Influencing People and Getting Your Items Into Their Backlog

Problem

As a senior IC, you don't have a team of engineers. Instead, you will have to influence other people to make things happen. While your seniority helps provide some "respect" and a positive halo you still have to make your case to PMs, EMs, and execs. Sometimes what you believe would be beneficial is very different from their current priorities. They are the ones accountable for their team's execution, while you are -- in spite of your seniority -- more of an influencer.

I was interested in having our teams work on a few reliability-related initiatives that I thought would be valuable for us and our customers. What I was lacking were engineers to make my ideas a reality.

Actions taken

First I looked at the data trying to understand what the problem was and understand potential solutions. I analyzed past incidents, the overall state of our system, and customer tickets. I summarized the problem and tailored my message to different audiences: Product Managers (PM) and Engineering Managers(EM), and executives/directors.

When approaching PMs to convince them to add my proposed initiative to the backlog I would typically say: If you do this, you will have fewer interruptions and be able to work on your product backlog for a longer -- and undistracted -- period of time. Also, your customers will be happier because your product will be working all the time.

When talking to EMs I would usually explain how this would reduce reactive work and continuous fire fighting and would result in a more stable product, which means engineers are happier and more productive.

When engaging in conversation with executives I would talk about brand damage, negative impact for the company, and how the system's reliability needs to be on par with other features. I borrowed a hugely persuasive argument from one of my peers, Reliability is the one feature that everyone uses -- if our system is down no other feature is working.

Being aware that teams are usually time-constrained I would pitch solutions in iterative steps, following the 80/20 rule. Instead of just throwing the problem at them, I proposed solutions that would dissect the problem into incremental parts that they could progressively address and that would not consume most of their quarter. With a concrete plan in place, I met with each individual team explaining what we specifically needed from them and also told them I would be available for all of their questions throughout their work on the project.

Lessons learned

• Align your own goals with the organization’s planning cadence. In our case, it happens quarterly. We discovered much of the problem close to the beginning of the planning cycle and thus were a bit late with presenting it to the teams, which made it harder for them to include it in their backlogs. The sooner you can present the problem to the teams, the better. You want to get to quarterly planning with them knowing about your important work.
• Whenever you are talking about improvements to "non-functional requirements" it is advisable to tie things to business outcomes and customers’ needs. I would frame the problem around faster and more confident delivery, making customers happy and even upsell add-ons we could offer. I pitched a business case for the add-ons and provided an example longer-term roadmap. Attaching reliability improvements to tangible things like $$ makes it easier to make a case for them.
• Talk to all stakeholders separately. When lumped together, people tend to behave with a mob mentality and are more likely to resist changes to plans.
• Don’t leave any of the stakeholders out. If anyone is missing, people will start to speculate why that is the case. That almost happened to me, but then I rushed to talk to a person that I thought I could omit.
• Don't just talk the talk, walk the walk. If something is important to make yourself available. I would block time in my calendar and would reassure people that they can ping me anytime. If it wasn’t important enough for me, I can't expect that it would be important for them.