MARCH

DATE: THURSDAY, MARCH 1, 2012

BANK BALANCE: $145,855.88

CASH RELATIVE TO START OF YEAR (“NET CASH”): $8,701.56

NEW-CONTRACT VALUE, YEAR-TO-DATE: $407,271

New orders arrived at a nice pace in February: twelve jobs worth $213,669. Nick continued as best salesman, closing six deals worth $150,104. Dan did four orders worth $37,768. I sold two, adding $25,797 to the total. Dan still trails Nick by a wide margin, but the total for the year exceeds my target. If current trends continue, I’ll need to figure out what to do about Dan, as he’s not producing at a rate that justifies his salary. The conventional fix for a slumping salesman is to pay him only with commission or get rid of him. Nobody would be surprised if I fired Dan—that’s what bosses are supposed to do—but I just don’t have the stomach for it. I can picture how unpleasant it would be to announce a decision that will probably destroy Dan’s life. He’s new to the area, with an unemployed wife and four small kids. He needs a stable paycheck. It’s in my power to provide that or take it away. Nick is carrying the load for both of them, so I just let the whole thing slide.

Our cash position is solid. In the past four weeks, we took in $232,475. Some of this included payments for jobs we did for the federal government in 2011. The government pays the whole contract amount thirty days after delivery, and we had shipped a couple of jobs to the Air Force in December. The rest is incoming deposits. We spent a good deal of money in that same period: $185,782, or $9,289 a day. More than I wanted, but the overall picture is good. We’re ahead for the year.

Since we have cash, I decide it’s time for the company to start making interest payments on the money it owes me. That number is substantial: $387,098. Most of that dates back to my years with The Partner and before. I would make loans to the company whenever the business ran out of money, once or twice a year. Incoming payments have always been unpredictable, but rent and payroll arrive like clockwork. I squirreled away as much of my pay as I could and usually had between ten thousand and twenty-five thousand dollars in a personal account, separate from the money I used for family expenses. I had no other financing options. Before The Partner, my books were in such disarray that no bank would lend me money. I remember well the humiliating day in 1994 when I dressed up in my best suit, sat down with the manager of my bank, and was told that I was far too risky for a loan. So I financed the company out of my own pocket. During the sixteen years I was sole proprietor, I loaned the company $167,650 and was paid back none of it.

The Partner brought some change. His wife sorted out my books, but she died before she could teach me to manage cash. Bookkeeping, which records what has happened to a company’s money, and cash management, which looks forward to figure out what is about to happen, are very different. Bookkeeping and accounting are standardized. Cash management is not. Amazingly, it’s up to each company to invent its own methods of predicting cash flows and planning expenditures. The Partner had no idea how to do this. He left it up to me, and my method was to watch my bank balance, make payroll first, let other bills accumulate, and hope for the best. We frequently ran short of cash and had heated discussions about how we had gotten into this fix, but we never came up with a solution. He didn’t think that losses were unexpected for a growing company, and I didn’t know any better. Our agreement was that we would each contribute equal amounts when the company needed more cash. Before the crash, we made twenty-one loans, averaging $31,776. We managed to repay ourselves just nine times, with an average repayment of $20,286. In 2005, we opened a bank line of credit for one hundred thousand dollars. In those days, banks handed out loans like candy to any company that showed revenue growth. They were eager to give us money, as long as we pledged our personal assets (my house, his money) as collateral. We used it all in eleven months. The Partner repaid the loan in October 2008 by raiding our cash (without asking me). He believed that we were about to fail and that his investment was gone, but he didn’t want a bank default on his credit rating. His action left the company free of external debt, which turned out to be a good thing.

After the disastrous fall of 2008, I was forced to run the company without borrowing. It was painful, but I managed it—at least until the following summer, when I realized that our Web site was coded in a way that made it hard for Google to tell what it was about. Developers that we hired in 2004 to spruce up the site knew nothing about search engine optimization, so the source code for every page and picture was done without words, only letters and numbers. I was driving traffic to the site by paying for AdWords clicks, but we never appeared on the free rankings. I knew that as times got tougher, I would need better marketing or we would die. So I emptied my children’s college fund ($31,251) and spent it on a new Web site. Development took seven months. The new site rolled out in early 2010, just as the economy and my fortunes began to recover. Coincidence? Maybe. Worth it? Definitely. If I hadn’t spent that money, I’m sure we would have failed.

