Incentive Pay for Workers

This long, detailed thread takes a close look at the whole concept of bonus and incentive pay for employees. January 9, 2008

A lot of people on this forum express frustration with managing employees, whether it is how to keep them and motivate them, or how to implement a profit sharing plan that doesn't get taken for granted, etc. Over a period of several years, I have come up with something that hits most or all of the requirements for a good plan. It is working incredibly well for my company, and may be useful for yours.

I started a door shop 13 years ago, and now have around 45 employees. We have had lots of changes and growing pains over the years, but now have a great company and a stable work force. I work around 10 hours/week most weeks and make great money, so I assume we are doing this business thing, especially our incentive plan, right.

Let me start from the beginning. It is important to start from a solid premise. Some of the basic premises I try to follow are the following...

You don't make money unless you are getting lots of quality work out the door. This means the sales have to be there, sales and production must work together, production must get a good quality product out the door, etc. Everyone must work together as a team to make this happen.

Reward people on what they have control over, but keep it simple, and have a strong orientation on production output, don't overemphasize other things that don't directly impact profitability. Small incentives for good attendance or such are great, but don't overshadow the main goal: getting quality product out the door.

Profit is what the people with "profit responsibility" get rewarded with. If someone does not have control over pricing and sales and production and purchasing and investing, then they should not get rewarded based on profit numbers. This means the owners, and president if there is one, are the only ones that get rewarded based on profit.

Avoid a system where bonuses can be paid out at the same time the company is losing money. But how to do this without a true profit-based plan? By setting a hurdle, and by being conservative in its design. This I will try to explain as we go.

Now we are at the point of designing an incentive system that focuses on production output. This will apply to almost everyone in the company. As mentioned, all the different functional aspects of the company will have to work together. It doesn't do any good to reward Sales for great numbers if Production can't get it out the door. Plus, if they sell more than you can make, you will just be ticking customers off, etc. What if Sales can't keep up with Production? Now you have a problem that everyone in the company is interested in getting solved. You should have a great sales force/person that can get the work. If not, this is a real problem to the health of your company, and will need to get fixed. Problems will have to be solved, as you know. Plus, if Sales sells jobs that are difficult to get out the door, then they will be hurting their own production bonuses.

In a nutshell, and I will explain in more detail, we bonus people for production that exceeds the level needed for my minimum acceptable profit. Each product we make has what we call a Work Unit, but is just a restatement of Labor Minutes, so it will fit onto a daily work schedule. In other words, I can't put 7000 minutes per day on Microsoft Project, so we use a much smaller Work Unit number so we can create a daily schedule. The more work units (labor units) they get out, the more bonus.

Too complex for now? Just use minutes. You probably use a time factor when figuring out how to charge customers, right? Let's say you get out 20 cabinets a day on average, and the labor is 14 people x 7 1/2 hours x 60 minutes/hr = 6300 labor minutes.

6300 labor minutes per day is your current average output. But you may get people arguing over the minute assignments to each product and service, so it helps to separate the issue with my Work Unit, which is much harder for people to argue with.

1) So for now I will use minutes. Assign labor minutes to each product and service you offer, regardless of how you charge. If it is a job that has to be produced or a service rendered, you want the crew to do it as efficiently as possible. Do not distinguish between cheap and expensive labor. Sometimes a carpenter has to take the garbage out, and sometimes a helper is learning to do a carpenter's job, so it doesn't matter, and you want to keep things simple. The labor minutes are the same for all materials, and you should also be tracking labor costs per labor minute.

6300 labor minutes / 20 cabinets = 315 labor minutes per cabinet in our example. (No, it will not be that easy to assign labor minutes to all the actual products you make!)

2) The owners must decide what their minimum acceptable profit is for each month, except maybe December. They deserve this. This is their return on their investment, just like you want a return on investment if you buy a stock or a bond. Owning a business is risky, so it should be at a level they are okay with it if the company just does okay that month. 6 - 8% is okay in my book. This is an individual thing. This is theirs, no one can touch it. They can take it all home, or decide to invest it back into the business. Their call.

3) Determine how many labor minutes of output are required to make their 7% profit in a 20 day month. Remember, I said this would be a conservative calculation? Some months have up to 23 days. We must use 20 days of output.

Let's say you get $500 per cabinet (I know my numbers may be way off - just examples). This is $500/315 labor minutes = $1.587 revenue per labor minute. The owners will want to track this number (Revenue per Labor Minute) to see when it drops off, and to watch it grow after a price increase. Yes, it fluctuates with the different materials used, but still is a useful number.

