As the adage goes, the best laid plans go awry, and that also applies to building software. The planning phase, including maintaining alignment, is critical in engineering, but even for the most mature teams, plans can change and evolve as they’re executed.
Engineering leaders need to keep teams aligned with the overall goals of their organization, and that includes setting and managing expectations for stakeholders so they won’t be blindsided when roadmaps need to change. How do CTOs successfully align on strategic priorities, navigate planning, and course-correct when things go off-track?
As part of Code Climate’s Velocity Next event, we invited engineering leaders Catherine Miller, CTO at Flatiron Health; Juan Pablo Buriticá, SVP of Engineering at Ritchie Bros.; Ryan Grimard, SVP of Engineering at EverQuote; and D Orlando Keise, Head of Banking Foundational Platform at UBS, to share their experience leading engineering teams while keeping the C-Suite at ease.
Read on for highlights from the panel discussion, led by Peter Bell, CTO and Founder of CTO Connection.
Peter Bell: An engineering leader’s responsibility is keeping the team aligned with a larger organization, and that starts at the top. What do you do to ensure that you’re on the same page with the rest of the C-Suite around things like resourcing, roadmap planning, and strategic priorities?
Ryan Grimard: Our teams use the Scaled Agile Framework, SAFe, for our planning and execution, and we involve our business leaders way in advance. They’re helping us with strategic planning for the company, and then our product and engineering teams are working through portfolio management and bringing things into the conveyor belt. When things are ready for final prioritization for a particular quarter, we use that process to go through “big room planning” and a one-week prioritization planning process. The end of that is effectively a confidence vote for all of the teams. We have 17 Agile teams, and the business leaders are in those meetings and hearing the confidence vote, and giving the thumbs up that they agree that these priorities that the teams have picked actually match up with the OKRs that we set at the company level.
Juan Pablo Buriticá: I have two techniques. One is to force the C-Suite to compromise on a single thing that they care about through a guiding principle. So, do we care about speed, do we care about quality? And then, I use that guiding principle to force decision making on the group for alignment.
The second thing is, a set of cascading strategies: you have business strategy, the product strategy that cascades from the business strategy, and the engineering strategy, which should enable both. And then, it forces resourcing, staffing, and prioritization to be aligned with the guiding principle. That guiding principle is the tiebreaker for everything.
D Orlando Keise: What I’ve found is important is to align on the mission. And I’m using “mission” in a sense that’s more than just the objective. I actually mean it almost in a military sense. What is our target? What are our threats? What’s the landscape and the terrain that we’re operating in? I think we all know that we’re going to come up with a plan, that we have a roadmap, but I find that the plan is not going to survive first light with the outside world.
I want to make sure that when that happens, we’re aligned not only on what we’re trying to do, but why we’re doing it, and the environment that we’re doing it in. Because when we start changing that plan, if we’re not aligned on all those other things, we’re going to run into problems.
Peter: What kind of conversation do you find the most difficult? Is it when you’re in the planning phase, and you have to say, ‘We’re not going to fit all that into this quarter,’ or is it once you’ve said, ‘Sure, we’ll do this by the end of the year,’ and then it’s December 12th and you’re realizing that Christmas might have to be canceled this year?
Catherine Miller: Planning conversations are about feelings, and execution conversations are about data. I like the execution conversations better, because the planning conversation ends up being a very abstract conversation that is based on trust. The more trust you have the better that goes, but you’re all just guessing, and at the end of the day, you are trading on some relationship or some extrapolation, and you know it’s wrong.
Then you get to the execution, and first of all, no one actually expects it to go on track, but what I love about execution is, you can check in along the way, you can see how it’s going, and it forces a conversation. What are you going to do? We literally cannot hit this deadline because of where we are. That is a fact. There is no hoping or wishing that will make that go away. So you’re forced into decision-making in a way that less mature teams often avoid at the planning stage where they just hope everything will be okay.