The debt is always in the back of my mind. If I had invested it in the stock market, I’d still have a lot of the principal available, and in some years it would have produced a handsome return. But I put it into my company, and since we were always broke, I have had no return, neither principal nor interest. But now, with cash on hand, I decide to at least make interest payments. As boss, I can structure them any way that I want. I can set the interest rate to my liking, as long as it conforms to IRS guidelines: the rate must be similar to what the company would pay on commercial loans from outsiders. And I can decide whether to repay principal, or pay interest only, or skip a payment if I want to. This flexibility is very useful as I manage cash flow. It’s the main reason that I have dipped into my own pocket first when the company needed money. If I borrow from a bank or another commercial lender, I have no flexibility whatsoever. The bank will want its money, and if they don’t get it, they will foreclose.

Having decided to make monthly payments, I choose an interest rate. Ten percent is on the high end of bank rates but lower than I might get from a lender who specializes in small, risky loans. I also decide not to make any principal payments. They’re a further drain on cash flow. I write and sign a check for the first payment: $3,225.82. I can’t bring myself to deposit it. Old habits die hard. I feel secure when the company has money, guilty when I take it out.

I CLOSED A DEAL on February 29 with Old Style Packing, who first called us the year before. They had a relatively low budget and wanted a set of tables that could be reconfigured on a regular basis. We had been getting a good number of calls asking for a table like this and I decided to design one.

We had several modular tables in our portfolio, but all were large, expensive, and awkward to operate. A low-cost, easy-to-use version needed a compact and sturdy folding mechanism. We can’t produce that ourselves. I found a company in Michigan that made exactly what we needed: a set with wheels that was inexpensive and looked good. I ordered a sample set in December and slapped together a prototype—I screwed the legs to a piece of plywood. It worked great. Easy to fold, sturdy, rolled around at the touch of a finger, and the wheels could be locked. I made a table design that incorporated this set, but also let us build in custom sizes and woods, and include wiring, at a reasonable price.

Shortly afterward, that call from Old Style Packing came in. It was a great opportunity to move from drawings to prototype, funded by a paying customer.

This is how I have always introduced new products. I love the challenge of designing useful items, but making things just for the heck of it is not smart. First of all, what to build? How do I know that some new design will succeed in the marketplace? And if it doesn’t, what do I do with the prototype? Furniture is large, bulky, and hard to store. I don’t want a perpetually expanding museum of failed designs. The solution to this problem is to listen carefully to what potential clients ask for and build that thing. Everybody wins: my R&D effort is funded by the client, who gets a useful finished product, and I can use the new design in my ongoing marketing efforts.

In the old days, it could be tricky to convince that first buyer to pony up some dough. Drawings of any design concept are bland and difficult to understand, but making a finished prototype before the sale was closed was hugely expensive and risky. That problem has disappeared. Modeling software shows the client exactly what the design looks like and how it works. I deployed this weapon in my proposal for Old Style Packing. Once I had a fully functioning virtual model of the leg mechanism, the complete table design was easy. I sent the proposal off, and a month later, the deposit check arrived. The job will be in production in the middle of March.

The first day of March is a milestone on another project. The Company S table is ready to go into the finish room. Making furniture from scratch has two phases: building and finishing. Building involves cutting lumber up and putting it back together in more useful shapes. Finishing is applying liquid to the wood surfaces, which then dries into a protective layer. Correctly applied, the finish coat makes the surface smooth, shiny, pleasant to touch, and protects the wood from dirt and water.

In our shop, we use a durable, water-resistant finish called a “catalyzed polyurethane.” It is very tricky to work with, however. It consists of three parts: resin, solvent, and catalyst. These must be mixed in exactly the right proportion. Before the liquid finish hardens—about a fifteen-minute window—it must be sprayed onto the wood in an even coat. The spray gun and hose weigh about two pounds. They must be held out at arm’s length, at the proper distance from the wood, and moved at the right speed. Spray too little, and the surface looks dry and blotchy; spray too much, and the finish will drip and run before drying. We often modify the natural color of the woods with stains and dyes. Stain mixes must be measured with the same precision as the urethane. And everything must be clean—dust in the finish is unacceptable.