Let's also say that you determine it takes 6000 minutes a day x 20 days = 120,000 labor minutes in a month, or 120k x $1.587 revenue per labor minute = $190,440 revenue in a 20 day month to make their 7% profit. Cool. No one gets a production bonus at 6000 labor minutes a day, but the owners do get their profit, even in a short month.

4) Determine Incremental Profit for Incremental increases in Production output. Say your margin is 50%, so if 1% of $190,440 = $1904, you have to invoice twice that, $3808 of product, which is $3808/$1.587 rev per l.m. = 2400 labor minutes of work to gain an additional 1% profit.

See where I am going? 2400 labor minutes/20 days = 120 labor minutes a day over the required 6000 labor minutes will net an additional 1% profit for the company. This is what you will start sharing with the crew.

Make it easy. Use round numbers. I will do this as an example: 200 additional labor minutes a day, on average, will equal about $3173 additional profit at the end of the month. It makes more sense to use easy numbers when motivating based on numbers. We will use 200 for our example.

5) Figure out how to share it. I think that about 25 - 30% of the incremental profits should be paid out in production and other incentives. Now you make a spreadsheet with everyone's wages, and start plugging in percentages of their wages as bonuses, and play with it until you get the right mix of bonuses. A couple of things: At this point, use 23 day months to figure what someone's wages are for the month. We are going to be conservative, to guarantee we are making a profit before bonuses are paid out, right? This is important. Item 2 - Your payout must take into account all payroll costs. For us, this is about 1.18 times base pay, to account for taxes and workers comp, etc.

So... 25% x $3173 = $793 to be paid out, but divide by payroll costs, $793/1.18 = $672 gross paid out, payroll costs will equal the $793 amount.

Take out the owner(s), and in our 20 person company, you have 19 people to divide the $672 among. On the spreadsheet, everyone in the same position will get the same percentage. People talk, don't they? You don't want one floor guy bragging that he makes a higher percentage than others. He already makes a higher dollar wage, so his bonus will be higher since it is a percentage of his hours worked in the month x base pay. Everyone besides the owners gets a production bonus offered. Even maintenance, bookkeeping/accounting, human resource, everybody. This keeps a team focus on production. And junk that gets remade doesn't count toward their bonus! Supervisors get a higher percentage, and the Production Manager gets a huge percentage for each incremental increase in output, so a significant portion of his pay is from attaining high production output. If the Production Manager just achieves the minimum acceptable profit, their pay should be very modest as a result. If the Production Manager cannot regularly achieve and exceed the minimum acceptable profit, assuming the work is available, then the owners should be on the lookout for a replacement, or take some other remedial action.

You have now determined what you think is a workable production incentive program, and have tweaked it with other things, too, like the 2% attendance bonus for perfect attendance we pay, and are ready to present it to the shop. Tell them this is a work in progress and will change every month for awhile until you get it down right. You give them a sheet saying what their bonus will be, and show how the percentage will accrue for each increase in production output.

It will essentially say something like, "For each 200 units of average daily output above 6000 units, you will receive 1% of your base pay x hours worked in the month as a performance incentive, paid the following month." You also will note in the meeting that in March and June you had average outputs of over 7600 units per day, so that would mean you would have made an additional 8% plus other (attendance?) bonuses. This is a time to challenge them. Ask them what they want the output to be this month, and I guarantee someone will be piping up with some big numbers, big enough to make you bite your tongue to keep from telling them that is just not possible. Give this some time, and you will be shocked at what they can accomplish once everyone gets focused.

Note: It is important to keep track of the completed jobs at the end of each day. One day might be 2000 units, and the next might be 15000, so you will daily announce the previous day's output, and the monthly average-to-date. To do all this you have to be determining the number of units in each job as they get booked, and of course any modifications to the job that get made mid-job.

I also do an incentive program for our accountant that rewards collections, to minimize accounts past 60 days and minimize bad debts. She gets dinged bad for both of those.

To top it off, if the shop runs out of work, they get to go home with pay for the rest of the day. This keeps them humping when they think the work is getting low. We found that if they slow down, it is hell to get them to speed back up. Plus, if they have been getting great numbers, they deserve this. To keep this from happening, we give Sales two hours off if they go the whole month without letting the shop go home early. But if they oversell, they will just have ticked off customers, but they are very good at the balancing act and keeping just the right amount of work on the schedule.

