March 26, 2023

You’re the Boss Blog: Running the Numbers: Why Our Productivity Improved

Staying Alive

The struggles of a business trying to survive.

Paul Downs: I don't know how many tables we made in 2011.Courtesy of Paul Downs Cabinetmakers.

My company has 14 employees. In the office we have me, two sales engineers, one design engineer, a part-time bookkeeper and a part-time administrative assistant. The other nine are out on the shop floor. They build, finish, and ship our product.

Our office is much like any office: people manipulating ideas about stuff, and moving money around. The factory, however, is set up to make a very particular product. The equipment, the arrangement, the skills of the people are all aimed at making wood conference tables. I have found, after 26 years of messing around with various forms of woodworking, that conference tables can support a business better than anything else I have tried. So that’s what we do. We can, when we want to, make other things. Some of our clients want us to build them credenzas or reception desks, and we do.

You might wonder, since this is about our production (Wednesday’s post, part of a five-part series, was about sales), how many tables we made in 2011. The answer: I don’t know, and I haven’t bothered to find out. I could spend some time in our database and come up with a number, but it wouldn’t mean much. Dividing our output into the unit “tables” doesn’t make much sense, because the tables we make vary so widely in size and complexity. This and this and this are each one table.

Instead, we take each sales contract that we write and break it into one or more work orders, each associated with a piece of furniture or, in the case of modular tables, with a set of identical objects. Each work order has been assigned a price by the sales team. That number is generated by estimating the number of hours required to design, build, finish, and ship the product, along with a tally of the materials we are likely to use and their costs. We use custom Excel spreadsheets to do this. These were complicated to write, but they have the virtue of allowing for quick and easy pricing of variations on our designs. (I explained more about this system in a previous post.)

Every day, our workers keep track of the hours they spend on each work order. They do this any way they feel like it — I presume that they take notes with pencil and paper as they go along. I don’t know how accurate they are. I have explained to them many times that accuracy is important, because that is the only way we can determine whether our pricing spreadsheets are kicking out reasonable estimates. They are smart guys, and I believe they do their best.

Anyway, at the end of each day, they enter the hours they spent on each job into our database. This is easy to do, and just takes a couple of minutes. The database has a report that compares the estimated hours with the hours actually used for every job in the system. All employees have access to this data, and I know that many of them keep careful track of whether they are under or over the estimate.

In 2011 our estimates suggested that it would take 14,618 hours to complete the projects that we built. The actual hours reported by the workers were 14,588. That’s a difference of .2 percent, which is pretty good! In aggregate, over a year, that is. In reality there is wide variation on particular jobs.

There were 163 work orders, but only seven of them came within 1 percent of the estimate. The others were all over the place. It appears that there are three reasons for the variation:

1) Estimates for complex work and new designs can be wrong. These are errors by the sales guys.

2) Problems on the shop floor. Machine malfunctions and mistakes happen — an odd thing I have noticed over the years is that they tend to cluster on one job. There are certain projects that just go straight downhill, with difficulty at every step. I don’t know why this is.

3) The spreadsheet models are designed to kick out a few more hours than a project that goes smoothly should require. So when we look at the individual jobs we see some over, some under, some way over, some way under — but adding up, for 2011, to almost perfect agreement between the model and reality.

One set of stats that I wish I had right now is a time series showing our aggregate over/under for the estimates, recording how we were doing at the end of each week, or month, would be useful. (I will be tracking that this year). I do check this now and then, and my recollection is that at the end of the first quarter we were over by about 10 percent, by the end of the second quarter we were over by 7 percent, and by the end of the third quarter we were over by 3 percent. We had a great fourth quarter, which brought the difference close to zero. So far this year, we are about 15 percent under our estimates.

I do have numbers on the value of tables built each month in 2011. When divided by the number of work days each month, we see an encouraging trend. In the first quarter of 2011 we were building at a rate of $6,537 a day. Second quarter: $7,833. Third quarter: $8,199. In the last quarter we got to $9,862 a day. I hired one worker for the shop floor last year, an additional finisher who came on board in April. But that doesn’t account for a 50.8 percent increase in productivity.

That outstanding result came, I believe, from three factors: mastering a number of new machines and processes that had been giving us difficulty in 2010; instituting error reviews as a regular part of our Monday meetings, so that everyone knows how particular technical issues should be solved; and teaching all of the workers to use the database, so that they could see in real time whether they were ahead or behind on estimates.

