Joel Koppelman has spent much of his working life attending to things that are unfinished. First came transportation projects, with their engineering decisions and financial consequences. Then came Primavera, the software company he co-founded in 1983 to help people manage projects. Later, after Oracle, came a fuller commitment to making art. A carving has a useful advantage over a corporate status report: if it is unfinished, there is generally nowhere to hide the evidence.
The neat version of his biography would end with the acquisition. Build a company, sell it, take the congratulatory photograph. Koppelman’s story keeps going. The retired executive is also a sculptor and painter in Santa Monica, with an earlier life in Philadelphia and a continuing interest in institutions that teach people to make things. There is room in that life for a spreadsheet and a studio. Neither need apologize for the other.
Before the screen, the railway
His qualifications joined two ways of looking at a project: civil engineering at Drexel, where he graduated in 1970, and finance at Wharton, where he earned his MBA in 1976. One concerns what can be built. The other asks what it will cost. Anyone who has attended a construction meeting will appreciate the convenience of being able to follow both conversations.
Before Primavera, Koppelman worked at Booz Allen Hamilton and Day & Zimmermann. At the latter, he became a vice president responsible for transportation capital projects. His assignments included establishing initial management controls for the design and construction of an 18-mile Los Angeles Metro Rail project, and evaluating alternatives for a $300 million replacement of Philadelphia’s Frankford Elevated line. Planning, in this setting, meant decisions attached to actual places, actual structures, and substantial commitments.
That background gives the software career its footing. Koppelman arrived with experience of the work a planning system was supposed to serve. Alongside his fellow founders, including Dick Faris and Les Seskin, he established Primavera in Pennsylvania in 1983. The desktop computer offered a new place to do work that had previously depended on larger computing systems.
Primavera Project Planner, known as P3, went on sale that year. Its early delivery medium was the 5¼-inch floppy disk. The modest object carried an ambitious proposition: put project planning within reach of the person working at a desk. The office computer could become a place to explore consequences before a change made its expensive journey into the world.
A useful suspicion of tidy numbers
For a software executive, Koppelman could be refreshingly interested in everything surrounding the software. In 1998, he spoke about the changing relationship between people and computers. Earlier systems obliged users to organize themselves around the machine. Better technology should allow the machine to accommodate the person. He also pointed toward communication and the ability to understand information, rather than merely operate a program.
“Ease of understanding.”
Joel Koppelman, on what project software should offer users, 1998
The phrase is small enough to overlook and demanding enough to organize a product around. A person can learn which button produces a report without knowing what the report means. A manager can possess an immaculate screen full of numbers and still misunderstand the work. Koppelman’s distinction asks the software maker to care about the moment after the button is pressed.
His authorship belongs in the same conversation. With Quentin W. Fleming, he wrote Earned Value Project Management. Their 2001 article on a simpler approach to earned value argued for a practical subset of a larger system. The attraction was to make the method usable beyond organizations prepared to carry an elaborate administrative apparatus.
The underlying question is wonderfully rude: how much work has actually been accomplished for the money spent? Expenditure alone cannot answer it. A project may have consumed half its budget without completing half its work. Earned value puts planned work, completed work, and actual cost into a shared frame, making differences harder to wave away.
$90k spent to produce work budgeted at $60k. A cost gap of $30k deserves a conversation.
By 2006, his comments about portfolio management were equally direct. He questioned how business requests were justified and how return-on-investment estimates were constructed. If the inputs were guesswork, sophisticated tools would still be working with guesswork. The practical concern was accountability: someone eventually needed to compare the promised result with what happened.
The bicycle horn in the software company
Primavera had its own opportunity to discover how uncomfortable that comparison could be. In 2002, a software release arrived three weeks late and was considered incomplete by management. Developers had sacrificed weekends to finish it. Selling planning software had conferred no exemption from the difficulties of planning software development.
Development leader Bob Schatz investigated a different approach. With support from Koppelman and Faris, Primavera brought in Scrum and Ken Schwaber. The work moved toward monthly increments, with demonstrations of functioning software. Later, the company added Extreme Programming practices to support engineering quality.
