I’m delighted to welcome Jeffrey Nolte, Founder of Nolte, and Yanna Lopes, Head of Products, to today’s conversation. Nolte is a software delivery firm that has been building products since 2006 — nearly 20 years of hard-won delivery experience.
We discussed something that’s genuinely rare: they stopped estimating. Not as a philosophy experiment — as a business decision. Today they guarantee outcomes. Fixed scope, fixed timeline, fixed price. And they have three years of delivery data to back it up.
I wanted to understand how that actually happens. What it takes to get there, what breaks along the way, and what it looks like on the other side.
The Question Every Client Asks and Why It’s So Hard to Answer
Every agency knows the moment. A client asks: how long will this take, and how much will it cost? And the answer, if you’re being honest, is that you don’t really know.
Nolte spent years trying to get better at answering it. They tried story points, T-shirt sizing, planning poker, just to name a few. “We found that none of it ever worked,” Jeffrey said. “It would put a lot of strain on our relationships with clients.”
The problem isn’t the method. It’s that estimation is still a guess. No matter how rigorous the process, you’re projecting a future you haven’t lived yet. And clients feel that uncertainty even when you dress it up in confidence.
Around 2021, Jeffrey started looking for something different. He wanted to stop guessing and start measuring.
What Changes When You Measure Instead of Estimate
Flow metrics cycle time, throughput, WIP aging record the past. How long does work actually take to move through your system? How many items does your team consistently complete per week? Once you have enough of that history and put the effort into optimizing your delivery process, the future starts to look a lot more predictable.
Yanna has owned the data model since she joined four years ago. Her starting point was the Nave aging chart — using it to track work in progress and monitor throughput week over week.
Early on, consistency was the challenge. They’d deliver 20 outcomes one week and five the next. “What is the magic we need to implement here?” Yanna asked. “And it’s not magic — it’s data analysis.”
Getting to a reliable pattern took time. It required clear definitions of what ‘done’ looks like, a process the team actually followed, and the discipline to look at the numbers even when they were uncomfortable.
“Do not be afraid of data,” Yanna said. “Sometimes we are afraid of looking at it. But you need to understand it. See what is going on, see why it’s happening.”
That’s the unglamorous part of predictable delivery. Not the tool, not the framework — the willingness to keep looking at reality and making small steps to improve, every single day.
From Reliable Data to a Client-Facing Promise
Once you have a consistent delivery process, you stop saying ‘we think this will take around eight weeks’ and start saying ‘based on how we’ve delivered the last 75 projects, here’s what we can commit to.’ That’s a fundamentally different conversation.
Nolte now runs at 75 guaranteed deliveries per month, with a 95% forecast accuracy. May 1st is their cutover date — the point at which hourly billing disappears entirely and every engagement moves to delivery-based pricing. What finally pushed Jeffrey over the line was a question from his head of engineering: “How do I log time for this when I have four AI agents running in parallel that are building code?”
There’s no good answer to that question inside an hourly model. And Jeffrey knew it. Agentic AI doesn’t produce trackable hours — it produces outcomes. The model had to follow.
The Internal Work That Makes It Possible
Predictable delivery isn’t a dashboard you turn on. Behind Nolte’s numbers is a whole layer of process work that most teams never get to.
Yanna built NolteOS — an internal tool that pulls from Nave, Jira, and Nolte’s own process data and surfaces it in a format clients can actually read. Deliveries this week, what’s remaining, what’s blocked, what the agreed priorities are.
But the more important shift is what the tool enables in the client relationship. When a client asks to change button colors, the conversation becomes: how does this delivery connect to your actual goal? “By focusing on delivery value attribution, that changes the game completely,” Yanna said. “Now we are having a more productive and intelligent conversation.”
They’ve also built a public-facing scoping tool at os.nolte.io. Describe a product idea or feature in plain language and it generates a delivery plan — scope, timeline, cost — drawn from Nolte’s historical data. Jeffrey demonstrated it live: a single prompt about a PayPal integration brought back 25 deliveries across strategy, discovery, UI, webhooks, compliance, and documentation. Complexity a client wouldn’t see coming, made visible before the first call.
What This Takes — and What It Makes Possible
What Nolte has built is defensible in a way that hourly billing never is. “I feel really confident we’re on the forefront of that,” Jeffrey said — and after seeing the delivery record, the tooling, and the client relationships they’ve maintained, it’s hard to argue otherwise.
The model works because it’s grounded. Not in confidence or methodology, but in a record of what the team actually delivers — measured consistently, over time, until the pattern becomes something you can stake a promise on.
If you’d like to connect with Jeff, follow him on LinkedIn or explore the scoping tool at os.nolte.io
I wish you a productive day, and I’ll see you next week, same time and place, for more managerial insights. Bye for now
