The lessons are in from the first proxy season that included the Dodd-Frank legislation’s "say on pay" provision. Say on pay, which gives shareholders an advisory vote on executive compensation programs, was a response to the public outcry about oversize executive pay packages following the 2008 financial crisis. Here is what we’ve learned:

Lesson 1. Proxy advisers hold all the cards. Because say-on-pay voting falls largely in the hands of investment funds ill-equipped to analyze thousands of proxy statements, the funds accepted the analyses of proxy-advisory firms, most notably Institutional Shareholder Services and to a lesser extent, Glass-Lewis. Compensation committees largely followed the advisory firms’ recommendations, effectively allowing them to dominate say-on-pay voting. The firms will continue to wield great influence until more investment funds begin to conduct analyses in-house, an innovation I would encourage. So far, few have moved to do this.

Compensation committees generally resent the power of proxy advisers. (I should point out that as a compensation specialist, my firm doesn’t compete with such companies.) I urge compensation committee members to better understand what drives proxy advisers’ voting recommendations because directors will be responding to them for years.

Lesson 2. Eliminate questionable pay practices. The vast majority—some 88 percent—of companies got positive recommendations from ISS on their executive compensation programs. For those that did not, one of two key reasons for negative recommendations stemmed from "problematic pay practices" such as tax gross-ups (which offset the tax burdens of reimbursed expenses) or excessive perquisites, supplemental executive retirement benefits, and severance.

To prevent problems next year, committees should review pay practices now. If your company has problematic practices with no strong business rationale or legal obligation to continue them, they should be eliminated. A preemptive strike will prevent one year’s say-on-pay votes against the compensation program from turning into votes against compensation committee members the next year. Bear in mind that most of the practices that shareholders and proxy advisers commonly oppose do not belong in a well-designed performance-oriented executive compensation program in the first place.

Lesson 3. Stand your ground on "pay for performance." ISS uses a flawed methodology by which it calculates that the company’s total shareholder return fell below median, compared to industry peers, for the latest one-year and three-year periods, while the chief executive officer’s year-over-year total pay went down by "less than a marginal amount." ISS defines total pay as salary, bonuses, all "other compensation" (including perks), pension accruals/contributions, and value of long-term incentive grants—which is by far the largest single element of CEO pay. The flaw is that ISS fails to take into account the value of actual compensation earned and realized from long-term grants. In many cases, these grants have not delivered any real value and never will unless some specified performance goal is met, particularly for stock options and performance shares. ISS looks at the potential value of long-term incentive grants, based on accounting rules applied at the time of grant, not the real payout that may or may not occur later, based on performance.

For example, I am familiar with a large company that had what ISS deemed a "pay for performance disconnect" because the CEO’s total pay increased from 2009 to 2010, due to higher values of long-term equity grants, while the company’s one- and three-year relative total shareholder returns were below median. Over the past five years, this company granted equity to the CEO valued at approximately $50 million in the form of performance shares and time-vested stock options. Since the company and its CEO failed to achieve performance goals and stock price was flat to down, performance shares were unearned and most of the stock options remain under water. Virtually no pay has been received by this CEO because of lagging performance. How can it be argued, therefore, that this company’s pay has no connection to performance? Compensation committees should prepare themselves to explain to shareholders exactly what has happened when such disconnects occur.

Lesson 4. Go directly to shareholders. Complaints to proxy advisers about calculation errors, black-box methodologies, and irrelevant peer groups will likely fall on deaf ears. Should a compensation committee believe it has legitimate complaints about the proxy advisers’ voting recommendations, it should register them directly with the investment funds that hold its shares. At one large company whose compensation committee I advised, management made more than 200 calls to investors in order to get a narrow say-on-pay win. Compensation committees need a systematic shareholder-contact plan so they can quickly make their own case.

Lesson 5. Make proxy statements clear. The proxy-statement narrative describing executive compensation programs is largely incomprehensible because of legal rules—a primary reason investment funds need proxy advisers. It’s dangerous to vote on something you don’t understand, especially when you have fiduciary responsibility. Many companies that were successful in 2011 say-on-pay votes put short and simple "executive summaries" at the front of their narratives. The best of these serve to concisely connect pay decisions to the underlying business rationale, including relevant performance. All companies should do this in 2012. They should write the executive summaries as if few will read or understand all of the legalese that follows.

Will say on pay turn into an annual point of contention between shareholders and management in years to come? No one knows. Compensation committees need to prepare for such a possibility. The 2011 proxy season was a trial run. You can expect institutional investors to step up their involvement in 2012 and beyond.

George Paulin is Chairman and CEO of Frederic W. Cook & Co., Inc. Founded in 1973, the firm focuses on independent evaluation of performance-based compensation programs that help companies attract and retain key employees, motivate and reward them for improved performance, and align their interests with long-term shareowners.