The only way to fix a flawed finish is to wait until it dries, sand it off, and start over. Even if one tiny part of a tabletop is bad, the whole thing has to be redone, because a refinished section is never a perfect match for the rest.

Eye of an artist. Knowledge of math and chemistry. Physical strength and endurance. Meticulously clean. Performs under strict time pressure. It’s unusual to find all this in one person. A good finisher is hard to find, and worth a lot.

Dave Violi, my finisher, has unleashed his superpowers on the Company S table. At the end of the first week of March, we reassemble it for final inspection. It is magnificent. Everyone in the shop stops to admire it. And, without exception, the first thing they do is run a hand across the top. A perfectly smooth, even topcoat, with no dust specks or other flaws, has been applied to each of the three oversize top pieces. Dave is a master of martial arts, and his strength and agility made the difference. We have eight days before our client’s board members gather in their new headquarters. We need to get the table to the customer.

Back when we made dining room furniture, I did all the deliveries myself. I learned a lot from watching a buyer’s first encounter with their purchase. Clients are very nervous on delivery day. Our sales process is designed to construct an image in the client’s mind of what the finished item will look like. Their hope that we will do a good job will be confirmed, or not, by the actual product. When I was doing delivery, I could make sure that everything went smoothly and that the client was pleased. Happy clients make the final payments. Unhappy clients cause delay and distraction while their problem is addressed.

Google changed our market from local to continental. Our clients are now scattered across the United States and Canada. We have to ship a large, delicate product over long distances, as quickly as possible. Damage is a disaster. Our products are one of a kind, so we can’t pull a replacement part out of inventory.

We have to trust strangers to transport and deliver our work without damage. Unfortunately, both truckers and installers vary widely in quality: some are careful and competent, and some are not. How can we find good people? How can we get them to take extra care? And what, besides paying them, could we do to make sure that they succeeded? I arrived at three methods: use middlemen to find the trucker and installer; redesign the tables for ease of shipment and assembly; and optimize packaging design. How does this work?

First: the middlemen. Our volume of business is too small to impress a large trucking company, and we need to hire installers in places that we have never been to. So we use a freight broker for trucking and an installation broker for installation. Both firms take the parameters of our job and get bids from interested vendors. It’s their job to identify quality vendors and deal with them if things go wrong. The vendors get significant business from the middlemen, who represent lots of small companies like us, so they pay some attention. And we give significant business to the middlemen, so they pay attention to us. This is an instance where it is not in my best interest to shop around for a low price. We need to commit to our middlemen so that they act as our champions. This raises shipping and installation costs, but saves us a lot of time and trouble. I chose one freight broker and one installation broker years ago, and have stuck with them ever since. Our person-to-person relationship is as important as the money in making it work.

Second: the design. Traditionally, a person moving and assembling furniture is presumed to have specialized tools and some skills. Our clients have neither. Detailed instructions won’t help. Our tables are all different from one another, so a universal instruction book won’t work. We’d need a new one with every project. I tried this a few times, and it took huge amounts of time. The clients ended up calling me for help anyway. I decided to rethink our construction details with ease of assembly as a primary goal. All tables would henceforth consist of components small enough that one person could pick them up and move them, but big enough so that there weren’t too many pieces. The parts would self-align so that they could go together only in the correct position. Hand knobs would join all the pieces together: no tools required. Even an untrained person could see at a glance exactly what to do. We adapted this system to everything that we built. As it turned out, this also speeded up construction considerably. In order to build a table, it needs to be put together and taken apart multiple times. So thinking about assembly from an ignorant client’s point of view had benefits for our sophisticated workers as well.

