Roger Allen built houses during the first half of the century in the town where I live. His contracts were likely to read: “I, Roger Allen, do hereby agree to build a house for Guy and Ruth Emerson on their land in Chilmark to consist of 3 bedrooms and 2 bathrooms for the sum of $4,500 and to complete this home and leave the premises broom clean by June 15, 1935.”

Roger probably greeted the people when they  arrived at their new home. Was everything as they had expected it would be? Was it “broom clean”? I have a feeling broom clean meant a lot. I think it meant trim and functional, well-fitted and just right. If something wasn’t just right, Roger probably discussed it with the clients over coffee and then took care of it. 

Roger started building in 1910, but by the time of his death in 1967, he had long since stopped building new houses. Maintaining, repairing and adding to the houses he’d built was more than enough to keep him and his crew busy. Roger understood that the dynamic process of designing and building houses doesn’t end at their occupancy.

Like a lot of architects and builders, my colleagues and I used to bang into a major trust barrier around the time houses were completed. After working comfortably through a long design and construction process, relationships with clients began to deteriorate as the project neared completion. The difficulties came from minor matters, like a billing error or the location of an outdoor faucet. I’d think: This has been a good collaboration; these annoyances shouldn’t be so important. But they were. Why?

Time and money are obvious problems. But diminishing trust has shown up even in projects with neither schedule nor financial difficulties.

Having a house built is a major emotional undertaking, and as the project winds down, many clients feel unfulfilled expectations, questions, uncertainty. And when little things don’t work, it aggravates the discomfort.

“A strong, collaborative relationship is hard to maintain near the end of a project when the pressure is on and our attention has strayed.”

I don’t know about you, but in our houses lots of things don’t work. Doors stick, and fixtures are crooked. Everything needs tuning up. Doing a good job with this final phase of a project brings great rewards. Doing it poorly has grave implications. If relationships crumble, your clients may remember only these difficult moments and forget all the positive aspects of the project. Whether directly or through referrals, our clients are the source of our business—that’s all we know about marketing—so it’s critical that we maintain their goodwill. And if we don’t stay on good terms with clients, we also cut ourselves off from our best sources of information: the houses we build.

A strong, collaborative relationship is hard to maintain near the end of a project when the pressure is on and our attention has strayed. The clients feel abandoned and figure every tradesperson in the hemisphere should be at their house. In recent years we’ve adopted specific strategies to ease this passage.

One strategy: To the client goes the benefit of the doubt. It’s a waste of time figuring out who’s at fault. A client once came to see his house as construction was ending. A large and important window didn’t feel just right. It was the window shown on the drawings, but in real life the size of the panes was wrong in relation to the other windows. It was subtle, but it was so. I said, “Maybe you ought to live with this one a while, and see if you get used to it as it is.” I’ll never forget his reply: “I can get used to it. I can get used to almost anything. I can get used to a hatchet in my forehead, but do I want to?” Right. We don’t want people to get used to what’s wrong in our buildings. We want to make them right. We replaced the window.

“Buildings need post-occupancy care. . . .Nurturing is required.”

Our most important strategy is simple: We don’t let go at the end of a project. We’re there whenever the clients need us. We have a post-occupancy budget in our original pricing, and one person’s job consists entirely of punch lists, minor repairs, maintenance and small improvements. The clients’ needs gradually disappear, and the problem of eroding relationships has dramatically declined.

The job I speak of is also a way to keep track of successes and failures, and it’s a mechanism for improving our designs and practices. When we go to a house to repair something, we inspect the whole house. We look for other problems, fix them and report back. As a result, we abandon the cheap weatherstripping that gets replaced yearly in favor of interlocking bronze that lasts for decades. We notice where compact fluorescent lighting is acceptable and where it’s not. We discover that our meandering flagstone path doesn’t go where people are most comfortable walking. Our understanding grows, and our buildings get better.

Buildings need post-occupancy care. They don’t automatically work right. Post-occupancy is like childhood. Nurturing is required. But often the design/build team is nowhere around. Stewart Brand, in his book How Buildings Learn says, “I recall asking one architect what he had learned from his earlier buildings. ‘Oh, you never go back!’ he exclaimed. ‘It’s too discouraging.'” Can you imagine if surgeons never did follow-up visits with their patients?

When our clients move into their new homes, a welcoming letter invites them to call us frequently and promises that in a year’s time, we’ll schedule a walk-through with them. We examine unresolved problems and discuss how the house works and how they’re feeling now that they’ve settled in. This is a service to them, but it’s a resource for us.

Recently, we surveyed past clients to find out what has and hasn’t worked well in their houses over time. We got detailed, articulate responses. To the question “What should no new house be cursed with?” one client, for whom we had built a house in 1980 with a dyed-concrete slab that hadn’t set just right, answered, “A floor that rubs off on the bottom of your feet.” Past clients are a great store of information. Blunt, too.