DNV Digital Solutions
Leading a portfolio of software businesses to engineer the energy transition.
I came to DNV from the other side of the screen. At Frazer-Nash I had been a Bladed customer, and in 2018 I joined DNV to sell it, first in one region and then leading its business worldwide.
In 2021 DNV Digital Solutions decided that the way it ran software had to change fundamentally, towards the way the best product companies run theirs. The business area’s chief executive introduced product leaders who each owned a P&L, and advertised the roles. I applied. I was the youngest candidate and had the least management experience, but I had the technical depth, I was the one selling and facing customers, and as Bladed’s business lead I had grown its revenue by 90% in a year, mostly through upsells to existing accounts in China. I was also the main author of the offshore wind business case, combining Bladed and Sesam, that had just been approved. I was already doing the job; the role gave me the mandate.
Onboarding was two books, Marty Cagan’s Inspired and John Doerr’s Measure What Matters. I read both several times, took the transformation as the job to be done, and kept reading: the SVPG list, Melissa Perri’s Escape the Build Trap and Patrick Lencioni’s The Five Dysfunctions of a Team.
The portfolio was specialist engineering software, most of it with decades of history. Bladed is used to design wind turbines and calculate the loads they are certified against. Sesam analyses the strength and motions of offshore structures, fixed and floating. Alongside them sat energy yield tools for wind and solar farms, and rule-check tools for ship hulls and machinery.
The people who use them have one job that matters: to design something that must not fail, and put their name to it. A turbine, a jacket, a floating platform, a hull. The software’s job is to let them do that faster without being any less right, because when an engineer signs off a design against the output, credibility is the product. Being wrong is not a bug report; it is a business-ending event for whoever put their name to it. DNV has been in the business of that kind of trust for 162 years, and it publishes that 81% of offshore wind projects entering operation in 2025 involved a Sesam customer. Bladed, when I ran it, held around 75% of the wind turbine design market.
What I inherited was organised around technology rather than customers. Sales had no leader. Two engineering departments had experienced heads, both of whom had wanted the role I was given. Technical support sat inside software engineering, and its spare capacity went into building product. Product management existed in name, but decisions about what to build were made by senior stakeholders rather than from customer evidence, and customers reached the developers second-hand, through sales or through those stakeholders. Engineering was organised by technology, so the software carried a fragmented architecture and duplication across the stack, and work was planned once a year to keep everyone busy across every use case. The team was 120 people across 20 countries.
Becoming customer focused, and getting the best out of the value streams we had
Getting the best out of what we already had meant turning the whole organisation round to face the customer: product management in practice rather than in name, then design and product marketing, a direct line to customers, and teams shaped around the work customers actually do. It happened one step at a time, and mostly by agreement, because most of the teams involved did not report to me.
Two businesses, one product line
We merged the renewables software and the offshore structures software into one business, Renewables and Ocean Structures. The merger had always been possible; it needed someone with a reason to do it, and that fell to me. I kept both heads of department: one led the merged engineering department, and the other moved into my team as product manager for two growth products.
Product management first
I started with product management, advertising five roles organised by market segment: one for Bladed, three for Sesam, and one for the two energy yield products. The rule-check tools stayed in their existing model on purpose, because there the product was, in effect, the standards and recommended practice, built as they were published. Design and product marketing followed.
Small, fast steps towards outcomes
We experimented with working in small, fast iterations towards outcomes instead of big releases. The clearest case was the floating wind work: planned as one release at fifteen months, it put a first alpha in customers’ hands in three. We put visuals, alphas and betas in front of customers early, let scope follow their evidence rather than stakeholder assumptions, and counted success in what customers did, not in what we released.
A channel everyone used
The direct line to customers started small. I asked for every UX research meeting to be summarised in a Customer Insight channel in Teams, and product, engineering and sales started reading and responding. Then I found change leaders in sales and customer success to post the customer needs they were hearing. It built a whole-team feeling, everyone could see the benefit, and more of sales and customer success joined in. It grew to more than 300 research conversations a year, and became the most used channel we had.