Third: the packaging. This must do more than just protect the goods. It has to communicate to everyone on the path from our shop to client. We want to send a different message at different points in the journey. The packaging has to intimidate the warehouse worker, convert the installer from neutral actor to enthusiastic champion of our product, and delight the customer. After a lot of experimentation, we have settled on a two-layer strategy: wrap every part, crate the whole order. Every piece of the table gets wrapped in foam, then in cardboard, and is clearly labeled to show exactly what it is and how to open it. Then all the pieces are bundled into a custom-made crate with wood sides but no top, so that there is no temptation to stack another load on top of it. To the truckers, the shipment looks heavy, strong, and expensive—something that is worthy of extra care. When installers break down the crate, they find nicely packaged, clearly labeled, and easy-to-handle pieces that are a breeze to bring into the client’s premises. We’ve heard numerous times from clients that the installers told them that we made the best table they had ever worked with. Good packaging converts the installers into our ambassadors and puts the client in the right frame of mind.

Designing tables for easy assembly ended up saving me money, but crating and wrapping is expensive. My shipping manager, Bob Foote, needs a full-time helper to keep up with shop output. It’s interesting to break my whole crew down and see what proportion of the workforce works in the different parts of the operation: out of fourteen workers, four are doing design and sales, six are on the shop floor building the furniture, two are in the finish room applying coatings, and two are handling logistics. In other words, the people doing the woodworking, which is what most people think is our primary activity, are outnumbered by people adding value at other points in the supply chain.

It takes two days to package all the components of the Company S table and to build custom crates for the oversize top sections. After it ships, I keep my fingers crossed, and the crate arrives safely at the installers, and a day later we hear from them that the install has gone well.

I send the final invoice, for $7,551, to my contact at Company S, along with care instructions. I don’t hear anything back, which is slightly unsettling but not unusual. I presume they’re busy with their board meeting. They have ten days to settle up. They’ve made all the previous payments without delay, so I turn my attention to other matters.

Emma Watson has been talking to the government. The U.S. Department of Commerce is eager to help us. She’s also found that Pennsylvania has its own export assistance program and maintains trade missions in a large number of foreign cities. Both of them contract with a third outfit, the World Trade Center, to provide manpower. Emma makes an appointment with the WTC to come see us. Two guys show and give us three cards each.

Like everyone who visits us, it takes them a little while to wrap their heads around the concept of a cabinetmaking shop that makes nothing but conference tables. These export guys have been in lots of factories, and ours does not compare to bigger entities. In some areas we are very advanced—our robotic machines and our marketing on the Internet—but in lots of others, we look like what we are: small and undercapitalized. The government guys don’t seem to care about any of that. Their job is to promote exports, and to do that they need American companies to work with. They wax ecstatic about our inevitable triumph on foreign shores, aided by their services. While they are gassing on, I’m looking at them and wondering what their day consists of. How hard do they work? Do they ever wonder if they won’t get paid? (I suspect not, since the guys who print the money are issuing their checks.) Would I ever want a job like theirs, where I put on a suit every morning, do something entirely predictable, and then go home? Where I knew exactly how much money I would make today, tomorrow, and in the future? Would I be happier if my life had more security?

After the trade guys depart, we contemplate the pile of beautifully printed, expensive brochures that they left. Our first decision is whether to sign up with the federal program, run by the Commerce Department, or the state program, run by Pennsylvania. The local contacts are the same guys we just met with in either case, but the staff in our target countries is not. Emma has no doubt that the feds are the way to go. Her argument: every person on Earth will take a call from the U.S. ambassador, while nobody has heard of the Pennsylvania trade delegate. Go with the people who can open doors. I agree. That decision leads us to the various levels of services offered by the Commerce Department. We settle on something called the Gold Key Matching Service. (Who thought of that name?) For a couple hundred dollars, a Commerce Department office located in our target cities will call around to local merchants to see whether any of them are interested in meeting with us. The process starts with a questionnaire, in which we describe what we make and what kind of foreign business we are after, and continues with a phone interview with the trade rep. I’m still only half committed to the whole idea of exporting, but Emma is enthusiastic, so I agree to sign up for the Gold Key Matching Service. I have to spend money at this point—three hundred dollars—but I feel like that’s not too much to see what happens next.

