Project XD

McDermott runs dozens of large engineering-and-construction projects concurrently - offshore platforms, subsea pipelines, onshore terminals - each staffed by hundreds of engineers and managed from regional offices. Before XD, project tracking lived in Excel: status reports emailed weekly, rolled up by hand, broken into regional sheets that diverged from each other by the time leadership reviewed them on Mondays.

Three very different people relied on that weekly ritual. Executives needed to see portfolio health at a glance - which projects were on track, which were slipping, which needed intervention. Engineers needed to report status without it eating into the engineering work they were actually paid to do. Field managers sat between those two, translating between the engineers' granular reality and the executives' portfolio view.

My brief from the PM was clear: replace Excel, give executives a dashboard, speed up the reporting cycle. The brief didn't mention engineers' workflow - and that turned out to matter.

Fig. 01 · The paper-trail reality XD replaced.

XD had three design decisions where a less experienced designer might have gone a different way. Two I made cleanly; one I got pushback on and ultimately made the harder call.

The clearest evidence of Decision 01 is what executives actually saw. Before XD, they saw a spreadsheet. After XD, they saw a rotating globe with portfolio health at a glance.

Before · spreadsheet

Excel rollup

Weekly Excel rollup assembled by hand, sent as an attachment to a Monday email, frequently divergent between regions.

After · XD globe

Real-time globe visualization with projects sized by scale, colored by status. Executives see portfolio health in one view.

For the first six weeks of XD, I treated the project as a business problem. The brief was "replace Excel, speed up the reporting cycle." I anchored my discovery on executives and managers - the people whose pain was loudest and most visible.

Engineers got included late. When I finally ran engineer interviews - partly because the template design argument forced me to - I realized I'd misunderstood the product's core constraint. The bottleneck wasn't executive reporting; it was the fact that engineers were already overwhelmed, and a reporting tool that added friction would be sabotaged at the source. No executive dashboard works if engineers don't file.

The product recovered because I course-corrected - involving engineers in template design, weighting their feedback heavily, re-sequencing the build so their experience mattered as much as the executive view. But the six weeks I lost early are the clearest "what I'd do differently" on this project.