Forum Responses
(Business and Management Forum)
From contributor B:
It looks as if you have a nicely workable system in place, once the metrics are known concerning profitability and productivity. Let me applaud you for paying your workforce to help prevent bad attitudes and work habits. Riding the clock does not really help anyone in the long run. Managing sales as indicated is impressive. Your sales force is forced to be on the same team, instead of things becoming an "us vs. them" kind of situation as is all too common.

I have a question, though. How have you determined, from a percentage, factor, numerical point of view, what these metrics are? One of the biggest battles seems to always be determining what is really profit and what is really productive, and having these match up with bank statements, not formulas or estimates from office jockeys.

From contributor P:
That is quite a system - I like it. Bonus for getting above the minimum acceptable profit good premise. This is one of the problems in our economy - a person gets paid for being there, rather than for production. The problem for most of us is that what we make is harder to measure than what we do, i.e. custom work. Maybe this is a marketing problem in that we don't make something that is more controllable. For sure, with a product that is more controllable, the business is more sellable. But diversity keeps us from getting slow or from being knocked out of business by China. How would we go about setting bonus targets for custom work?

From contributor G:
Contributor P, it's just a thought (as I'm in the same boat, custom wise). Maybe we're thinking too deep. Don't we already know the target? When we bid it, we should know how many hours we have allocated toward it. Could we not use that? Anything faster than that is bound to be a good goal.

From the original questioner:
You can't get too hung up on getting perfect labor minutes. But you can get accurate profit numbers by using Generally Accepted Accounting Principles (GAAP). An accountant can help you set up Quickbooks or such correctly. You will want to use accrual profit statements even if you do your taxes on a cash basis. Adjust the profit statement to reflect for actual labor and material costs for the month. You will have to do a month-end inventory of raw lumber, ignore work-in-progress. Also, the owner's bi-weekly paycheck should be a true reflection of the job function performed, and not be influenced by their owner status. I take a very modest paycheck to reflect how many hours I work, and take most of my pay in profit distributions.

Back to labor minutes. Don't fret too much. You estimate time for jobs well enough to bid them and earn a living. Sometimes a hand will get a job done quick, sometimes it gets done 3 times. Use your best estimate, and continue looking for ways to better estimate. All you need is some consistency. If one of your products is always undervalued in minutes, then it just changes the average value of your minute assignments, i.e. you will show a higher revenue dollar per minute, and will need fewer minutes to meet break-even or minimum profit. The calculations will still serve the exact same purpose as long as you are consistent. Eventually, as you improve your process (think "lean" and "continuous improvement"), the actual minutes required to do a job will change, but do your research before changing minute assignments too much.

From contributor C:
I must applaud you for such an effort. I had almost lost all faith in a bonus/profit sharing plan until I read yours. Most plans fail because of inequity or perceived inequity in distribution. Your plan addresses this quite well. The teamwork and balance factor is crucial for success. I do wonder how you address major mistakes in your plan. A few that come to mind other than the standard mis-cuts and sanding mistakes are ones like wrecking the company truck, wrong stain color on a job, etc.

From contributor Y:
We base our bonus program on four Key Performance Indicators. Our existing Productivity program is similar, but we measure performance in boxes per labour hour. Our current goal is .49 boxes per labor hour. Or by your formula, 123 minutes per cabinet. We do custom kitchens, both frame and frameless. Our staff realizes that because of the custom nature of our business, some cabinets are much more time consuming than other simpler boxes. The .49 boxes per labour hour is an average that works out over the week.

We also include in our KPIs: On time delivery (5 week lead time), Costs (calculating rework labor, and waste percentages over goals), Quality (as reported from department quality audits). Program is now two years old.

From the original questioner:
A wrecked truck would not affect our numbers. We offer free delivery, and invoice before delivery. The delivery would just be delayed, the truck is insured. The driver better not be at fault, or his job may be in jeopardy. Incentives can only be paid out for invoiced jobs. Yes, we have had to redo jobs like everyone else. It can really hurt a bonus, just like it hurts profit. Everyone pays, and fortunately it doesn't happen too often.

From contributor P:
I think for custom work, the question with this type of system is, what is the unit of measure? If you don't have accurate prediction on the work, then any incentive plan will soon become a mass protest and disincentive at the least.

