CTO Corner #1: Why Tech Teams Need to Stay Customer Focused, Too
I work as a software developer at Jebbit, and I’ll be interviewing a new engineering department head for the CTO Corner. Andrew Groh is the VP of Engineering at ChoiceStream, a programmatic DSP. He manages a team of more than 20 engineers there. In this edition of the CTO Corner, I talked to Andrew about how to manage and scale an engineering team, how to best collaborate with other departments, and why engineers need to stay customer-focused.
How did you get into engineering?
When I graduated from college I had a computer science degree and I was reasonably happy doing it and it seemed like a good-paying field to get into, so I never really considered too much else. As I’ve gone forward, I’ve occasionally been an individual contributor and then I’ve gotten into management, and then I got back to being an individual contributor.
How was your transition from software engineer, to architect, and then to VP?
As an architect, but even more so as a VP, there’s so much more convincing other people to do it instead of doing it yourself. And if you have a good team, they are smart people who have their own opinions and sometimes they’re right and you’re wrong. If they don’t buy into it, they won’t do a good job with your vision. So a lot of it is learning how to balance that.
What are other challenges you face as a VP at a tech company?
Within my team, the thing that I also want to instill is a sense of ownership and risk taking, so I try very hard to support that. If people take a risk and screw it up, I try very hard to still support that because the value of people taking ownership so outweighs the cost of the potential mistake. You don’t want to have them come to you and do nothing on their own. I want to encourage people to take their own initiative.
Does this stem from you yourself taking risks and being supported by managers?
I certainly have had a mixture of managers over time but the ones that I really respected encouraged me to just do what needed to be done, to see a problem and identify it and fix it, without any hand holding.
What are some challenges you face in scaling your team?
I guess one problem for me is I don’t have a lot of people under me who necessarily want to be people managers. I have a lot of direct reports. In general, I only encourage people to be people managers if they want to. I try to set things up with team leads, so even if people aren’t reporting to them they can still handle technical issues. The mantra I always use is “smart gets things done.” I figure if I have enough of those people, everything will sort itself out in the end.
What I really get excited about is to see the output of engineering actually working — everything else is just stuff I have to deal with to make that happen.
My job is to keep everybody else effective and productive. Removing barriers — Do they not have the right monitor? Do they need to know how to do an expense report? — because I need to keep people productive.
How do you protect the engineering team from other departments while balancing collaboration?
I think within pure engineering, I really try to use an agile methodology and track everything. I encourage people to come talk and bounce ideas off each other, but if it ends up being real work it needs to get entered into the tracking system and get prioritized at our weekly meeting. I encourage conversations, but I try to dissuade something that could be a half day of work that I don’t know about.
How do you instruct your team and the company to think about building the right product and getting in front of customers at the right time?
It’s really important to understand how the visible parts of the system are going to work. It’s easy to latch into this technical discussion, but we really need to know what we actually want to accomplish first. That’s something that engineers can struggle with because they want to dive into the implementation without really knowing what the end goal is. If you only focus on getting tasks done, it limits your career because you can’t reach the architecture level. When I interview people I always ask about projects they’ve worked on and ask what they were doing and what they were trying to accomplish.
How do you transfer knowledge from marketing, sales, and account management back to engineering?
A lot of our technologies are also used by people internally. We have groups that are very similar to customers. So in that case, we’ll schedule an hour and watch you set up an ad campaign. Or say, show me what you’re trying to do. We’ll watch a person do a task and ask them why they do it a certain way. Engineers are good at figuring out how to automate things, so I like to hear our internal customers explaining why the product should work a certain way. As much as possible, I like my engineers to be able to watch the real customers of my product and see them go through real use-cases.
How does data affect your content strategy? Is there such a thing as too much data?
There’s too much data when you don’t have the tools to analyze it. But if you’ve got good tools, I don’t think there’s too much data. One of the key things is how you’re querying the data — if it takes me too long to query then I can’t move quickly. I have to be really sure I don’t ask the wrong question. But if I have a system that acts quickly, then I can do a much better job exploring that data and asking other questions without penalty. That to me is key, the response time of the system, so that you can work through these questions quickly.
For us, the data sometimes comes in out of order — so the question becomes how often does this actually happen? If it’s one percent of the time, I don’t care. But if it happens 50 percent of the time that’s terrible. So we can go look at our data and determine how often the data comes in out of order and determine how much time we want to spend fixing it.
For the past 10 years you’ve been in marketing and advertising — what is it about this industry that grabs your attention?
There’s money to be spent. Previously, I worked more with software that was sold to IT organizations. They don’t have a lot of money to spend, they can only justify it if there’s clear ROI, and it’s a long sales cycle. In advertising, there’s a lot of money. Yes, there’s a lot of competition, but there’s good money to be made here.
The other thing is I appreciate the data side of it. I like having a large data set that I can play with.
What makes a leader in ad-tech and mar-tech?
It’s such a tough question, because even if you’re doing something truly innovative, it’s hard to convince everyone that you’re real. It’s hard to stand out, because there are so many people marketing fluff. The next big player that comes out of our industry will have a clear way of making sure they are different and are offering something that no one else is.
That’s what’s hard about this industry — a lot of people are just saying “my black box works.”
How do you keep up with new technologies and trends?
On a purely technological perspective, I read Hacker News and threads on Reddit. They’re really into new stuff, so you have to read it with a grain of salt, but it’s a really good way of discovering new technologies. A lot of people might be talking about Docker, so maybe it’s time for me to look at how I could use it. Things like that.
I can’t write code anymore because my day is just too choppy, but I can go in and debug problems. So I will occasionally log into machines and look for errors in logs or do system analysis. It doesn’t keep my programming skills up, but it keeps my technology skills up.
Having spent the past 30 years being in engineering, where do you think tech is heading?
People always overestimate how fast things are coming along. That’s something people can make a mistake with, but they in general will get right where we’re going. I think better automation in transportation is coming — I don’t think you’ll see self-driving cars in five years, but I think 20 years from now you’ll see them at least in certain areas. Even that you can use an app that tells you when the bus is coming — that makes you so much more willing to use the bus. Little things like that change cities dramatically.
The other thing is it used to be really hard to do a lot of these technology plays that are now very easy. You don’t have to have a data center, you can use Amazon; things like that. For example, we have this site that we run polls on. It had been all internal, but now people can post their own. I was worried about people posting offensive things — but it turns out there are services and technologies that can monitor the poll content. That kind of evolution, having more and more of these services out there, will continue to make this ecosystem easier to build.