Ryan: I would say that our company has really evolved over the last two or three years. It used to be a much more difficult conversation upfront when business leaders asked, ‘These are the strategic priorities; how many of these can we complete in a quarter?’ Because they weren’t necessarily involved in the process that I described earlier, they were surprised when we couldn’t do everything at the same time and get everything out the door in that particular quarter. So since we’ve introduced this process, I feel like the more difficult part of that is through the backside, or partway through the quarter when the conversation might be: Why aren’t we executing? And those really get broken down into retrospectives. When something doesn’t quite go right, teams will do a full retrospective. They will come up with a root cause analysis through that retrospective, and line up whatever we need to fix, then present that to those business leaders. I think that builds confidence across the board.
More recently it’s been boiling down to either high levels of tech debt for a particular team, or our team isn’t onboarded to what we call our “paved path,” or our delivery pipeline.
Juan Pablo: I like solving problems by preventing them from happening, so I’m usually more uncomfortable when things have gone off track, because that means that we didn’t prevent what we were trying to prevent. Whereas, when I’m on the planning aspect, I feel very comfortable. Plans are lies we tell ourselves to get to the next stage and motivate our teams and motivate ourselves, so I have a much easier time on that end than the former.
Peter: When you realize for whatever reason that your team isn’t going to hit a date — there’s some commitment you’ve made, it’s clear now that it’s not going to happen — what’s the first thing you do?
Juan Pablo: The moment I find out, I broadcast that state to the organization, because rather than rebuilding trust, I prefer not losing trust. I’ve learned that people will usually handle transparency better than we think.
Since we’re building software and things tend to go wrong in software, I like building a culture where being off-track is not uncommon, and people don’t react with fear. We’re a good engineering team if we’re good at reacting quickly.
D: I want to underscore something that Juan Pablo said. The very first thing is to communicate it. The second thing is to understand what kind of problem we have. Did we plan poorly for the amount or type of resources that we have? Did we have great resources but the ground shifted under us?
I find that people think they’re solving the real problem but they’re really solving some symptom. So, I try to go a few layers down to find the real problem that caused us to be off track to begin with, and attack that.
Catherine: Part of what I would do is figure out whose problem this is to solve. Is this something that I should really be [calling on] one of my VPs or even directors for, and setting a framework for them to solve, or is this significant enough for me to dig into directly?
Ryan: Keeping it amongst the team is important, at least for a place to start, where the team is a handful of engineers and the product manager or the product owner. There’s a lot of trust that should have been built up over time. We’re pretty big on keeping teams whole and not rotating people around unless there’s some sort of internal mobility opportunity for folks.
Peter: So you found out things are off track, and you have to prepare for a difficult conversation with a senior stakeholder. Are there any things you do when preparing for that to try to make it just a little more likely to go smoothly?
D: A lot of my preparation for moments like that happens long before the moment. A lot of us have used the word trust at various times, and that’s huge. I’m trying to establish trust with stakeholders and partners from day one, because so many of these moments are going to rely on trust to go the way they need to go. The other thing I try to do in the moment, when it comes to explaining a difficult truth, is to take any of the emotions out of it. I try to lay out something that’s very fact-based.
Cat: I very much agree with the emotional regulation piece as a key part. If I have the leeway and the time, the best thing I can do is I can sit down with my coach. As we talk, I lose my emotional attachment to things, until I’m just thinking about options that are good and bad and what the cost to the business is.
Ryan: Data gathering, absolutely. Pre-reads I think are huge, meaning sending out what you plan on talking about beforehand, and making it clear that reading the pre-read is a requirement. You shouldn’t be going into the meeting and then learning about what you’re about to hear about. The pre-reads give you that opportunity to be informed before the meeting, and if successful, meetings are for making decisions.
We’ve run into situations where the pre-reads are pretty in-depth, all of the decision making and our suggestions on where to go next are already laid out, and the meeting is five minutes long.
Juan Pablo: I use paper as my coach. I believe writing is thinking, and I try to anticipate questions that might come from the conversation. If I’m walking into a board meeting, I try to minimize the times I say: I don’t know, let me get back to you.
