This section gives you some quick ideas on what works and does not work.
Over the past two decades there has been a fierce war over whether people were more motivated by cash or non-cash rewards such as recognition, affiliation, and status.
Many researchers with socialist political leanings sought to prove that paying incentive bonuses actually reduced employees’ willingness to work hard (that is, reduce intrinsic motivation). However, there is now a large and compelling body of research that says cash incentive compensation works (duh!).
That said, there is also a huge body of work that says non-cash rewards are important as well. The research says that cash incentives will work well without non-cash rewards, but that they will work much, much better with non-cash rewards.
Even when you tip well, you get better service from a waitress when you are nice to her. Non-cash rewards matter as well.
So give employees psychic rewards as well. Offer recognition to high performers. Create a culture that reinforces the company’s standards. Avoid work-place environments that are hostile to getting work done.
9.2. Don’t pay too little
We mentioned the trap of paying too little previously. You need a swing of at least 20% of base to really get people’s attention. 40% or more is better.
9.3. Don’t pay too much
Some corporate boards seem to believe that $30 million bonuses are required to motivate executives. While this may be required for particularly rich, jaded executives, this is not required for 99.99% of most executives and staff.
In particular, you seldom need bonuses above 50-100% of salary. It is OK and probably better to structure plans to reduce payouts at extreme levels.
Just as absolute power corrupts absolutely, extreme bonuses have the potential to motivate employees to cut corners and cheat. Most people know how the high-level executives at Enron cheated the system to make money (for example, the Enron CFO received $30 million from partnerships that did business with Enron). High and low-level staff can both cheat and game the system.
Perhaps the most famous example of cheating by low-level employees happened at Sears Auto Centers. Sears provided commissions to shop employees that recommended repairs. State officials in many states charged that Sears’s auto-service centers with defrauding customers by systematically recommending unnecessary repairs.
Sear conceded that the incentive compensation program and sales goals created an environment where “mistakes” occurred. The problem here was that the incentive system focused entirely on short-term quantitative measures that did not include value to customer, qualitative goals, or the long-term interests of Sears. Sears replaced the flawed program with a compensation scheme that paid for long-term customer satisfaction, not profits that day. Sears fixed the problem.
As discussed next, qualitative bonus components also help people to focus on the right things.
If the bonus is 100% determined by a formula, you increase the odds of the three-wish problem (unintended consequences). For example, you want the ability to penalize units that hit production targets by ignoring safety and killing a few employees along the way.
Rather than imagining all the potential ways the formula could screw up and adding factors for each one (for example, how each employee death affect payouts), you should keep 30% of the payout reserved for people who exemplify the company standards. These are usually well articulated and include the factors that allow you to correct for unintended consequences.
If you leave much more than 40% qualitative, you will lose the employee’s trust that you will objectively reward exceptional performance. If you put 20% or less qualitative, people will ignore the company’s values and game the payout formula for large, unintended consequences.
Some people say that you should increase incentive bonuses while cutting base pay. The base pay cuts are said to “fund” the increased upside.
Staff hates this approach. The morale hit is huge. They take a certain hit in pay for an uncertain upside. For most employees, the math here is very unfavorable.
The long-term sustainability of these plans is also doubtful. The first time the plan does not payout, the adverse morale is so serious that management scraps the plan.
Finally, cutting base pay is not necessary to fund the upside. As mentioned previously, the effect of incentives in a properly designed plan is to generate the cash that will pay for the plan. The way to increase the bonus-to-base ratio is to increase the bonus, not decrease the base.
Linking bonuses to beating budget forecasts results mainly in low-balled budget forecasts. This is unfortunate because 70% of companies set compensation targets with bonuses tied to budget forecasts. Thus 70% of companies push people to low-ball budgets and expectations.
You get higher performance from people by setting goals that are different from budget. Budgets are for minimum levels of performance.
Compensation goals should instead key off:
- Beating competitor performance
- Continuing improvements
- Giving staff a percentage of the incremental value they create (more on this later)
An incentive plan that does not pay out is a failure. Staff morale will plummet. You will scrap a plan that does not play out anything.
Management is often tempted to set stretch goals like “We will give every employee $20,000 if our stock price triples.” This sort of compensation plan is DOA (Dead on Arrival).
Instead, choose a pay plan that pays out a little to start and then gradually stairsteps it’s way to bigger payouts. Keep the stair-steps low and frequent. If the plan does not pay out in year one (or better yet within the first 6 months), the odds of failure go way up.
Remember that we are after performance improvement.
You have to set the plan so that big payouts only come with big improvements.
Obviously, there is a tension between setting hurdles too high versus too low. You need to make goals achievable and help people achieve them. Of course, goals cannot be current performance. Goals must be a stretch.
