. And sometimes to a fault. But they forced you to develop an opinion/position, and defend it across multiple fronts. It didn't matter if you were right or wrong: having conviction and a thorough argument was most of the battle. While this was frustrating over the years, they helped evolve my thinking and preparation for the better.
How to get employees excited about a seemingly boring topic (hosting)
- The David vs Goliath analogy really drove a sense of team and purpose. I loved working for the little guy going up against the most powerful corporations of our time. It put a chip on your shoulder and felt like you were really sticking it to the man. It really forced you to think differently.
- While I wouldn't describe DigitalOcean as a mission-driven business, the employees tended to operate this way. There was an extreme passion for community, education, serving underrepresented groups and technical literacy. The key word there is Passion.
- Things are fun when a community gives a shit about you. While there are a number of reasons for DO's success, this is what drives the flywheel. This is what makes marketing, sales and product-development fun. How many companies or products do you genuinely support? You can probably count them on one hand.
- "Developers are the new kingmakers." This was the lens through which we viewed the world. This isn't a strategy, but a core assumption. Having this clarity felt like an (exciting) secret.
- Early on, DO invested in its people. I was fortunate to spend a chunk of time with a career coach (thanks Rachel). This was easily the most valuable professional development exercise/perk I've experienced in my career. These sessions crystalized the importance and value of 'asking the right question' and having the right instincts on threads to pull. Having a therapist with the context of your business and cast of characters was very impactful both personally and professionally. Side note - my original boss at DO, Aaron, is also a fantastic executive coach and one of the better people you'll find.
Why I joined Grafana Labs
I joined Grafana Labs a few weeks ago as the Head of Self-Service (my wife asked if I'm now an ice cream man? 🍦😂). Generally, I see similar ingredients at Grafana that I saw at DigitalOcean 5 years ago. Specifically, wide distribution (OSS) + genuine love within the community + riding a software wave + a commercial team that seems to really get it = strong combination.
- Open source. I've seen the value and the fun in being part of a community built around software. I also recognize the distribution opportunity OSS presents.
- Spending time around developers and DevOps teams over the last several years, I've seen how Grafana is becoming the de facto dashboard within companies (especially with the rise of containers / Kubernetes, and the like)
- The Grafana flywheel rhymes with DO's: be deliberate in spreading love. Offer value. And good things will come. I've seen how impactful this approach can be on people's lives, employee morale, and ultimately the impact on the business
- Finally, and probably most important: I got the chance to know Doug, Raj, Matt and some of the team for quite a while. They are good humans who are great at what they do.