Monday, February 11, 2013

Gnip's Organization Evolution

Today timehop showed me a picture I took two years ago of a guess at an org chart that I thought Gnip (a social media company) might look like at fifty employees (approximately where we are at the moment). The timing may be off a little; I may have drawn this up a few months before I took the actual pic. Gnip is five years old, this month.


I drew it up after one of our board members suggested the exercise; "just imagine what you think you might look like." At the time, the thought of our business supporting fifty people was so far out there I could hardly imagine it. I went through the exercise, and the above picture resulted.

The actual org chart of what we look like today is of course different. Just how different is a matter of perspective. When I look at the above sketch, I'm surprised at how well it held up to reality. Some thoughts on how things actually evolved follow.
  • I learned at a First Round Capital CEO summit that the CEO should never draw up an org chart with the CEO at the top. Instead you invert the pyramid like Rand Fishkin did here. An inverted pyramid represents reality. The one I drew two years ago isn't how the CEO role works.
  • We almost never look at our team in org chart form like this. While everyone has their domains and expertise areas, we look and feel relatively flat.
  • A critical hire wasn't even on the map represented two years ago; the COO. Half of Gnip reports to our COO. I'm a technical CEO. What I'd originally mapped out, distributed much of the "business horsepower" across a handful of VP roles. Things didn't fully evolve that way in reality. While we have incredible talent (that I would hit the battlefield with any day of the week) in the leadership positions that drive Sales, Marketing, and Finance, our COO owns and manages those functions outright. Looking back, I'd likely have screwed things up with the above model; we'd likely have had stunted growth. Knowing what I know now, all of that is likely too much for a technical CEO, "alone," to pull off properly.
  • The COO, VP of Product, and VP of Engineering report to me directly.
  • We do not have an dedicated HR person. We split the functions across our Finance Director, and a remote HR contractor. I like it this way still.
  • We do not have a CTO. That level of technical leadership/representation is driven by myself ("technical CEO") and our VP of Engineering.
  • The code-level groupings for engineering are kinda close to where we are today; closer than I would've guessed they'd be after two years. Notably, the amount of continued code that has to be written in order to ensure reliability of the system, including keeping pace with continued massive growth of social data volumes from publishers, was underestimated. We feed the mainline real-time software beast a ton, everyday. I naively thought we'd write that code, and just have to babysit it. It requires constant revisions and new approaches.
  • The Product org looks today just like I'd mapped it two years ago; wild. It's small and simple, so I guess I couldn't have gone too far off the rails.
  • Finance is 2x bigger than I'd predicted. Wishful thinking at the time I suppose.
  • Marketing is bigger than I'd guessed it would be. Not surprising I suppose for a technical guy's prediction.
  • We're only just now formalizing Sales Engineer roles. We've gotten by with engineering and support handling that function until now.
  • Data Science wasn't even on the chart two years ago, and today, we have a small group doing precisely that. Gnip's Data Science function is centered around understanding the publishers we integrate with in order to ensure we can effectively market/position/sell their data. Gnip's often in a market maker position, so we have to know the data we work with inside and out. The degree to which we ultimately needed to invest here was a surprise. When I look at it now however, I can't imagine having drawn this map up without said function. That was a miss.
  • We also have an office manager and receptionist today.

Wednesday, February 6, 2013

Actions Speak Louder Than Words

We've all heard this adage a million times. I've been saying it at home, and at work, a bit over the past few months.

My son is a decent carbon copy of me (poor kid!), and if you know me, I'm generally a man of few words. If I find myself telling you about what I want to accomplish, I've already failed in my mind. I need to be showing you what I've already done.

I lead by my actions. My son tends to do the same. My daughter on the other hand has thrown out the proverbial "I'm trying" a few too many times over the past six months. The result, my inner Yoda comes out and I engage in "there is no try, only do" conversations with her.

Don't talk to me about trying. Don't talk to me about what you want to do. Don't talk to me about what you've accomplished or want to accomplish.

Get it all done and show me. That's obviously a little harsh for a child, but when you're an experienced adult on a high-functioning team, that's table stakes.

I'm way oversimplifying here of course. There is such a thing as empathy, and planning a course of action, and assisting, and sharing experiences in order to come up with a plan of attack to get something done (as an individual, and as a team). Those are all crucial pieces around getting things done, as individuals and as teams. Engagement around speculation and desire is healthy and required in order to succeed.

I'm just sayin' that actions speak louder than words. We are judged by what we did, not what we talked about doing.

Monday, February 4, 2013

When Are You No Longer A Startup?

In the past six hours, the topic of whether or not a company is a startup, or not, has come up a half-dozen times. I've been meaning to post my thoughts on the topic.


Gnip (a social media company) is a high-growth small/mid-sized company. Gnip is no longer a startup because...
  • our annual recognized revenue became "significant"
  • we successfully made it through our first clump/wave of contract renewals
  • the majority of code in production became tied to said significant revenue
  • we have a lot of customers, and aren't dangerously clumped around a small number of them
  • the product/engineering stuff that we told ourselves "we'd deal with if it became relevant downstream," became stuff we had to to deal with
  • things started working, and you could place a low-risk bet that they would continue to do so
  • the risk of the core product/revenue stream going away became relatively small
  • for the current version of the company and product to succeed, our reliance on outside capital infusion went away
  • we realized profitability was in our control, and something we could have, whenever we wanted to reduce our expense growth rate to something below our revenue growth rate. note, I did NOT say expense cutting
  • after the third time I'd mentally stressed about having a consecutive month that missed expectations, I realized the stress wasn't warranted (we'd exceed expectations *again*). at that point, it was clear to me there was a rock in our repertoire that was rolling downhill (instead of us struggling to push it uphill). even on an upward trajectory there are moments in which the growth rate fluctuates. what matters is that they're all net positive rates
Gnip has plenty of hard work to do; we're only at rev 1 of the company/product suite. I sometimes think about the new stuff we need to do as "startups within the current company." There are lots of relative terms in the above bullet points; I'm aware. This is all just a state of mind for the most part.