9.9. Shorter payout periods are better than annual
Remember the success that Five Guys had with weekly bonuses. All behavioral research indicates that more frequent feedback and payout is better.
How fast would your dog learn a new trick if you only gave treats for good performance once a year?
The closer you provide positive feedback to the behavior you want, the better.
Higher frequency gives you the opportunity to announce more frequent breakthroughs, and thus provide more motivation.
Remember that best performing companies had a plan for everybody, even the secretaries who did not directly affect production.
Plans often fail because of equity issues. Sometimes companies scrap plans because employees excluded from the plan complain.
Everyone who contributes to success deserves part of the reward. Nevertheless, the same plan seldom works for all. Different people need different plans, different performance measures, and different upside potential.
In summary, try to be as inclusive as possible.
9.11. Have only 2-4 factors driving the pay package
As mentioned previously, if you give people too many targets to accomplish, they often get none of them done well. People can only dynamically manage only two to four things.
Avoid what some call “Balanced Scorecards”. These are shopping lists of things for employees to watch. Employees cannot dynamically manage the 20-30 factors on these scorecards.
Good pay plans base bonuses on 2-4 factors.
If you put all your salespeople on 100% commission, you will lose some good performers when either they have a run of bad luck or the company has a bad quarter of product releases. If making the goal looks hopeless and the goal period looks far away, people will get discouraged and leave.
While it is good to put incentives in, do not go crazy. Winner-take-all systems lead to “run-and-gun” cultures that often go awry.
Do not get rid of all base salaries. Try to prevent situations where employees with temporary bad performance feel hopeless for a long period. Keep goal periods short so a good employee with accidental bad performance can always look forward to next period.
As mentioned above, employees need the connection between their effort and pay. If you pay all farmers on a big farm the same bonus, you have just duplicated the Soviet collectivist system.
Team-based bonuses are OK. Just try to keep the definition of team as small as possible. Try to keep the bonus group size well under 100. Try not to compensate the whole factory with one bonus system.
9.14. There are a variety of EWYK incentive plans to use. All have their uses, but some argue that quotas work best.
The same analysis of 45 compensations studies found that there are four general types of tangible incentive program types or schemes:
- Quota-Based Programs– incentives are given for meeting or exceeding a performance goal.
- Piece-Rate Incentive Programs — for increasing rates of performance or doing more of something.
- Tournament Programs– where individuals and/or teams compete with each other for incentives.
- Fixed-Rate Incentives– salary-based compensation.
Which works best?
- Quota-based incentive programs were the most effective.
- Piece-rate incentive programs were significantly less effective than quota-based plans. But piece-rate programs are significantly more effective than either tournament schemes or fixed-rate incentives.
- Tournament and fixed-rate incentives were equal to each other in terms of their impact on performance
Using the company’s shares as incentives is a very good idea for top executives in all companies and all employees of small firms. For these situations, employees can see how their actions affect the value. For them, shares are value-linked and motivating.
They are much less good as a motivating tool in other situations, particularly mid-level or lower employees in large firms. Few of these employees can directly affect value and fewer still understand the linkages. For these employees it is more motivating to use other tools and other rewards.
Studies have shown that while giving equity and ownership to employees can increase productivity. However, productivity only increases when employees participate in decision-making. If employees do not participate in decision-making, productivity does not increase.
It still can make good sense to offer stock and stock options to employees more broadly. It creates a sense of community and can be an efficient way to compensate employees without using cash.
In fact, there is an argument that all employees should be paid 10% or more of their compensation in restricted stock and stock options since it is a very cash-efficient way for employers to pay staff. Just do not pretend that it is especially motivating.
Over-use of options can also be a danger. Studies have shown that over-use of options seems to lead management to “swing for the fences” with risky acquisitions, excessive cost cuts to make earnings targets, and other bad behaviors.
Finally, you need to make sure that options pay out only when the person creates value. Many option programs are just get-rich-schemes for management. The following graphic shows how just stock market gyrations can cause problems:
Some people like to use indexed options, a type of option that only pays out when the company outperforms competitors. This approach works well when you want to avoid compensating management for just a general lift in all stocks.
These approaches are tricky as well. In a recent project for a gambling and hotel company, management rejected our suggestion for payouts only when they create value. They insisted on an approach where the company paid executives on how well they did versus the competition.
After company implemented the plan, the gambling company did OK with some minor improvements in profits. However, their competitors suddenly announced massive accounting write-offs. Suddenly, management looked terrific “by comparison”. The company paid huge payouts under the “pay versus competition” approach, even though they did not actually create much value.
Net net, options and pay-versus-competition performance plans are tricky. You need sophisticated analysis of many future scenarios to make sure they work. In addition, you need to limit payouts to avoid windfalls for management just hanging around. Finally, boards sometimes need to listen to their consultants more (executives can try to tilt the game towards payouts in appropriate situations).
