out there on newer platforms and start talking about our own messages.
I encouraged every single employee in the CIO Office to talk about the projects that they were working on. Not because I wanted media attention, but because I wanted real-time collaboration with other sectors. Often in government, we see policy consultation with a beginning date and an end date. This way you are just limiting collaboration opportunities by not engaging continuously. With LinkedIn, Twitter, or other social media platforms, we could collaborate 24/7 now. The government way of doing business had not changed so far with the tools we had at our disposal.
Our way of being open by default meant saying continuously, “Hey, this is what I am working on.” I wanted the vendors to know, I wanted the provinces and regional governments to know, I wanted other departments to know. People just had to start talking about the work they were doing. It was great, even if it is a double-edged sword because it adds more attention to your work.
It also meant that we had to start talking about the failures better because some of these projects did not work well. Government does not like to talk about failure, right? Like I keep joking, fail is four-letter word that starts with the letter f and that you are not allowed to utter in government. We had to make it known that failing fast and small is better than failing slow and massive, like Phoenix.
You Did End Up Taking the Work of Saving Phoenix into Your Own Portfolio. How Did You Go about Fixing It?
The more you start to demand that the system of government or a large corporation change, the more you are preaching about it, the more people will say, “Well, do something.”
We were never charged with fixing the existing Phoenix; I still do not think that was doable anyway. Nothing worked in it: government staff were getting overpaid, underpaid, not paid. There was no architecture done, no data, no testing, no backup plan, you name it. There was no way to get out of that mess in any other way than starting a new system.
In parallel, we had been telling all the procurement officers in the GC doing digital procurement to please stop giving specifications out and instead give out the problems for procurement instead. Then the really smart people that do tech outside of government can come up with a whole bunch of different solutions. And we might use them all, why use just one? So, we gave ourselves the task to run the country's most important open procurement up to that point to replace Phoenix. We started putting our policies to work to execute, pulling all the tools we had brought in and out.
We went to the market and said, “We want a cloud-based software-as-a-service human resources and pay solution; who is interested? We got about a dozen interested parties in a room. Then we said that the process is going to be like the Hunger Games, with gates to go through where we would up the requirements as they passed through the gates. We did not want any paper response; we wanted to see the actual software change and work against our requirements at each gate.
The procurement people at first thought we would have to pay the suppliers as they were going through the process. I said no to that. So, we did three gates, five months, the last gate was with real anonymized user data and suppliers had to show that they could pay the people properly. We ended up with three final suppliers—not by prescribing a solution, but by working with suppliers at each gate and working from the problems up. In less than six months, the government had three viable options, whereas nothing had been achieved for years before.
We put our money where our mouth was, and we showed what can be done with a different approach. It shows that you have to get your hands dirty on some of this stuff.
How Did You Build or Rebuild the GCIO Setup to Achieve the Changes?
Just for context's sake, Government of Canada has up to 25,000 people who work in Information Management and IT professions in forty-three departments. The Office of the CIO is about two hundred to two-hundred fifty people to coordinate privacy, access to information, AI, data, service delivery, architecture, project oversight for major projects, open government, people hiring policies, you name it.
I had a whole bunch of silos within my own team. I could not ask the rest of GC to integrate digital with everything if in my own team, say, privacy people did not even talk to architecture people. Or if tech people were creating products without the privacy and policy people along. We had to get our functions working together before expecting the rest of government to do it.
I also remade the immediate team around me. We went from eight to ten direct reports to three as I built my team on the three legs of any kind of change in a modern organization: people, governance, and technology. I consolidated all the policy work together, all the people work together, all the tech stuff under a government CTO to oversee architecture and project oversight. Over time, as we continued influencing the system and got more money to do more things, those roles became more senior.
Also, I wanted to make sure I did not become the single point of failure myself. If you have eight to ten divisions with silos and you are not around to coordinate and orchestrate, like if you get hit by a bus, there will not be a lot of integration left. My three reports had to work together so that if I was hit by a bus, the next day the organization would continue delivering in an integrated way.
I also needed to change the culture in the office. I needed to have my direct reports comfortable with the fact that I would not go to them for updates but would have the team come into my office for an update—like we did with Skunkworks. It takes a lot of trust to work that way. They needed to trust their teams to brief properly. So, I did on purpose create a different flow of conversation within my office, leading to everybody able to replace everybody or at least understand everybody's job.
Getting past silos and getting the conversation flows going took a lot of informal effort. You know, people want to have fun when they work hard. If they have fun together, they are increasingly working better together as well. The office had never even had a Christmas party. We did an offsite party and against all doubts a whole bunch from our team showed up. We started doing Winddown Fridays, basically once a month sitting together and going for a beer or glass of wine. Winddown Fridays even started attracting people from other departments to join. All of a sudden, we were running out of room.
It meant that we got people talking and it reduced the government hierarchy and silo structure. It was sort of our attempt of attacking the Industrial Age organizational structure of the government bit by bit all the time. Unless you do it relentlessly with a series of micro adjustments, like our Winddown Fridays, the system will win because it has been there longer than you.
I Know That You Also Made a Big Push for the Team to Work in the Open. How Did That Work?
We would have our GCIO office all-staffs quarterly and have them mandatory. We also started recording these meetings, because we were the policy center, and I wanted the operational departments to know what we were talking about. We would livestream the meeting or just send the record out to all other departments. The direction setting to me is always about open and transparent communications, plus letting people challenge the direction.
With my own team, we would also do periodical “ask me anything” hours so that people could air out some of the issues there.
Openness and transparency are huge because they mean that we are also forced to deliver. Enabling the staff members to talk themselves about their projects and their programs was a small thing that turned into a big deal. They made themselves accountable by talking about their work online. The effect of open-by-default working on self-discipline at the staff level became apparent through the work we were doing.
We also very openly sang praises about
