Job 10: Hulu
Sometimes, the timing works out.
After eight years in Seattle, all but one at Amazon, we decided it was time to take a hard look at our future. Both the weather and the long commute to the city (we lived closer to Tacoma than Seattle and I drove ninety minutes each way on a daily basis) were starting to wear on me. If we were going to stay in Seattle, we wanted to move closer to the city before our oldest (now in seventh grade) started high school. If we weren't going to stay in Seattle, now was the time …
The message in my LinkedIn inbox could not have come at a better time.
How would you like to come to sunny Santa Monica and work at Hulu?
Funnily enough, I just went back in my LinkedIn history to get the exact quote, and the recruiter didn't actually play up the weather in Santa Monica at all. She didn't need to. I saw the phrase "headquarters in Santa Monica" and my brain did the rest. The message came in November 2014. For those who have never experienced a full year in Seattle, the sun only really comes out consistently for July and August, so November is early enough that you can remember that the sky is supposed to be clear and blue, but that only makes the realization that you won't see such a day until July of the following year that much worse. The possibility of moving someplace with actual good weather was more than intriguing.
The role at Hulu was to head up their Advertising Technology department. The team was small, but the opportunity seemed big. Advertising was about half of Hulu's total revenue, and in 2014 that amounted to about $500 million. Hulu was owned by the television networks — Comcast, Disney, and Fox — and they had recently made a major investment into the company and expected significant growth.
In March of 2015, we closed up the moving van, loaded the kids into the car, and said goodbye to Seattle. After two days of driving, we moved into an apartment in Santa Monica just a few blocks away from Hulu's office and started a very significant new chapter in our lives.
I showed up for new hire orientation on my first day thinking things would be off to a quiet start. They weren't. My hiring manager, as it turns out, started his three month paternity leave that day, putting more or less on my own to figure out what my team needed from me. It wasn't as bad as it might seem, though, because Hulu seemed to be full of supportive teammates everywhere I turned. My team had a Program Manaager who filled tons of gaps; our partners from the Advertising Sales teams were some of the best in the business, and always willing to help me understand their business and what they needed from my team to keep all our customers — advertiser and viewers — delighted with their experience on Hulu. The engineers on the team welcomed me to the team, helped me get up to speed on the technology, and treated me like I'd always been there. (In a way, I had been. The former manager and I shared the same first name. On his way out, the team had gone to Dave and Busters, I believe, and won a large stuffed bear. They named the bear Jeff and put him at the manager's desk, so I guess I fit right in.)
The first thing I remember noticing when I joined the team was that our team mailing list got a lot of requests for help from Hulu's advertising operations team. We were fielding tons of questions about how our campaign management software worked, why the ad server was under-serving (or over-serving) ads from particular campaigns, bug reports, and various other random requests. I remember at the end of the first week trying to review the email in my inbox and trying to identify which of all of these email threads had been resolved, and if any requests had slipped through the cracks. The team used JIRA for their planned projects and feature development work, but the email threads were untracked and unorganized. I think the only code commits I made as a manager on this team was to add some buttons into our campaign management tools that allowed the operations team to open tickets directly into our JIRA instance so this work went from untracked to tracked and managed.
How our team (and most of Hulu) planned and organized work was a theme I spent a lot of time and effort on during my time there. A few weeks after I started, all of the engineers from Hulu's remote offices — Seattle, Marin, and Beijing — flew to Santa Monica so we could spend a couple of days in some vacant office space we had rented and plan out all of the work for the next quarter in what we called Big Room Planning. (This was not an idea unique to Hulu, but part of the Scaled Agile Framework for Enterprises (SAFE).) There were about 200 engineers who spent two full days (many after traveling halfway around the world) sitting in vacant office space we had rented across from our permanent office, trying to figure out how to plan all the work, including dependencies and inter-dependencies, for the next three months of work. In the end, we had a plan. There were post-it notes on a wall describing the work items for different teams and pieces of yarn connecting notes to represent the dependencies between projects. Our PM team transcribed those physical artifiacts into JIRA for the teams to work over the next quarter.
