Leading engineering with vision

I recently joined another call through CTO Craft about leading engineering with a vision. The previous one is here.

The call was hosted by Andy Skipper and the panel was Meri Williams (CTO Pleo), Rob Zuber (CTO CircleCI), Jossie Haines (CTO and Engineering Leadership Coach).

The following are my disorganised notes from the call:

Rob - Often we realise too late that we need to set a vision. People start off aligned but then realise too late that things have drifted. Over time there is a lot to do and people get worried about so many different things and are unable to do all of them. So people need to buy into the vision and it will provide clear direction on what should be done and what should be dropped.

Meri - Gave an example where the team started off well aligned. It tripled in size in a year. It became much more challenging to stay on the same page - need to identify the where and why? She introduced the technical and architectural vision above the overall vision.

Jossie - A vision is 2 to 3 year direction - it’s about the why that drives the what. A strategy - needs to provide enough clarity, resourcing, build v buy, etc. The strategic roadmap - is for the next 6 months and includes the milestones. Each layer gets more tactical. The technical vision has to serve the product vision.

Vision or strategy first?

Meri - A change in mission can impact both. You can live without a business vision but not a product vision.

Andy - Product vision is based on what the customer wants. Tech vision is separate - it covers things like compliance, performance, size of team, etc.

Rob - Product vision always informs the technical vision. He explained that you can still have the same philosophies independent of the actual product being produced and it is important to write them down. Time horizons are longer on a technical side than a product side. Think about “What are the reasonable things we might be asked to deliver?” Understand we will iterate and adjust so take that into account when building.

Communication

Technology vision should be easy to communicate to technologists.

Meri - Write the documents so that others can stop reading when it stops being relevant to them. Start with the why then the what then the how and when (the latter 2 are mostly for engineers).

Jossie - Engineers want to make an impact - they need to understand why they are building what they are. Create a framework and get the team to co-create the details - especially the how and when. This increases their buy in.

Rob - Understand the market and industry to see where it is going. You need to speak the language of peers (financial terms for the financial people, etc.) when relevant - most of them are not engineers so won’t understand technical terms.

Product / Tech

Meri - Sometimes the product vision can change too quickly for the others to keep up - usually from the CEO rather than CPO. They need to understand the “tax of change”. It can have an impact as it can be disheartening, there are sunk costs if it is never delivered. When the mission is very clear it happens a lot less.

Jossie - Founders often don’t have the experience so train them. You need to have flexibility.

Visibility

Rob - Visions tend to be the north star direction. You need to embed it in the day to day.

Meri - Tech leadership team - what is needed to make this real? A 6 monthly check in was to make sure progress was being made. Push the vision through all hands, roadshow, discussions. She needed to present to the board and measure progress - used the DORA metrics - it helped to sharpen the focus. Progress > Perfection.

Andy - He would go into a room with a couple of developers and set a short term vision. Use champions to do the pushing and reminding.

Tension - vision v delivery

Jossie - Ideally they are aligned. There is a difference in the product v tech vision length. How do we build it step by step so we don’t stop delivery of features while building the vision? Need some buy in from product - it’s a balancing exercise, we still need to deliver features.

Meri - Technical debt is like financial debt - not evil, but when interest payments get too much then you are in trouble. Show the CFO how much time is being spent on keeping the lights on. You can’t just do tech debt though. Need to spell out clearly the reasons. Logic is not the only way to persuade people - compromises, expectations, benefits.

Andy - You can create a vision but you must also sell it as well.

Rob - If I am not getting what I want (the best thing for the business) then listen for feedback - something isn’t landing. How can I help people see the challenge for the business? I am failing to provide what they need.

Meri - There is a vision of what type of engineering organisation you are - what career path do people have etc.? Not as concrete but it is really important.

Jossie - This is crucial at the moment with rolling out AI. Lots of implications on the SDLC. A lot of this is a leadership challenge. For example clarity on things such quality is important (PRs reviewed etc.)

Andy - Vision components - Culture, team size, team building, amount in parallel, etc.

Rob - If a senior person pushes back and doesn’t buy in. Not uncommon if the team has scaled. Understand their motivations. Find the most believable person who disagrees with you and find out why (they might be having a bad day). You want your most senior people to say what is broken.

Andy - The order in which you communicate the vision and the people you inform and when is important.

Meri - If you get buy in from others then they might fight the case for you. Understand who is the most influential in the company - it might not be given by the title.

Jossie - Start any difficult conversation with acknowledging that they might truly care rather than wanting to pick a fight. You want people to be debating these things - may need to educate them on how to do it.

Rob - Once you stop receiving feedback, then you have lost. Your vision will have holes in it so if you lose the direct feedback then you have real problems.

Meri - Give feedback to people when you notice them using the vision language. A quick DM is the most impactful.

Jossie - Make sure you acknowledge the positives even when they are not the most visible - DevOps keeping a system running for a year for example. Don’t just acknowledge the fire fighting.

Rob - It is not what you preach, it is what you tolerate. If people are ignoring the vision then you need to act on it. It quickly spreads. “The standard you walk past is the standard you accept”.

Book recommendations

• Good Strategy, Bad Strategy by Richard Rumelt
• Our iceberg is melting by John Kotter