For cabinets, a box is a perfectly good unit of measure. On a nurseís station, reception desk, store fixture, a box as a unit of measure would not work.

You could take a macro or a micro point of view on the more custom work. The only thing that is consistent from one of these jobs to the next is the processes involved. This would be a micro point of view. You could look at similar jobs that you have done before - this would be the macro point of view.

If you take the micro point of view, you could look at the labor aspect of each type of operation as associated with a certain part, e.g. a partition, drawer, Corian counter, die wall, etc. This method would be used to create your estimate and your labor tracking targets and associated bonuses. I have not used the micro method (for estimating), as I never had the need for it, as I go after work that has some repetition, and in this way I can use the macro approach for estimating purposes.

I donít currently do the incentive system since I downsized due to changes in the market - not because the incentive program didnít work. The system I used was to break down the tasks in larger jobs into smaller targets, usually in about 2 hour increments. I have some Erp software that I tracked the labor with. The bonuses started at 90% of the target time and increased up to 145% of the target time. The estimated time was based on the fact that only 80% of their time was productive. So 90% was really above the estimated time by 10%.

Now I expect a reply from the original questioner stating how wrong my system is. Since it worked and it always made me more money than it cost me, I say it is workable and created a very productive shop. Iím open to other ideas, however.

To the original questioner: you mentioned how you changed your door process that involved inserting some molding in a mitered door before you cut the miters, and saved a lot of time by sequencing the processes differently. Well, we are up against that sort of thing a lot more than I would like to admit, and it sometimes makes prediction more difficult. However we arrive at it, we have to gain control of the processes and the people performing them, and prediction is a big part of that. Considering this, it is harder to come up with a system.

From contributor W:
I am on the fence as far as profit sharing goes. I realize that paying people by the hour does little to provide incentive to improve, but profit sharing is mainly directed at certain employees, and those are the ones who are not motivated to improve their lives by improving their workplace. We have good employees who work steady and with care. They do so because they know that by doing so, they will improve their wages and advance their careers. We have other employees who putz around. They do so because they believe things should be handed to them regardless of the effort given. In a sense, we have a profit sharing plan. We reward the good employees with steady wage increases and advancement, and we eventually weed out the bad employees.

That being said, I do believe that there needs to be a clear link between productivity and reward, so the questioner has some great ideas to provide this link. The difference is that I provide this reward only to employees who are deserving, and I decide who is deserving and who isnít. A profit sharing plan rewards employees regardless of their productivity as long as the group is productive, with the hope that good employees will put pressure on bad employees to improve. A bad employee will think that the owner is benefiting off the backs of the employees regardless of profit sharing plan. I think the most important goal is to increase the good employee/bad employee ratio. Itís a constant struggle, which is totally dependant on the employee market. We try to provide other incentives to increase morale, like training, bi-weekly barbecues at lunchtime, etc.

I do not want to criticize the questioner, who gives some great advice, and who obviously has created a system that works great for him and possibly others with a similar business. I am looking at it closely to see if it might fit for me. I do want to play the devil's advocate and question the whole concept, because as I said, Iím not sold.

From contributor P:
If you reward production, you get production. If you reward non-production, you get non-production. Government workers get paid the same regardless of their production, consequently the game becomes to do as little possible for the same pay.

Individual accountability. The "Team" concept doesn't hold water. Go to Home Depot and observe how motivated they are to help customers. They are, however, very social. Also, bonusing on the group basis creates a disincentive for good workers, while some workers are not going to be motivated, no matter what. The good workers don't want to carry dead weight.

Accountability is a two way street. Your employees hold you accountable every week for their paycheck. I feel that I have to do the same and hold them accountable for what they exchanged for that money.

The accountability needs to be objective. The end of the year bonus doesn't work; the workers come to expect the bonus. They don't have any idea what they did to earn the bonus, so it doesn't motivate them. The reason for the bonus needs to be clear. I think graphs are the way to go regarding this, based on weekly production. The graphs are also a management tool.

You have to have enough prediction so that overtime pay (time and a half) does not become the incentive that overshadows your system. If you do get into that situation, then only use the guys whose graphs are up (the most productive guys) so that no matter what, you reward production.

As an example of how effective this concept can be from another industry... Some friends of mine have a clothing company, and they instituted a plan similar to mine. They were able to reduce their labor cost (70 workers) by about 25%. Another example is when I worked for another cabinet shop as an installer. The installers would work piecework for most installations, but when we did the model homes, we were paid hourly. I would conservatively guess the difference in production was about 50%. In other words, half speed. Or the framers would say that when they were on hourly, they felt like they were on vacation. I recommend a book called "The Game of Work."