One major bet per product
Each product rallied around one bet instead of spreading itself across every use case. For Sesam, the bet was to win floating offshore wind by making structural fatigue and code checks ten times faster, and later a hundred times, which produced new unit load and unit structural response methods: far less reading and writing, computation and storage, with enough fidelity for concept design and the approach to detailed design. For Bladed, it was a hard push to Bladed 5, re-architecting the data model so calculations are easy to automate and scale, and customers can run their own models inside Bladed runs.
The machinery rule-check tool’s bet was to keep doing what worked. The hull rule-check tool’s was to become part of the classification value chain rather than a software business of its own. The energy yield products’ was to go API first, with API usage as the one metric that mattered.
From an annual plan to a four-monthly rhythm
Planning moved from once a year, to quarterly, to outcome objectives reset every four months. Progress was shared openly across the whole organisation every three weeks, and a weekly tactical session became the place where problems got solved rather than reported.
Teams shaped around the customer’s experience
Finally we restructured the teams to match how customers experience the service and the product, combining the value streams of Sooner Safer Happier with the team shapes of Team Topologies. Development teams were grouped around the stages of the customer’s engineering workflow, with a cross-functional team above them owning the workflow as a whole, a shape a team a layer below me worked out and ran first. Engagement held at 7.5 out of 10, on over 90% participation, through the restructuring.
Managing the portfolio for profitability, and to make room for innovation
Owning the whole P&L changed the question from how to improve each product to which products belonged in the portfolio at all. We answered it with a vision and mission for the portfolio: “Enabling the world’s engineers to accelerate the energy transition and sustainable use of the ocean.” With that in place it was obvious which products fitted and which did not, and together with how each was doing against its goals, portfolio decisions became straightforward. Some products were software businesses in their own right. Others were really there to serve another part of DNV, or had not found their market, and moving those to where they could do better freed money and attention for the bets that mattered.
The hull rule-check tool, back where its purpose lived
The hull rule-check tool existed to support DNV’s ship classification business rather than to stand as a software business on its own. So it moved into the classification business, where that purpose lived.
Two products, a pivot and a new home
The energy yield products’ API-first bet did not land in time: they kept a desktop focus and lost their first-mover edge. I initiated a transfer into a larger data business elsewhere in DNV, whose go-to-market could take them to customers we could not reach. We developed the pivot strategy together across the two business areas, and the receiving business took on the cost of executing it.
I ran the teams through the handover for months, keeping customers served and the technology running while the API-first direction got going, and coached the receiving business’s leader in a weekly one-to-one.
Investing where products belonged together
Where products belonged together, we invested. I won £2.5M a year at board level to make the Bladed solver drivable from Sesam, so offshore wind customers can simulate turbine and structure together, and the organisation was restructured to mirror the product. Bladed 5 shipped publicly in early 2026.
Moving customers to subscription
We moved customers from perpetual licences towards subscription in four steps. First, a bias towards leasing, with some added friction for perpetual. Then a subscription offer that customers were drawn to rather than pushed into. Then iterating until we found the subscription recipe that worked, and only then introducing the sticks.
Along the way a three-year £2M renewal with a flagship account closed at a 15% discount, against a historic 35%, and rebuilt commercial credibility there.
What it added up to
Over those years the business went from £20M of revenue at break-even to £30M at 15% EBIT, 21% before restructuring costs, against a plan of 10%. 35% of revenue came from China.
Offshore wind grew over the period, and some of the revenue line belongs to the market. The margin and the shift to recurring revenue came from how the business worked, not from the market.
- Revenue
- £20M to £30M
- EBIT
- Break-even to 15%21% before restructuring costs
What I have not done
The trust I stewarded was not trust I created: DNV’s name was 162 years old when I arrived, and I have not built technical credibility from zero with a first customer. I know what earns it with an engineering buyer and what destroys it, because I worked inside the institution whose business is that credibility. What I built rather than inherited was a product management function, a leadership team that worked as one, and a customer channel everyone used.