On March 9, our new folding tables are ready to ship. I take a number of photographs of one of them, both raised and folded, to add to our Web site. I can’t get a good picture of it under our glaring lights, so I generate a nicer image with a rendering program. I only post renderings of pieces we’ve actually built and always include some shots of the table on our shop floor, even though those are often terrible photos. I want to prove that we actually do the work we show.

Photos are easy; pricing is hard. What should I charge for this new product? What should I charge for any product? It is a surprisingly difficult question. There are two ways to think about this: first, what is the customary market price? and second, what does it cost us to make it? The first approach doesn’t work for us. We have almost no information on what our competitors are charging because they don’t publish price lists. We have some idea of what factory-made tables cost, but that doesn’t help us price a custom project. So we work from internal cost projections.

Our pricing spreadsheets predict the costs of each project, but when I check the actual build data, I find that the forecast was often wrong. Errors and differing skill levels of the workers on each project make for wide ranges in build hours, even for very similar tables. And the actual material costs are murky, too. The spreadsheet predicts how much wood we need to buy, but the area calculations are very imprecise, and we haven’t updated our cost data for the different woods since 2007—I haven’t had time to track down hundreds of (volatile) prices. I’m assuming that the recession has brought demand and prices for wood down. Given what happens on the shop floor, inaccurate cost data may not matter. Wood comes in random sizes, with some defects, so yield is inconsistent. Sometimes we can use scrap from earlier work in a subsequent project, sometimes not. Errors and rework require duplicate material orders and extra time. In order to give ourselves a margin of safety, we mark up the theoretical material costs by 40 percent to cover uncertainty. Then we add that number to our labor costs, calculated at $78 per hour, no matter which worker, with associated pay rate, is working on the project. The sum of those numbers is, theoretically, our minimum profitable cost. We then mark up that number by 7.5 percent as an additional safety margin and add 2 percent for the sales commission. That’s our theoretical price for any of our products.

I have told Nick and Dan to get at least the calculated price out of the customer, but sometimes we need to cut a deal to make a sale. We might also change the price if the spreadsheet kicks out a number, say $10,032, or $40,151, which is just above an obvious pricing cut-off point. I’ll gladly give up a couple of dollars to bring a number from five digits to four, or from one decile to its lower neighbor. I’ve read enough about retail pricing to believe that people really do respond to slightly lower numbers in those situations.

In aggregate, this system works. Our material costs are lower than what we charge for them, and we use about 5 percent fewer shop hours than the spreadsheet predicts. This system has produced positive cash flow and profit in the past two years. But in a particular case, I am wary of the time-cost predictions. They are wrong as often as right, sometimes in our favor and sometimes not.

Which brings me back to my modular table. I’m putting it up on our Web site and I want to put a number on its page. I have the build hours from the Old Style job. We had predicted that it would take sixty-nine hours to build four tables, and it had come in at fifty hours—28 percent under. The material costs were predicted to be $2,234, but we had ordered only $841 worth for that job. (But we might have used up materials we had on hand, so that $841 is not our actual cost.) Based on that data, I offer a lower price for future iterations than we charged to Old Style. They paid $9,210 for their set of four walnut tables, two of which had data ports. I rerun our pricing spreadsheet using the actual build times, and find that the new number is $1,594. Per table in a set of four, that is, and taking advantage of the efficiencies inherent in producing multiples. I expect that we will get orders for more than one table, and that the actual order size will be more than five thousand dollars.

I set up the product page so that you see a picture of the four tables arranged in a U configuration, with pricing per table of $1,594. If you read the body text, it’s clear that I am talking about the per-table price in what will be a multiple-table order. I like the idea of featuring a low number, as I figure it will grab the shopper’s attention. We’ll sort out the particulars when a potential buyer calls us.

We now have 182 tables on the site, sorted by shape, features, and price. This makes it easier for buyers to find the kind of table they are looking for, and easier for Google to serve an exact match when people search for specific terms. We run different ads for a wide variety of table types: large, small, round, custom, boardroom, and so on. We haven’t been pushing modular tables, so I write a new ad intended to drive traffic to that page. I check to see what kind of traffic the keyword “modular tables” will generate, and Google assures me that this is a heavily searched term, but without a huge amount of competition from other table sellers. I write a catchy headline, set an amount I’m willing to pay each time a viewer clicks on the ad ($3.50), and set a schedule. It will appear from seven a.m. to ten p.m. every weekday. This ad is just one of 126 we’re running. Each is tightly targeted, so no single ad generates a huge response. In aggregate, they produce a steady stream of calls.