9.16. Use tangible, business unit-based outputs as performance measures, not behaviors or company-wide financial measures
Pay systems generally use a mix of three types of performance measures:
- Behavior measures (accidents, absenteeism, safety inspection, etc.)
- Output measures of business units (production, costs, quality, on-time delivery, customer satisfaction, cycle time, unit profit, unit economic profit, etc.)
- Company-wide financial measures (stock price, company profit, company economic profit, ROA, etc.)
Employees like behavior measures (accidents, absenteeism, safety inspection, etc.). Employees see these as directly under their control. For them, behavior measures are the clearest, easiest to affect, and the most motivational. However, there is often an uncertain link between better behaviors and better company value.
Managers like company-wide financial measures (stock price, company profit, company economic profit, ROA, etc.). The company-wide financial measures have the surest link to company value. By linking compensation to these measures, they can ensure against large payouts when the company is doing badly. Unfortunately, these measures have very low motivation for employees for several reasons:
- The unreliability of payout (employees can put in high effort and create huge value, but be cheated by a general stock market downturn)
- Lack of understanding (what does getting ROA from 3% to 4% mean?)
Profit share plans tie to company-wide results are overused. Profit-sharing plans are not motivating. Profit sharing is only a way to share the upsides. It is only as motivating as distributing lottery tickets to employees. Profit-sharing plans are not line of sight.
The optimum “Goldilocks” middle is output measures of business units (production, costs, quality, on-time delivery, customer satisfaction, cycle time, unit profit, unit economic profit, etc.). They link to company value and are easy for employees to see and understand.
A good example of a useful output measures is often customer satisfaction scores. Five Guys is a good example of how a company focuses employees on customer satisfaction and drives business success. The secret shopper score clear and transparent. Employees understand and trust the measure. This measure thus works for both the company and the employees.
9.17. If you want to use business unit profits as a measure, use economic profit (defined below) and not operating income
Businesses often over-focus on operating income, without due consideration for taxes and the cost of capital. Let’s illustrate.
Say you owned an apartment building. The building’s operating income is the rent minus the operating costs of the building (the super, maintenance, and utilities). It is nice when operating income is positive (the rent covers operating costs). It is even sweeter when you have enough left over to pay the taxes due on the building and cover your mortgage costs.
Economic profit is the profit left after paying for taxes and the capital you have tied up in the business. The capital charge is usually about 10% of the book value of assets. Think of the capital charge as “covering the mortgage” (it actually is the weighted average cost of the company’s debt costs and its shareholder required return).
By incentivizing people on operating income alone, employees will under focus on capital and taxes.
In one real example, a LEK corporate client recently highly rewarded a business unit executive who doubled his businesses operating income over 5 years, building a multi-billion revenue business. However, during the same period, he invested $15 of capital in his business for every $1 of operating profit created. Under any reasonable cost for capital, the unit did not create any real economic profit or value. The company could have done better by paying down debt, investing the money in public markets, or giving money back to shareholders.
By the way, companies like Disney that look at economic profit well out-perform their competitors. Disney’s stock has quadrupled since 2009, and a good part of the reason is the corporate discipline that economic profit provides.
If you want to use business unit profits as a measure, use economic profit and not operating income.
For most companies, compensation is usually the biggest secret they have.
Why is it a secret? While there are many reasons, the biggest is that compensation is often hard to justify or unfair. It is easier to try to keep it a secret than to explain it. Even if it all leaks out when employees talk to each other, management can refuse to discuss it.
If you cannot fairly tell someone why he or she is paid less than someone else was, the pay system is at best arbitrary and at worst unfair.
I am not saying that you need to advertise everyone’s bonus in the local paper. I am saying that you should be able to discuss the reasons why two people in similar jobs are paid differently.
If you find it hard to have this discussion, you do not have an objective Eat-What-You-Kill system.
9.19. Be careful and model out results of your pay plan in advance
“You get what you pay for!” As noted early, people will usually work hard to make money. But beware. People are clever, and people will find ways to game the compensation system.
Executives may try to get short-term profits by cutting payroll. Executives may try to goose this quarter’s results by shifting the timing of payments.
Further, there may be things that happen that were unforeseen. Perhaps inflation may skyrocket in a South American division, and the South American executives suddenly are paid more than other executives because inflation triggered payouts.
Put simply, there are a huge number of different scenarios to consider when planning an “Eat What You Kill” incentive system. You need somebody to model out all potential scenarios, and show the probability of excessive payouts or pay other pay problems.
We now will turn to how to implement a successful “Eat What You Kill” pay system that drives firm performance and employee pay. We will discuss in the next section.