A 50-percent increase in productivity is, when you think about it, a staggering achievement. But keep in mind that one of the factors in the result, the dollars associated with each project, is controlled by the sales people. If they get more money for a table, the productivity as measured by dollars/day will go up even if build operations are unchanged. But we did not raise our prices by 50 percent. I did change the estimate model to include an extra 2 percent for sales commissions in July, and that’s all. I conclude that the boost in shop-floor productivity is real. This bodes well for 2012. In the last quarter of 2011 we had few of the equipment and process problems we saw at the beginning of the year.

In 2012, I’d like to see our output increase to $2.4 million, which would be a little more than 10 percent over 2011. And I’d like to do it without hiring anyone, because the training period for these positions is long and the cost of new workers is high. That means another productivity increase will be required. The sales guys are going to have to either make more sales, or get more money for each sale. Getting more money can be tricky, but there is one surefire way for us to sell more tables: do deals with people who need something in a hurry.

The clients who currently place orders are willing to wait six weeks or more for their orders to be delivered. We also get calls every week from people who have to have their table in three or four weeks and would happily pay our rates. For them, time is the deal buster. We are turning these people down right now, because I don’t care to lie about when tables will show up. But if we can find a way to decrease our backlog, or at least shoehorn in a couple of extra jobs now and then, we can get that 10 percent without any further investment on my part.

It’s mostly an information problem — we’ll need to know exactly how long it will take to complete our current order book and exactly when we need to ship to see if there is any excess time. Further increases in shop-floor productivity would go a long way to creating the capacity we need to do this.

Paul Downs founded Paul Downs Cabinetmakers in 1986. It is based outside of Philadelphia.

Article source:

Staying Alive: Running the Numbers: Why Our Sales Increased

The Paul Downs sales and design team (that's Paul in front).Courtesy of Paul Downs Cabinetmakers.The Paul Downs sales and design team (that’s Paul in front).

Staying Alive

The struggles of a business trying to survive.

My New Year’s resolution a year ago included a vow to do a better job of recording statistics on a wide variety of performance metrics. As it turned out, 2011 was the best year I’ve had in my 26 years of business life. I actually ended up with what seems like a reasonable amount of money in my pocket.

Looking back, I think I learned a lot by focusing on the statistical aspects of our operations, and I’d like to share some details with you. I’m going to do this in five installments, starting with sales, then production, then material and overhead costs, then labor costs and finally management pay and profits. If you stick with it, you’ll find out how much money I made in 2011.

Why Did Our Sales Go Up?

We use three ways to tally our sales for a year. I like to count the value of new orders received. QuickBooks prefers to sum the value of tables we shipped. And my accountant tells me that, for tax purposes, what matters is the amount of cash that arrived during the year. By any of these measures, 2011 was considerably better than 2010.

Let’s start with the value of new orders. This is simply the sum of the dollar amount of new jobs booked during the year. It does not reflect what happens on the shop floor, just the efforts of the sales people. In 2011 we sold 119 jobs worth a total of $2,155,193. In 2010 we sold 105 jobs worth $1,549,488. By this measure, sales increased 39 percent.

QuickBooks tracks the value of stuff that actually goes out the door. The total value of items shipped in 2011 was $2,005,361. This compares with $1,532,328 in 2010, an increase of 30.9 percent. That is a smaller increase than the rise in new contracts, mostly because of a bunch of jobs that came in at the end of December but won’t be built until 2012.

Looking at our cash situation, we find more good news. We received payments totaling $2,152,589. This includes deposits for new work and payments for finished work, and it compares with $1,580,759 in 2010, a 36.2 percent increase. Our expenditures totaled $1,955,532, leaving us $197,057 ahead of where we started. Out of that gain I paid $79,128 in bonuses, leaving me with a net gain of $117,928. In 2010 we ended up $90,538 ahead of our balance at the beginning of the year, meaning that our cash tally increased by 30.25 percent.

Let’s look at my available working capital over the last two years. At the beginning of 2010, I had $16,239 in the bank, the equivalent of three days of working capital. At the end of 2011, even after paying the bonuses, I had $187,154. That’s enough to run the shop for five weeks. Increasing that cushion has lowered my stress level from Code Red to Easy Blue.

So what led to this turnaround? I think it was an effective AdWords campaign, which feeds an expanded sales operation, which supplies an efficient shop floor. Today I’d like to talk about AdWords and how we process the inquiries it generates. Tomorrow, I’ll talk about what happens on the floor.

In 2010, 90.2 percent of our sales came from AdWords. The rest were repeat orders from my old residential clients and one big order from a local architect we have known for a while. In 2011, clients who found us through our Web site placed orders worth $2,080,157. This was 96.5 percent of our sales. The other 3.5 percent came, again, from old clients.