I have my talking points. What is the outcome of the conversation that I’m seeking? Am I trying to inform? Am I trying to convince? Am I trying to drive to a decision? Based on that, I bring options or I shape my talking points.
Peter: Things didn’t go according to plan, you missed the deadlines, how do you rebuild trust both within your engineering org, and with your executive stakeholders?
Cat: It’s not inherently breaking trust to miss your deadlines if you’re being transparent about your metrics and what you’re seeing as it goes along. It’s mostly that transparent communication that will leave no one surprised.
Ryan: I think rebuilding trust is really about consistency and showing continuous improvement. We’ve been using some of the DORA Metrics that Code Climate Velocity has been putting out.
We have custom reports for DORA Metrics, and they’ve been really awesome. Being able to show that the team is performing, that the issue that we ran into was kind of out of our control, and showing there’s an upward progression to how they’re performing…it’s big to show that they’re not not performing.
Juan Pablo: I usually lead product engineering teams, and so the focus is less the plan or the deadline and more the outcome, so ensuring that the teams understand the goal that they’re supposed to be driving. And in product, your goals are so variable, and you can invest 6 months launching a feature that tanks or spend a week and suddenly you have a great product, and so what I care about is… less about times and deadlines and estimates and plans and more about, are you being impactful and are you shipping as fast as you can so we can learn?
D: I would underscore that rebuilding trust is really about being reliable in the things that you say. Communication is important. We had talked about diagnosing the real problem, and now when you’re going and acting on that and executing. The second time around, are you hitting what you aimed to hit based on the problem that you diagnosed? Hopefully you are doing a good job of identifying what that real problem is and then when you drive the team to execute towards that milestone, it actually solves what you said it was going to solve … as you do that you’re rebuilding trust.
Audience member: [To Juan Pablo] Tactically, how often are you resetting your principles because you can’t make the mark or realize priority two has to be priority one?
Juan Pablo: I’m doing a good job if I’m not resetting them that much, or they at least have an 18-24 month lifespan. I do start from a mission. When I was at Splice, the mission of engineering was to enable Splice to learn faster than the market before we ran out of money. So that was driving engineers, that was our mission, anything you do is in service of that, because we’re serving musicians.
And from there, the principles that we set were about sensible choices in technology. We were like, yes it’s fun to learn, but our job is not to learn; our job is to learn faster than the market. I also crowdsourced feedback from the team as far as what resonated with them in regards to our principles, because usually I don’t have the full perspective. If there’s ownership from the group, it’s easier for them to also embrace the principles.
Audience member: There’s a known standard in the tech industry of listening to your customers versus understanding business objectives from leadership. What has been your experience trying to carve in both business objectives and customer feedback?
Cat: More and more, I think every dichotomy in tech is long- and short-term planning. For example, I think that tech debt is long term and product and desires is short term, and I think what you’re talking about a little bit here is what your customers are beating down your door right now, whereas presumably your business metrics are about your long range plans, what your business needs to grow, what our burn rate needs to be… And I don’t necessarily have an answer for you, but I think dispassionately thinking about how much are we investing immediately — how much do we need to do this month or this year, versus three years from now? — is an interesting way to think about your portfolio across a lot of different things like this.
Ryan: I worked at a company where it was very referral based, and the business metrics were: How many referrals did we make, and how much revenue did we make from that referral? The other side of that is, how happy is your customer? Does a happier customer actually produce more revenue? Maybe not in the very short term, but repeat customers, and happier folks in the marketplace of that particular referral business goes a long way. The business is able to interact with a consumer that’s happy and potentially upsell them and so on.
Juan Pablo: Usually we’ve done a good job as a leadership team if we’ve set business metrics that empower teams to take these customer needs and act on them. Because customer needs and business metrics should not be diametrically opposed, and if they are, there’s something wrong.
Actionable metrics for engineering leaders.Try Velocity Free