IN THE THIRD WEEK of March, I get a call from Nigel at Eurofurn. He has spoken to his superiors at the home factory, and they have agreed to host me for a couple of days. I have to buy the plane ticket, but they will pay all other expenses. The trip will cost me about fourteen hundred dollars. I’m happy to lay out that cash if I get a peek into their factory. I have never had such an opportunity before—my domestic competitors have no reason to let me see their operations, so I haven’t tried to set up any visits. And, frankly, I’d think long and hard about letting any of them into my shop.

We settle on the last week of April. I’ll fly on Monday and return on Friday. Also, he wants me to start design work on a table to be placed in their New York showroom. It’s to be an upgraded version of their current table, which doesn’t have up-to-date wiring capabilities. They won’t pay me for my design time, but do agree on a reasonable price for the table itself: $6,523. This will be our first attempt at combining the Eurofurn look with our own design details. The table will be challenging: the top will be an equilateral triangle with radius corners, made in three identical pieces. We will be integrating power/data lids into the top, and the wood grain needs to run across the top and lids without disruption. I have a good idea how we will produce it, but I want to get a better sense of how Eurofurn would approach a project like this first, just in case they have some special tricks that will work better than our methods.

The next Wednesday, I get a call from my contact at Company S. My stomach drops. If we don’t hear from a client after an installation, all is well. But if they do call, it can be good news, or bad. Sure enough, they’re unhappy. After the board meeting, they found scratches on the table. Not big ones, the executive assistant tells me, but she could see them from a certain angle. She is convinced that there is something wrong with the finish. That seems unlikely; I carefully examined the top before it shipped and everything looked good. Sometimes clients scratch a table, but they almost always take responsibility for it. Anything other than laminates (commonly called Formica) and granite—including glass, marble, metals, and wood—can be easily scratched if a sharp object is dragged across the surface. Every wood table eventually picks up a lot of tiny scratches, mostly from users’ jewelry and laptop bottoms. It’s an unavoidable consequence of normal use.

I ask her to e-mail me photos of the damage. They arrive the next day. They’re terrible, taken at very close range with a shaky cell phone, but I think I can see something. In the universe of scratches, they rank about 1.5 on a scale of 1 (undamaged) to 10 (chewed by wolves). In other words, marks that are normal for any table in actual use. Meanwhile, I asked Dave Violi: did anything go wrong? Not as far as he knew. He built up a very thick layer of finish to get the look the client wanted, but the process had gone well.

It looks as though the client is overreacting to something that they did, but they don’t see it that way. They still owe me more than $7,500. Blaming them isn’t going to make them happy or get me paid. But it’s impossible to repair a finish like this on-site. The chemicals in the sprayed finish are noxious, residue would end up all over the room, and if you don’t recoat the entire top, the newly sprayed section will look different. If I am going to respray it, we will have to ship the top segments back here. Despite the logistical headache, I’m leaning toward agreeing to respray. Any alternative is likely to turn into a Stalingrad—an endless, damaging battle, most likely ending with bad feelings from the customer and an unpaid balance. A respray will cost a little bit less than the balance due, but at least I will have satisfied the customer. Unless, that is, they damage the table again. I really need to figure out what happened out there.

I make the trip on March 27. My journey begins at four-thirty a.m. For the convenience of nobody, Philadelphia TSA has decided that all travelers from four terminals will have to go through one security line. That’s an extra forty-five minutes that I didn’t plan for. I end up jumping to the head of the line and sprinting to make my plane. Two flights and a long drive later, I walk into the boardroom, the executive assistant at my side. From the doorway, the table looks perfect. “You have to look from a certain angle,” she says. I look from a certain angle: nothing. “You have to get up close.” I get up close. When my head is inches off the tabletop, I can see some fine scratches. Normal wear. But I make sympathetic noises. “Do you think the finish is defective?” she asks. The honest answer is “No, you did this and it’s your fault.” Instead, I ask her what had been on the table. She goes through the inventory: glassware, placemats, folders with notepads. She’s been very careful. I ask about unglazed mug bottoms—common culprits—and she tells me she thought of that and sent their mugs to a local potter to be reglazed. At this point, the CEO comes in. He says he’s disappointed that this happened. He really loves the table; in fact, one of his board members even asked who made it because he’s interested in having us make him one. And as soon as we fix this, he will tell me who it was. (CEOs of billion-dollar companies don’t get there by being stupid.) I surrender. I agree to retrieve the pieces in mid-April and return them three weeks later, before their next board meeting. The boss offers to pay for a rental truck, but we have to fly out my shipping manager to manage the packing and loading process and drive the truck back ourselves. It will be a huge, expensive pain.

I am still mystified as to what caused the scratches. We take out all the dishes and identify the culprit: it’s the mugs, after all. The local potter didn’t reglaze the bottoms, only painted them with varnish. I dragged one of the mugs around on the table, and voilà! Tiny scratches.

I’m back by midnight, exhausted. I feel like a chump for knuckling under, but arguing would have resulted in a much bigger mess. When I get home, I find a surprise. On our front door, my wife has taped up one of her drawings—a cartoon showing me with my arms raised in triumph over a kneeling businessman, who is handing me a large stack of dollar bills. It didn’t turn out that way, but it’s nice to have some family support.

This is a business book, but you can’t understand a boss without knowing what he goes home to. I have been blessed with a happy family life. I met my wife, Nancy, in college, when she was eighteen and I was nineteen. Immediately after graduation, we moved into our first apartment together and I opened my shop. Nancy has been my bedrock during many turbulent years. She grew up in a family with intermittent income and she can deal with a life of financial uncertainty. And she has proven equal to the challenge of raising a child with special needs.

We have three sons: eighteen-year-old fraternal twins, Peter and Henry, and a younger boy, Hugh, who is sixteen. Hugh and Peter are typical kids and attend the local high school. Henry is severely autistic. We knew something was wrong quite early, as his twin brother developed normally. Henry lagged behind in every way. He received an official diagnosis at the age of two. After that came years of therapy, special classrooms, speech therapists, and behavior experts. Nothing cured him.

At age eighteen, Henry still has the mental development of an infant. He cannot speak—he tries, but the brain development that allows the tongue and lips to form a full range of words never happened. He cannot read. He has little interest in other people. He won’t watch TV. He likes only a few things: to be driven around in a car, to listen to a particular Beatles CD over and over (turned up as loud as the boom box will go), and to eat. He is 6-foot-3, 205 pounds, and still growing fast.

Henry is volatile. Since he can’t talk, he has great difficulty communicating what he wants. We’ve learned to anticipate his needs, but sometimes he weeps with frustration, or slaps his own head repeatedly, or jumps up and down bellowing at the top of his lungs. And if someone gets too close while all this is happening, he will attack them—grab them around the neck and try to throw them to the ground.

Before age twelve, Henry was usually very calm and cooperative. When he started puberty, the violent behavior appeared. By age fifteen, he would attack my wife out of the blue, once or more a week. Now, at eighteen, the hormonal surges are abating and the tantrums becoming less frequent. But he is much, much larger. The episodes are harder to handle. Nancy is wary about being alone with him, but even so she’ll take him with her in the car as she does errands, and he behaves well. He likes driving and going to the grocery store—it’s like a food museum for him.

At first glance, he’s a tall, handsome teenager. It takes a moment to see that something is off. Fortunately, people are quite tolerant of him. Even if he bellows or bursts into tears, they take their behavior cues from us. If we treat whatever he’s doing as normal and expected, everyone stays calm. It’s embarrassing, though, to be in public when he’s difficult. It doesn’t keep us home all the time—he gets bored, and so do we. Not to mention that two hundred pounds of autistic boy jumping up and down has cracked our plasterwork and loosened our stair treads. So we take him out, and take our chances.