At the beginning of 2011, I started tracking the number of inquiries we get, which arrive as either phone calls or e-mails. We received a total of 668 inquiries. The vast majority of these came from our Web site. We know this because many of them come as e-mails through the site, and most of the people who call mention that they are looking at the site. I don’t know which calls were generated by organic search results and which came from people who clicked on one of our ads, but, as I concluded after an experiment in November, AdWords is doing the heavy lifting.

In 2010 our AdWords campaign yielded 25,288 clicks, which cost me $100,090, or $3.95 per click. Each click yielded $61.27 in sales. Our Web site received 40,249 unique visitors. In 2011 the ads generated 28,350 clicks, which cost me $109,613. That’s $3.85 per click, pretty close to the 2010 cost. Each click in 2011 yielded $73.37 in sales, a return on investment of 1,897 percent on my AdWords spend. Not too shabby.

In 2011, the number of clicks rose 12.1 percent, the cost per click went down by 9.7 percent, and the revenue per click went up 24.1 percent. How did that happen? Was it just more traffic to the site? The number of unique visitors in 2011 increased to 47,630 from 2010’s total of 40,249. That’s an increase of 18.3 percent. (We have to take into account traffic from this blog, but the site actually received fewer visits from The New York Times in 2011 than it did in 2010. I have only talked to one client who found us through the blog, so I don’t think it has much affect on sales.)

I believe that the number of clicks rose because more people were shopping this year. I think our cost per click went down because I restructured our campaign by increasing the number of ad groups we targeted, so that we were getting results in very specialized searches where the cost is not so high. And I increased the revenue per click by putting more resources into our response to inquiries, so that we could give plenty of attention to likely customers.

In 2008-9, I was the only person in the company who made sales. In fact, taking incoming calls and writing proposals took up most of my time. Each proposal takes several hours to put together, and in those years, I couldn’t let any slide by, as we were desperate for sales. I don’t have good numbers for 2008, but in 2009 I wrote 145 proposals. Running the company also took some effort, so I was working long days. This took a toll on my body. By June of that year, I was almost crippled from carpal tunnel in both hands, but the phone was still ringing every day.

I knew I needed help. I brought one of my best cabinetmakers, Nathan Rossman, off the shop floor and taught him how to make sales. Taking a highly skilled craftsman off the floor was expensive in two ways: I had to keep paying Nathan while he learned sales, and I lost his capacity to produce work. I was worried, when I first started him on his new job, that I would find that my ability to present and sell our tables was unique. I’m very good at talking, writing and putting together proposals.

If I was the only one who could sell our product, it would mean that the company would never be able to expand beyond $1.25 million a year in sales, which is the most I am able to turn out before breaking down physically. Fortunately, our sales methods proved transferable. I put Nathan’s desk in my office. The physical proximity made it easy for him to ask questions and get instant answers. He was also able to listen in on my phone calls so that he would get a sense of what to say. In 2010, Nathan and I completed 357 proposals. I sold 82 (!) jobs totaling $1,233,542. Nathan sold 24 jobs worth $312,833.

At the beginning of 2011, I initiated a redesign of our Web site that made it easier for customers to find particular types of tables, and I rejiggered the AdWords campaign to better align search terms with the page we show clients. This effort would take six months, half for the additional Web development and half to add new content. In April, assuming the changes were going to work, I hired a second salesman, Don Wuest. He started in June. I moved the whole office staff, including me, into one small office so that we could all talk to and listen to each other. As with Nathan, it took Don several months to learn our sales method, but by the end of the year he was selling at a pace equal to ours.

In 2011, the three of us wrote 545 proposals. I sold 49 jobs worth $903,976; Nathan sold 54 jobs worth $972,601; and Don sold 16 jobs worth $278,870. Next year I hope to switch totals with Don. I’ll be throttling back my own involvement in sales with the intention of turning my spotlight on the business itself. Not that I don’t enjoy interacting with our clients, but I’ve been doing it for 26 years and I’d like to take a break.

In a world where people want their stuff delivered fast, it’s important that an increase in sales be matched with more output. Otherwise the wait period, which we call backlog, gets to be unacceptable, and we lose prospective clients. In 2011 our production kept pace with our increasing sales. This was due to an increase in our productivity on the shop floor, which I’ll talk about in detail tomorrow.

Paul Downs founded Paul Downs Cabinetmakers in 1986. It is based outside of Philadelphia.

Article source: