Easy as Pie.

When you know how, linking communication to organizational performance is

For communicators to move from tacticians to strategic partners with management, they must demonstrate how communication affects organizational performance and show which programs have the most significant return on investment.

But here's the rub: Traditional survey analysis may lead you to implement tactics that do not advance your organization's business goals.

Much of the research conducted today provides an analysis of the strengths and weaknesses of communication and stops there, leaving it up to you to determine which weaknesses to focus on and which strengths to leverage, making assumptions about how the actions you take affect the performance of the company.

Note the operative word "assumption." The purpose of research should be to decrease the assumptions you have to make. But because traditional research processes do not tell you about the impact of communication on the organization, you are forced to make assumptions because you do not know for certain which programs are most effective in building trust, employee loyalty and commitment, or even in contributing to bottom-line performance.

Now, thanks to advances in statistical analysis procedures, it's possible to demonstrate quantitatively the effect that communication has on organizational performance -- significantly reducing the assumptions in your plan of action. By measuring the effect of communication, you can predict how changes in communication programs may affect the overall organization.

To measure the effect of communication on performance, it is necessary to measure organizational performance. How is this defined? There can be any number of measures of the effectiveness of a company. Ideally, it would be a bottom-line measure, but other types of measurement are acceptable. See Figure 1 below for some examples.

EIGHT STEPS TO AFFECT RESEARCH

Here is an eight-step process to measure the effect of your communication:

1. Clearly define the reason for conducting the research. What do you want to accomplish as a result? Picture yourself at the end of the project, reading the report: what kind of information do you need? How is it helping you make decisions about what to strengthen? What to change?

2. Develop a research method. Should you use focus groups or a survey? Or both?

3. Conduct an internal assessment of your communication programs and activities. Before going to employees, conduct your own analysis of the strengths and weaknesses of your communication programs, from the employees' perspective.

4. Develop a hypothesis for how communication might affect your organization. How does communication affect people's understanding of the business goals and how they contribute? How does communication affect the company's bottom line? This will serve as your model for measuring impact.

5. Conduct the research.

6. Analyze the results. To move beyond traditional research methods and into effect analysis, use statistical tools such as analysis of variance, correlation analysis, multiple regression and structural equation modeling (these are described below).

7. Develop and implement a strategic plan of action. This is the most critical step of your research. Failure to take action will reduce your credibility and contribute to a break down in trust within the organization.

8. Measure your performance against the plan. Within two years, conduct follow-up research to benchmark your progress.

ANALYZING THE RESULTS

The rest of this article will focus on Step 6 above, describing what it means to measure the effect of communication though analysis of survey data.

The following exhibit shows the various methods of survey analysis, ranked according to how complex the method is (in terms of understanding statistics) and according to the extent to which the methodology measures effect.

TRADITIONAL ANALYSIS

Frequency analysis and demographic tabulations are common to traditional research methods. But neither one of these provides a measure of effect -- leaving it up to you to make assumptions about the strengths and weaknesses on which to focus.

FREQUENCY ANALYSIS simply provides the percentage of respondents who checked each response option to each question. For example: 60 percent agree with Question 1, 20 percent are neutral and 20 percent disagree. This tells you the strengths and weaknesses, but it does not tell you which areas of communication have the greatest impact.

DEMOGRAPHIC TABULATIONS do the same thing as a frequency analysis, but the tabulations provide the responses to each survey question according to different groups of employees. For example: Management Respondents: 70 percent agree with Question 1, 20 percent are neutral and 10 percent disagree. Hourly Employee Respondents: 50 percent agree with Question 1, 20 percent are neutral and 30 percent disagree.

ADVANCED ANALYSIS

Beyond the traditional research methods, more advanced statistical analysis of the data consists of these processes:

ANALYSIS OF VARIANCE identifies statistically significant differences between various demographic groups on the survey. This is useful for identifying meaningful differences in opinions between groups of employees, such as management and non-management. Yet analysis of variance does not provide a measure of impact.

CORRELATION ANALYSIS measures the extent to which two survey questions "move together." For example, if two questions are perfectly correlated, then a 1-unit change in one item is expected to correspond with a 1-unit change in the second item. Correlation analysis does provide a measure of impact, but it is limited in that it can only provide estimates for two items at a time, and these estimates can be difficult to interpret.

MULTIPLE REGRESSION measures the effect a group of survey questions (called the "independent variables") has on a single survey question (or an average of a group of questions), called the "dependent variable." Regression analysis does provide a quantifiable measure of impact, but the limitations of multiple regression are that you can have only one dependent variable and it is difficult to examine the indirect influences that the variables have on each other.

STRUCTURAL EQUATION MODELING uses advanced statistical analysis processes to calculate the impact of multiple survey items on each other. The process is more advanced than multiple regression because you can have more than one dependent variable, it allows for the measurement of indirect effects, and it takes into consideration the interrelationships of all of the variables.

Sounds complex? Well, yes... but the rewards can be significant as you use the results to build a strategic plan of action.

BECOMING A STRATEGIC PARTNER

You may have seen the results of recent studies about the effect of communication on organizations. While these studies provide great support for the role of communication within companies today, using advanced statistical analysis with your own research results gives you the added power of being able to demonstrate the quantitative effect that communication within your own organization has on performance.

Demonstrating how communication affects performance and presenting a business case for investments in specific communication programs will help move the communication function from tacticians to strategic partners.

John Williams is president of Joe Williams Communications, Inc., a 16-year-old communication, research, training and consulting firm. You may reach him by e-mail at [e-mail], or visit the web site at www.JWCom .com

MEASURES OF EFFECTIVENESS 

BOTTOM-LINE MEASURES

Financial performance

Employee productivity

Customer service

Quality

Safety

Turnover/retention

And more...

WORK ENVIRONMENT MEASURES

Level of trust

Employee commitment & alignment 

Understanding of business goals

Understanding of individual roles

Overall work environment

Job satisfaction

And more...