From contributor W:
Some excellent points, contributor P. We were on a piecework system that worked great until we hit a hot economy and all our good employees who made a lot of money complained of burnout and left for less paying jobs. We were not able to replace them because we were unable to find highly motivated employees willing to work hard for high wages (if that makes sense), so we had to scrap the idea. We now try to find steady workers who do good work. Less production, but happier employees and higher quality.

Every employee is different. Some are fast but sloppy. Some are slow but do good work. Some are fast but not team players. How can you create a reward system that takes this into account? I think the simplest system is periodic reviews and communication, with raises and perks based on individual performance. This gives you the most flexibility and doesn't lock you into something that may be hard to change later. It also provides incentive for underachievers to improve.

From contributor P:
We used to do the piecework thing also. One problem is that in a shop, it is illegal in the communist state of California. But other than that, it worked fine. Regarding the burnout, I think that prediction is key. However you arrive at it, rewarding production works. But I hear what you are saying.

From the original questioner:
The team concept absolutely does work if you have a great team, which means management has to do its part in giving regular performance reviews, and getting rid of the bad employees. "Slow to hire, quick to fire" is a very useful tool. (We get rid of the deadbeats quickly.)

As far as rewarding the good performing and not the lesser performing employees, this is addressed by both performance reviews (and the resultant raises), and that the $15/hr person makes a 50% higher bonus than the $10/hr person. This is significant. Our Team Leaders get a higher percentage rate, so this jumps their bonus up even more. We still want the $10/hr person to see and work toward the common goal, but no, they do not get the same rewards. To not offer them this incentive really would turn them off instead of on.

A few years ago a consultant told me that I needed to always be interviewing and bringing people into the hiring process. We had about 34 people and I thought he was crazy, but after a few years of waiting until we suddenly realized we were 4 people short, I have come to believe he was right. At 45 people, we certainly do need to always have someone to interview and call in as needed. We hire a lot of entry level people, so regardless of their quality, a lot still leave due to job abandonment, jail, school, poor attendance, etc. - all the things that happen to a typically young worker. Yes, we do go through a few people to find a keeper. But even the ones that don't stay long are typically very good workers.

From contributor P:
This definition is what I'm referring to:
"A team is a small number of people with complementary skills who are committed to a common purpose, performance goals, and approach for which they are mutually accountable." (Katzenbach and Smith, 1993)
"People working together in a committed way to achieve a common goal or mission. The work is interdependent and team members share responsibility and hold themselves accountable for attaining the results." (MIT Information Services and Technology)

The idea that they hold themselves accountable may work if you have agreement between the good workers who are able to keep the bad workers in check. If the bad workers outnumber the good workers, I don't think so. Karl Marx had a similar ideaÖ The Team concept in of itself absolutely does not work. It may work in the above example, but not because of the team concept. I will agree to disagree on that point.

From contributor W:
If you were to take the bonus you give out and divide it amongst your employees, by giving them all a raise based on a number of criteria, like skill level, attendance, team spirit, etc., they would all be making the same wage, and it would cost you the same in wages. Do you believe your overall productivity would go down? And if so, might this not be offset by the caliber of applicants at your door who would be drawn by the higher hourly base wage?

As I said, I am not against using a profit sharing plan, and your plan sounds exceptionally well thought out. I am just playing the devil's advocate.

I am in a particularly hot economy at this time, and as such do not have the luxury of "slow to hire, quick to fire." I might get one qualified applicant every three months walking in the door that I have to hire on the spot. In my case, I believe that advertising a higher base wage is more beneficial than offering a bonus plan.

Nonetheless, I appreciate you posting your information. I would still like to get other opinions on it as there are many established companies large and small that don't offer profit sharing, or a bonus plan. I would be interested in knowing why not.

From contributor C:
Team: a group of people working toward the same goal. In this case production.

Is that a better definition? Whether we call it teamwork or some other term, the ability of the whole shop working together is crucial. As already stated, one bad apple spoils the bunch. Any reward or incentive program must be understood and followed by all. It appears that the questioner has overcome most of the challenges of this type of system. I will hazard it is a full time job to keep running.

