I love to read about other people's design process, unless it involves a graduate-school vocabulary or it smells like a beret and Gitanes are involved.

For me, designing things is a mostly mechanical process. Sure, there is inspiration involved in the process. But that part is easy if you have done your homework. (And there is a great deal of homework.)

As Pablo Picasso put it: "Inspiration exists, but is has to find you working."

This column will lay out the typical steps I take to design a piece of furniture. I'm now in the final stages of designing a simple sitting stool for an upcoming book, so I'm going to use that stool to illustrate my process. Though this stool has only eight pieces to it, my method is the same when I design a commission that takes six months to build and involves 100 parts or more.

If there is a fire in my workshop I will try to save several of my books, including these volumes on Kaare Klint, the father of Danish modern furniture.

Build a Library

I have two libraries at my workshop, a physical one made up of books I've collected for the last 30 years and a virtual one that is images I've cribbed from websites, plus photos I have taken while at museums all over the world. I am constantly organizing both libraries. I arrange the books into meaningful categories such as "European vernacular stuff," "Chair books I love," "Victorian-era catalogs" and straight-up Danish Modern Furniture Porn.

Likewise, my Dropbox is stuffed with terabytes of images. And I cram the objects' dimensions and other descriptive information into the metadata of the photos.

This sounds boring, but acquisition and organization of images is my No. 1 job as a designer. After I visit a museum, I dump my images into Dropbox and then organize them into meaningful categories. It's a habit, and it's something I can do at the dentist's office or while I wait for my family to get ready to go out to dinner.

These big image libraries help create an overall vibe of the things I love to look at. Exactly what I love has changed through the years, but that's useful information as well. I can see where I've been and where I'm headed.

Then, when I have a specific design task at hand – the stool! – I do two things. First, I see if I've collected any stool samples (I promise that is the only stool joke in this column) in my libraries. I start scanning books and copying images into a temporary folder of ideas I like. Then I do some searching on the internet to see if I can turn up any fresh chum. I've been on a bit of an Irish kick lately so I might turn over a few rocks at the online vendors of Irish antiquities.

I haven't built any chairs that were inspired by vernacular designs from colonial Canada, but I have a folder of beautiful (and odd) examples for when that day comes.

Oh, I forgot to mention something: Even if I'm designing something contemporary, I start with historical images. I want to know what's been done and what hasn't. Ignore the past at your peril.

This is not a brief process. I try to look at everything I can get my hands on. The idea is to look for items I can build upon. Or tropes I can reject.

Know the numbers before you draw. There are published guidelines for how to make a chair comfortable or a cabinet useful. Learn them so you can disobey them when the time comes.

The Mechanical Stuff

Figuring out the gross dimensions for the object is the easy part. Those limits are imposed by the human body, the customer or the book "Human Dimension & Interior Space" (my preferred reference book). These dimensions create an imaginary three-dimensional box to work in.

Even if I've designed 100 stools (and I haven't, yet), I'll reread the sections on popliteal height, seat depth and anything else that applies to stools in the reference literature. When I skip this step I regret it.

I am not trying to win my town's art contest and get my piece displayed at the shopping mall. I'm trying to capture an idea. This is not pretty, and I don't strive to make it so. These sketches aren't for the client, they're for me.

The 45-second Sketches

After all that technical and organizational drudgery, I get my sketchbook and a pencil. Some people work in pen because you cannot erase. I love to erase. Then I start making sketches and spend no more than 45 seconds on each sketch (including any erasing time). This is a lesson I learned from my youngest daughter, an artist.

After the first sketch I force myself to try other iterations. It's too easy to fall in love with the first drawing and stop sketching. I force myself to add stretchers to the stool. Subtract them. Make the seat comically thick. Add legs and shape them differently. Try different seat shapes. Plus including anything I can remember from my library that I liked (for example, steeply beveled edges to the seat).

I shoot for seven sketches because I love alliteration.

After I get my seven sketches, I'll redraw the two or three versions that I like best. This is mostly about correcting the overall proportions and refining the details.

CAD or Prototype?

The next step depends on what I'm building. Some furniture makers would be horrified by what I'm about to say, but basically we make boxes and platforms – not much else. If I'm building a box (a cabinet, bookcase, chest or the like) it is usually fairly rectilinear. So I'll scan my sketches and import them into CAD. Then I draw over them.

Quick note on this process: I don't try to convert my sketch into whole-number units, though it is tempting. Doing so seems to suck the life right out of the design. A number such as 4-3/16" is just as valid and achievable as 4". Many times it's best to obey the pencil.

This rectilinear piece required seven pencil sketches (at least). Then I scanned it into CAD and produced a bunch more to pick from. There were at least three that were decent enough to build. I picked the riskier one with asymmetrical drawers.

From here I can usually start construction of the piece. If I lack confidence, I'll build a full-scale prototype using pink or blue insulating foam and clear packing tape from the home center. Then I draw on the door fronts and ornamentation with a Sharpie.

The other furniture form is a platform (tables, chairs, stools, settles, beds). While some of these can be rectilinear and easily drawn in CAD and easily visualized, most are more sculptural and need to be viewed from more angles than the typical plan, elevation and profile before I start building.

That's when I sneak into the closet and snitch some wire hangers. I snip these hangers to become the legs of a half-scale model. The platform (seat, tabletop etc.) is a piece of wood cut to shape. To install the legs I drill a slightly undersized hole in the wood and epoxy in the legs. Then I dab some extra epoxy around the hanger to add some strength.

I keep the models as a reference. I mark all the construction information on the model and clip the legs so I can reuse them.

After the epoxy cures it's playtime. I get my needle-nose pliers out and start bending the legs until the model looks like my sketch. I can even snip the back legs to tilt the seat. To add stretchers to a table or chair I'll use pipe cleaners or even blue tape.

If I'm making a chair, I might add the stretchers and arms above the set as well.

This is much faster than CAD (and I'm pretty good at 3D modeling). And, this is the important part, I have never been surprised by the look of the final piece when I have built a model. I cannot say that about piece I have built right from a CAD or mechanical drawing.

Having the half-scale model also helps me calculate with ease the drilling angles during construction.

This stool was made using pieces I pulled from the garbage and the scrap bin in our machine room. It's still salable – after a coat of paint.

Full-blown Prototype or No?

With a CAD drawing or half-scale model in hand I decide if I'm going to build a prototype before cutting into the good wood. When I make prototypes I use inexpensive woods, such as poplar or red oak. If the joinery needs special jigs or is something I haven't cut before, this is my chance to do it when the stakes are low.

If, however, the joinery is straightforward and just time-consuming (mitered through-dovetails, for example), I'll use the Festool Domino to join most of the pieces. In some cases I might use pocket screws. The goal is to get a piece of furniture that is good enough to paint and then sell at a discount. If the prototype takes an ugly turn, however, I'll disassemble it and use its parts for other prototypes. (I eat dinner every night while sitting on a chair prototype.)

I prefer to make a prototype because the real project always turns out much nicer – I make most of the missteps in the prototype phase.

This three-legged backstool was a failure. But after drawing out some changes over a photo of the failed piece, I ended up with a piece I was thrilled with. And I didn't have to go back to the beginning.

What if the Prototype Sucks?

There have been a few instances where the prototype was a full or partial failure. The piece simply did not look right when it was brought into the world in full size and with real joinery. When this happens, I kick myself a bit. I then take photographs of the failed prototype against a plain background. Then I print out those pictures and start drawing in the changes that I think will improve the piece.

This last-ditch strategy almost always works.

The Finish Samples

Lastly, if the customer has requested a surface finish that I haven't used before, making a sample board is critical. Winging it almost always ends in stripping it.

Sample boards should be made using wood that is ideally from the same tree (or at least the same species) as you are using for the finished piece. The sample board should be planed, scraped or sanded just like you plan to do for the finished piece.

And after you get a finish that both you and your customer like, cut the sample board in half and give one half to the customer. This tip, which I received many years ago, has saved me a couple times.

Helpful Next Steps

There are lots of different ways to design furniture. My company, Lost Art Press, has published two of them. Apologies for the advertisement, but we wouldn't have spent years working on these books if they weren't helpful. "By Hand & Eye" and "By Hound & Eye" by George Walker and Jim Tolpin explore design using whole-number ratios and historical methods. I think whole-number ratios are indeed pleasing; I simply find them in my work after the fact instead of installing them in there intentionally.

Also good: "The Intelligent Hand" by David Savage. David, a trained artist, comes at the process from a different direction that involves an imaginary friend, George. You'll love George.

If you want to start building a visual library, here are a handful of my favorite sources for images. When I am despondent about the ugliness of the built world, these online places cheer me up.

The Rijksmuseum, the museum of the Netherlands. It has a huge digital library that is open to everyone.

Colonial Williamsburg's online collection.

Victoria & Albert Museum's online collection.

Bukowski's Online Auction – great for European stuff.

1stdibs, a sprawling worldwide auction site of decorative objects.

Furnitureindex.dk takes you on a deep dive of Danish and mid-century stuff.

Christopher Schwarz is the editor at Lost Art Press and one of the founders of Crucible Tool. He works from a restored 1896 German barroom in Covington, Ky. You can see his furniture at christophermschwarz.com.