Henry’s twin, Peter, is a direct refutation of both astrology and the notion that Nurture is more important than Nature. Born two minutes ahead of Henry, he couldn’t be more different. He’s been accepted to the Massachusetts Institute of Technology on early decision and has inherited his mother’s gregarious nature and easy charm. Our younger boy, Hugh, has different interests, but he is smart and a very diligent worker. Maybe we’ve been compensated for the difficulties of raising Henry with these two.

I’m not a hero for running a business while raising a special-needs kid. Everyone has some kind of trouble in their life. Every boss needs to make a decision as to how to deal with pressure. Work all the time, drink, cheat on your spouse, yell, road rage? Or more benign choices? Take your pick. If you can. High levels of stress drive you to your worst behavior.

I had some tough days before the twins arrived, but in retrospect those years were a lark. The day you have children, you enter a different world. Mix parenthood with the problems of a business that doesn’t actually make money, and stress is a much bigger issue. I’ve tried to keep the work problems at work, but sometimes I’d bring the stress home and would suddenly explode over some stupid thing. It was usually an innocent request made by my wife, assuming that of course I would be able to supply the money or time required. Then the fight would start, with me shouting that she had no idea what was going on at work, and she replying with equal vigor that I was always promising that someday the business would go well and I would have more time and money for the family, and it never happened. Every couple has a fight that just keeps coming back, and this was ours.

When the recession arrived, and I was struggling every day to keep the doors open, the stress returned to unbearable levels. I decided that the only way I could deal with it would be to tell Nancy everything that was happening at work, good or bad, every day. She hated hearing it at first, as it reminded her of difficult times in her childhood. But we ended up getting along much better. Keeping a barrier between my work and home lives was a mistake. When my wife had a clear picture of the situation, she became an ally instead of an adversary.

My kids are old enough to take care of themselves now, and Henry is away from home much of the year. When he comes back, life is harder. But as long as the business does well, I’ll be in decent shape.

ON THE LAST FRIDAY of the month, I review our sales numbers and they aren’t good. We end up selling just $135,732 in March, drastically undershooting my $200,000 target and leaving us well short of our quarterly target of $600,000. Sales for the first three months are just $543,003. Nick had the worst month, booking only $25,502 in orders. Dan did better than January and February, closing deals worth $49,783. Neither of them sold anything after the eighteenth. I am the champion of the month, with $60,447 in orders. “Best Salesman” is a prize I didn’t want to win. Watching Nick and Dan falter makes me feel very uneasy.

The sales pattern for March is odd. The number of orders hasn’t decreased—it’s actually grown from eleven in January and twelve in February to sixteen in March. More customers should be a good thing, but in this case, the additional orders don’t amount to much. The value of the five smallest orders adds up to $5,854. Peanuts. And the remaining eleven average just $11,807. This is well below our average order for the first two months of the year, which was $17,688.

The size of our orders varies widely. The majority of our jobs fall in the five- to twenty-thousand dollar range, but their aggregate dollar value is less than half of our total. Big jobs are important, and the biggest—more than fifty thousand dollars—have a disproportionate effect on our fiscal health. Land a few whales, all is well. Catch only minnows, and it’s hard to make our target. And March brought us only little fish. Nick and Dan assure me that they are working on some jumbo orders that will arrive soon. We could be back on track any day now.

One thing is undeniable: undershooting our sales target has affected our cash flow. If we’d hit our goal, we’d have another thirty-two thousand in cash right now, from deposit payments. A couple of months of strong sales followed by a weak month is not unusual—I’ve seen that pattern many times. But I have also seen sales shrink for several months in a row, and that’s a disaster. The shop is staffed for a certain production level, with all the attendant costs of payroll and machinery. If we fail to bring in work at the same rate we produce it, we get hit two ways: incoming cash falls below our spend rate, and eventually we will run out of work to do. Then come layoffs.

I’m not sure what I need to do right now, but my confidence that I could back out of my sales role has evaporated. I can’t rely on the other two, and don’t want to take a chance that they muff a deal when we are behind our quota. I am going to work more deals myself, starting with sucking up to Eurofurn. Maybe they’ll throw more business my way.