Before joining Plandek as CPTO, Craig Smiley had already been a customer three times.

Across more than 20 years in technology leadership, he has led digital and agile transformations, built product-led organizations, and managed engineering teams numbering in the thousands. At three different companies, he chose Plandek to understand how those organizations were really performing.

Now, he is helping build the product he once wanted as a customer.

We spoke to Craig about why he kept choosing Plandek, what engineering leaders struggle to see in conventional dashboards, how his experience shaped ProductivityRadar™ – Plandek’s board-ready productivity insights feature – and why AI is making productivity measurement increasingly important.

You’ve spent more than 20 years leading engineering and product organizations. Tell us a little about your background.

I’m a mechanical engineer by background. I like to understand how things work, so systems engineering has always been a big part of how I think. I moved to the UK 30 years ago and started working in technology. Early on, I worked on digital transformations in music and television – things like digital track downloads with Napster and Apple, and video on demand with the BBC and Channel 5.

For the last 15 years or so, my focus has been much more on large-scale corporate transformation. I spent seven years at Worldpay, helping the organization move toward scaled agile ways of working, and eventually became CPO for the UK business. I then led similar transformations at Bottomline and EQ.

A lot of that work has been about moving organizations away from, “We’ve built this widget, now let’s sell it,” toward being genuinely product-led: understanding what customers need, building around that, and creating the conditions for teams to perform well.

Across all of it, I keep coming back to the same things: set good goals, measure what matters, understand the system, and create strong feedback loops.

Before joining Plandek, you were a customer. What first led you to the platform?

We were going through an agile transformation and trying to understand whether it was actually working. When you have hundreds of software engineers, it’s really difficult as a leader to understand what that system is doing. You can have a lot of data, but that doesn’t necessarily mean you have visibility.

Plandek helped us see the flow of work through the engineering organization and understand the patterns behind what was happening. That meant we could move beyond saying, “We’ve implemented agile,” and start asking whether the way we were working was actually improving.

You then chose Plandek again in two subsequent leadership roles. Why did you keep coming back to it?

I really liked the platform and the visualization, but importantly, I also really liked the people. The Plandek team understands productivity and how engineering organizations work. They don’t just give you a tool and leave you to figure it out. They work with you to understand what the data is telling you and how you can use it. It felt much more like an extension of my team than another supplier. So when I moved into new organizations, I knew I wanted that visibility again.

What could you understand with Plandek that you couldn’t get from the dashboards and metrics you already had?

The biggest difference was being able to look across the system and understand where the constraints were. Raw data is difficult. You can look at a collection of numbers and struggle to see what they are actually telling you. Bring that data together and visualize it, and you start to see patterns.

That was the real “a-ha!” moment for me.

Once you can see those patterns, you can ask much better questions. Why is one team performing particularly strongly? What can another team learn from them? Why is another team struggling? Is the problem within that team, or are they waiting for information, dealing with dependencies, missing the right tooling, or being constrained somewhere else in the system? With systems thinking, the place where you see underperformance isn’t necessarily where the underlying problem sits.

That visibility also works at different levels. You can explain what is happening to your executive team or board, while having much more constructive conversations with engineering leaders about where teams need help, where good practices can be shared, and where the organization needs to remove a constraint.

After choosing Plandek three times as a customer, what convinced you to join?

I’d spent the previous 10 or 15 years working at scale. My last organization had more than 1,000 people in the core team and closer to 1,500 including partners and support. I wanted to get back to building something from the ground up. But the timing was also incredibly exciting. AI is changing software engineering very quickly. I could see what was beginning to happen in the market, and some really interesting things happening inside Plandek too.

I wanted to be part of a company doing both sides of that problem: changing how we build software ourselves through agentic engineering, while helping customers understand whether those changes are actually improving productivity. There’s a personal motivation too. I’ve spent a long time in software engineering and product, and I’d like part of my legacy to be helping the next generation of teams have better tools and better ways of understanding how they work. Right now, measuring AI adoption, productivity and token economics is one of the most interesting places you could possibly be.

How did your experience as a customer influence ProductivityRadar™?

For a long time as a customer, I’d been asking for something very simple:

I need a single page I can show to my board.

As an engineering leader, I want somewhere I can quickly understand: How are we doing? Are we improving? Where do we need to focus? And I need to be able to take that view into an executive or board conversation and put our performance into context.

That is what ProductivityRadar™ is designed to do. It gives leaders a single view of engineering productivity, but then lets them move beneath that view and understand what is driving it. Benchmarking is a really important part of that. It’s useful to know whether you are getting better, but leaders, boards and investors also want to understand how performance compares with other organizations. Then you can take it down another level. Which teams are performing strongly? Which teams need help? Where are the constraints? Where could training or better processes make a difference? And finally, you can apply business context. 

Not every engineering inefficiency deserves investment. You might find a constraint, but fixing it may have very little impact on revenue, risk, or strategic priorities. Somewhere else, getting product to market faster might be incredibly important. The value comes from understanding productivity and then deciding where improving it is actually going to matter.

AI is changing software engineering quickly. Does that make understanding productivity more important?

I don’t think productivity itself matters more or less because of AI. What changes is that organizations are making a significant new investment, and they need to understand whether that investment is working. Whether you have a huge agentic transformation program or you’re simply starting to introduce AI tools, productivity in your organization is going to change. If you aren’t measuring it, it becomes very difficult to understand the effect.

AI also introduces another cost into engineering. Organizations are going to have to think carefully about token costs and tooling costs and ask: What return are we getting?

Broadly, that return has to show up in three places. You either grow the product and create more revenue, make the organization more efficient, or reduce risk. If you are investing heavily in AI, you need to understand which of those outcomes you are actually generating.

What do you think agentic engineering will mean for organizations over the next few years?

Initially, I think we’ll see people reclaim some of the time currently spent on more mundane work. Engineers multitask a lot, and there are plenty of repetitive tasks around software development that AI can increasingly take on. I think we’ll see quality improve and product output increase as a result.

But that creates another interesting systems problem. If your engineering organization suddenly becomes capable of producing much more, can the rest of the business absorb that output? You might reach a point where you’re building more product but can’t sell or distribute it quickly enough. The constraint has moved somewhere else.

Organizations will have to start asking much broader questions: How much are we spending on AI? How much are we spending on people? Where is the constraint now? Do we need more product capacity, or is the problem somewhere else in the business? We’ve seen versions of this before with other technological shifts. AI will accelerate it, but the underlying challenge remains the same: understand the whole system rather than optimizing one part in isolation.

If there’s one principle you’d want engineering leaders to take away, what would it be?

Measure what matters, visualize it, and think in systems. Measurement gives you something concrete to work with. Visualization helps you see patterns that are otherwise difficult to find. Systems thinking helps you understand what is actually causing them.

But there’s one other thing I wouldn’t want people to forget. At the end of all this, we’re still people working with people. You can have great data and great tooling, but collaboration, communication, and the conversations you have around that data are what turn insight into improvement.

From customer to CPTO

Craig’s journey from three-time Plandek customer to CPTO gives him an unusual perspective on the problems engineering leaders are trying to solve – and on what they need from the tools designed to help them.

Today, he is applying that experience to Plandek’s product and technology strategy, with the same principles that shaped his work as a customer: measure what matters, make complex systems easier to understand, and use the insight to make better decisions.

Meet more of the people behind Plandek and the work they’re doing to help engineering organizations understand and improve performance.