The measure of work for custom is a problem for some shops. I would use a job or phase unit that could be taken from the bid/estimate paperwork. Using a nurse's station as a project, breaking down all the components to a time unit, would be overly complicated. A better option would be to use the maximum time allowed for the unit; anything under the preset time would be in the bonus. In other words, if you set a time of 24 shop hours to complete, and the shop completed the project in 18 hours, that would be a win for the production incentive. A word of caution: If the time to beat is consistently low or close, morale will suffer and the desired result will not be met. Likewise, if too much time is allowed, there is no incentive to push.

On the subject of merit pay increase, I never felt good about anything more than a cost of living increase. My main concern was that an employee will step it up when review time is near and can fall back after the raise is earned. I think if a close examination is made on productivity around review time, a pattern will show.

A production based pay hike or bonus is only paid if the work is done. A hard working dedicated employee will reap the benefits of more money while the slacker will not. Hard feelings? Maybe, maybe not. No matter, you are not stuck with a higher pay rate for non-producers. How many times have you been able to hand out a pay cut without trouble?
I do not believe there is one answer for this problem. Dealing with people has too many variables to make it easy.

In closing, if you have good employees that make you money, treat them well and reward their efforts. Bad attitude, non-producers should hit the road as soon as possible. Employees are people too; they will make or break you.

From contributor A:
Contributor B, I really like using sales dollar per hour; the minute is a more refined number. Knowing workable sales dollars per hour makes estimating, budgeting, and pricing cross checks very simple. It's the inverse of production dollars per year/per man to make money.

There are high dollar items that need to be over the norm and low dollar under the norm, but you can't do a lot of under the norm hours or you can't crack the nut. I like the concept of this system. To the original questioner: have you implemented this?

From the original questioner:
Been doing it solid for a year and a half.

Several of you have mentioned how weakened the team is with the bad apples mixed in. This is a fundamental problem that needs to be dealt with for this type of system to work. If several employees have complained about the same person to you, why would you not take corrective action? If the employee is a good employee, but your hot-shot is complaining about them, that's when I say, "That's why you are making a lot more money than them." They smile, stand a little taller, and clearly appreciate the compliment; problem solved.

Contributor W, unlike the cabinet business, I can't hire people with door experience, so they are all entry level. Hiring people at a higher wage rate has backfired badly for us. They never have stayed on. When we start them at a modest pay of $9 - $10, train them, give reviews (and, most often, raises) at 30, 60, 90 days, 6 months, and annually, and see them experience the team work and additional incentive pay, soda and popsicle days, t-shirts, a clean and safe work environment, a crew that laughs together - that's when we have an employee that will stay with us for years and bust their butt every day. It's a package deal. If we had all that, but were racist or sexist or had unsafe machinery or jerks for supervisors, none of this would work.

From contributor Y:
Without question, management has the responsibility to build the team. Our staffs are not eligible for the production bonus until they have completed a three month probationary period. During that time, we complete necessary training, assess their performance and provide feedback, set performance goals and expectations. Poor performers are eliminated before they begin sharing in a bonus that they didnít earn.

Besides the bonus, we complete performance appraisals annually for all staff. Merit increases are measured on individual performance under 5 performance indicators -Productivity, Quality, Adaptability (cross training), Reliability (attendance management) and Team Work. We have noted improvement in all areas and a reduction of staff turnover.

From contributor P:
Regarding assigning an overall time to something like a reception desk, that would be fine if you knew how long it was going to take. If you don't know the time required, being able to use the time required for components is useful. Complicated or not, you have to have prediction.

Whether you try to assign time to the tasks or money to the tasks, it is much subject to over- and under-estimation. If you have a product, this is easy.

Regarding the team thing... When a team is winning, they have high morale. In an organization, morale comes from production. The esprit de corps, mutual respect, etc. does not come from some sort of team spirit.

From contributor W:
When we were on piecework, we used to assign time to custom work and the problem we had was that our time estimate was always high to account for contingencies. If nothing went wrong, we could make good money, but a lot of times, things go wrong, resulting in substantial drop in profit or even loss. The company needs to make a higher margin on custom work to account for this. If the employees are losing their bonus because of one measurement that was misinterpreted, it can make for bad feelings.

The questioner has developed a great system. I believe that in order to apply it to custom work, it would take a lot more thought, and also a lot of administration.

One other point I would like to make... Our piecework system fell apart because I was too slow to react to a changing economy. Any system needs to be fluid, constantly monitored, and constantly evolving.