After the meeting, I asked the engineers on my team what they thought about the process, and most thought it was a waste (or, more politely, an inefficient use) of their time. The advertising systems were almost entirely independent of most of the other systems at Hulu, so we were mostly able to finalize our quarter plan independently of the rest of the organization and, therefore, much more quickly. Over the next few months, I took this feedback to Hulu leadership, and within a couple of quarters, the advertising teams had a separate planning process working directly with our business partners in advertising sales. I continued to attend Big Room Planning together with my partners in product and program management (who usually did most of the work) to handle any coordination with the rest of the engineering teams, but our engineers were able to remain focused on what they did (and enjoyed) the best: building great software.
During the three-and-a-half years I worked at Hulu (I have yet to complete four contiguous years at any company) Hulu launched major features like an ad-free version, premimum channels, and live television on Hulu. Each of these meant a lot of work for our engineering teams, so collectively meant that the engineering organization overall, and advertising technology teams in particular, grew substantially during this time period. I went from being responsible for nine engineers, to an organization of over fifty engineers, testers, and support engineers, with another 12-15 program and product managers supporting us in those organizations.
Throughout most of my time at Hulu, one of the unique challenges we had to deal with was managing our relationship with Hulu's network partners at ABC, NBC, and Fox. Hulu was founded as a joint venture between these television networks, and remained owned with a one-third stake to each partner. (Discovery Networks invested in a 10% share while I was there.) This partnership gave us access to the content from those networks, but it also prevented Hulu from being entirely independent. As it related to advertising, the network partners maintained rights to serve about 90% of the advertising in their current content, so Hulu's customer support team fielded a lot of complaints from our viewers about poor advertising experience (usually getting served the same ad over and over and over again) in those shows, which we would have to take back to the network partners and their sales teams to resolve.
Over the few years I was there, we made a lot of improvement in this and other areas related to this partnership, but when Disney (the parent company of ABC) announced their intent to acquire most of Fox, including their share of Hulu, I was nervous about the potential implications this could have specifically for my role. Within a short period of time after the acquisition was announced, there was a lot of turnover in the most senior leadership at Hulu. Specifically for me, my entire management chain — my manager, his manager (the CTO), and the CEO of Hulu — announced they would be leaving Hulu.
At first, I had every intention of not joining the mass exodus from Hulu. There were pretty quickly some warning signs, though, that in my opinion showed Disney wasn't interested in allowing Hulu to continue to be Hulu. The interim CEO was a sports executive from Fox who showed himself to be largely out of touch and ill equipped to deal with differences of opinion. (Famously, I heard stories that he would second guess decisions of senior leaders at Hulu who were among the best in the industry in their respective specializations.) I got involved in a series of discussions with him related to questions about how we had ended up in a situation where we hadn't sold as many ads as we had minutes of viewership, which was a condition that, ultimately, was caused because some of Hulu's projects had exceeded all of anybody's expectations. (He didn't seem to understand that our forecasting algorithms weren't actually able to predict the future; all they could do was extrapolate forward based on past behavior with manual adjustments from our business teams.)
1: Lessons I'm Learning
Of all the places I've worked, Hulu during this time period stands apart as the most fun I've had while working. I'm not talking about the obvious fun of having ping pong tables around the office we would play to let off some steam — though we did have that — but even when doing work at the office, we were having fun. When you like the people you work with and are working on meaningful projects that make an impact to your business and your customers, doing the work is fun. I cared for and trusted the people I worked with, and that mitigated many of the daily annoyances that can make work so deeply frustrating. Yes, there were meetings that weren't as useful as they could be, and of course there were occasions of miscommunication or unreasonable expectations, but for the most part, we were able to work through those because we trusted one another.
As I was starting the process of growing out the team, I got approval fairly early on to add another manager of the team. While interviewing for this role, we found two candidates I really liked. Since we anticipated a lot more growth, I was able to get approval to make offers to both of them. Over the next couple of years, they both grew their own teams, and promoted engineers from their team into management roles, and then, in something they assured be was pure coincidence, they both tendered their resignation within a week of each other. (The two of them had become good friends and later ended up founding a startup together, but that was not in the cards when they left Hulu.) Overwhelmed at the prospect of so suddenly losing two strong leaders, I moved quickly to open up a headcount for a new senior manager to lead their teams, with the newly promoted managers reporting to that position instead of directly to me. While I really like the person we found to fill that role, I realized later that those newly promoted managers saw this as a vote of no confidence from me in their own abilities to run their teams independently. Their view of this was perfectly understandable, even predictable, and I wish I had not been in such a hurry to fill this position.