One detail deserves to survive every management presentation made about the episode. At reviews, a bicycle horn told visitors when to move to the next development team. Executives and colleagues encountered the work, asked questions, and moved on. The sound is less dignified than a strategic transformation announcement, and considerably easier to remember.
Koppelman’s place in this episode was as a supporting executive, alongside people who did the investigating and implementation. That distinction matters. A founder’s biography should leave room for colleagues. Here, the change depended on leadership being willing to let the organization work differently, including making progress visible while there was still an opportunity to alter what came next.
A customer brings his own golf clubs
The company’s product evolution brought another kind of negotiation. In 2002, Primavera was shipping P3e/c, with a more substantial database architecture and more flexibility in how users could organize information. Koppelman described customers wanting both an overview and the ability to examine detail. “Owners want a real database,” he said.
The new system could subdivide project elements through many more levels than P3. It also offered templates and ways to reuse material from previous projects. Existing customers were offered the product free, with the ability to run old and new versions concurrently. Even under those conditions, persuading people to change remained a task of its own.
Years later, a lunch with construction scheduling specialist Fred Plotnick supplied a useful illustration. They discussed construction users’ attachment to P3 and SureTrak, and Koppelman questioned the case for substantial development to make P6 behave more like P3. Plotnick compared those users to left-handed golfers: they needed clubs fitted to their requirements.
The exchange captures a difficulty familiar to anyone who has built a product for specialists. A feature can look marginal when counted across the whole market and essential from one particular desk. The customer’s stubbornness may contain a specification. Equally, a software business has to decide which demands justify the work. Lunch does not resolve the tension; it makes it wonderfully concrete.
The acquisition was a chapter, not an ending
By the middle of the 2000s, Primavera had grown beyond its original setting. In December 2006, the business had recorded 22 consecutive years of revenue growth and profitability. Its reported revenue for 2007 was $176 million. The numbers establish the scale of the organization Koppelman was helping to lead; they also imply a great many people whose efforts had to meet.
In December 2007, Primavera and Autodesk announced an agreement to support interoperability between their architecture, engineering, and construction software. The link with Navisworks addressed coordination across design and project management. A schedule gains usefulness when it can participate in the rest of the work, rather than sit apart looking handsome.
Oracle announced its agreement to acquire Primavera on October 8, 2008. The proposed combination connected project portfolio management with Oracle’s wider applications and infrastructure. Koppelman was to lead the business unit, becoming senior vice president and general manager. The founder went with the company into its next organizational life.
In 2011, he retired. The years after a sale can disappear from a business profile as though the person has exited along with the shares. His took a different direction.
A different kind of unfinished work
Art had already involved years of learning at the Pennsylvania Academy of the Fine Arts. In 2016, he moved from Philadelphia to Santa Monica. His work includes stone carving, terra cotta, and painting; his galleries offer another way to look at someone usually introduced through a software company.
Look at the two works here. A pale carved face sits among leaves, exposed to the untidiness of a garden. In the painting, buildings and structural elements occupy the foreground beneath a crane. They are images made by someone whose professional history also includes construction. The visual connection is there for the viewer to enjoy without turning every brushstroke into a business lesson.
His institutional connections also continued. PAFA announced his return as a national trustee in July 2019, describing him as an abstract painter and former continuing education student. He wanted to strengthen its West Coast connections with artists, prospective students, collectors, and contributors. The earlier student now had another way to help keep the classroom connected to the world.
Drexel had recognized his work too, inducting him into the Drexel 100 in 2000. He supported a virtual collaboration laboratory there and served on the university’s board. Across engineering, software, and art, teaching institutions recur in his biography. They make a pleasing complication of the accomplished executive: a person can have considerable experience and still have reason to show up for a lesson.
There is no need to force the engineer and the artist into a perfect likeness. The pleasure is that one life contains both. Koppelman helped build tools for work with budgets, dependencies, and delivery dates. He also gave more of his time to objects that ask to be looked at. The company belongs in the story. So does the workbench.