Just as there are many tools and techniques that PR practitioners can utilize to begin to measure PR outputs, there also are many that can be used to measure PR outcomes. Some of those most frequently relied on include surveys (of all types), focus groups, before-and-after polls, ethnographic studies (relying on observation, participation, and/or role playing techniques), and experimental and quasi-experimental research designs.
Best practices for both qualitative and quantitative research are covered in the Advertising Research Foundation’s two documents: “Guidelines for the Public Use of Market and Opinion Research” and the ARF Guidelines Handbook: A Compendium of Guidelines to Good Advertising, Marketing and Media Research Practice. Both are available from the Advertising Research Foundation, 641 Lexington Avenue, New York, NY 10022.
Ultimately, one intent of public relations is to inform and persuade key target audience groups regarding topics and issues that are of importance to a given organization, with the hope that this will lead those publics to act in a certain way. Usually, this involves four different types of outcome measures:Awareness and Comprehension Measurements, Recall and Retention Measurements, Attitude and Preference Measurements, and Behavior Measurements.
1. Awareness and Comprehension Measurements
The usual starting point for any PR outcome measurement is to determine whether target audience groups actually received the messages directed at them, paid attention to them, and understood the messages.
Obviously, if one is introducing a new product or concept to the marketplace for the first time — one that has never been seen or discussed before — it is reasonable to assume that prior to public relations and/or related communication activities being launched, that familiarity and awareness levels would be at zero. However, many organizations have established some type of “presence” in the marketplace and, thus, it is important to obtain benchmark data against which to measure any possible changes in awareness and/or comprehension levels.
Measuring awareness and comprehension levels requires some type of primary research with representatives of key target audience groups.
It is important to keep in mind that Qualitative Research (e.g., focus groups, one-on-one depth interviews, convenience polling) is usually open-ended, free response, and unstructured in format; generally relies on nonrandom samples; and is rarely “projectable” to larger audiences.
Quantitative Research (e.g., telephone, mail, mall, fax, and e-mail polls), on the other hand, although it may contain some open-ended questions, is far more apt to involve the use of closed-ended, forced choice question that are highly structured in format, generally relies on random samples, and usually is “projectable” to larger audiences.
To determine whether there have been any changes at all in audience awareness and comprehension levels usually requires some type of comparative studies—that is, either a before and after survey to measure possible change from one period of time to another, or some type of “test” and “control” group study, in which one segment of a target audience group is deliberately exposed to a given message or concept and a second segment is not, with research conducted with both groups to determine if one segment is now better informed regarding the issues than the other.
2. Recall and Retention Measurements
Traditionally, advertising practitioners have paid much more attention to recall and retention measurement than have those in the public relations field.
It is quite common in advertising, after a series of ads have appeared either in the print or the broadcast media, for research to be fielded to determine whether or not those individuals to whom the ad messages have been targeted actually recall those messages on both an unaided and aided basis. Similarly, several weeks after the ads have run, follow-up studies are often fielded to determine if those in the target audience group have retained any of the key themes, concepts, and messages that were contained in the original advertising copy.
Although recall and retention studies have not been done that frequently by public relations practitioners, they clearly are an important form of outcome measurement that ought to be seriously considered by PR professionals.
Various data collection techniques can be used when conducting such studies, including telephone, face-to-face, mail, mall, e-mail, and fax polling.
When conducting such studies, it is extremely important that those individuals fielding the project clearly differentiate between messages that are disseminated via PR techniques (e.g., through stories in the media, by work of mouth, at a special event, through a speech, etc.) from those that are disseminated via paid advertising or through marketing promotional efforts. For example, it is never enough to simply report that someone claims they read, heard, or saw a particular item; it is more important to determine whether that individual can determine if the item in question happened to be a news story that appeared in editorial form or was a paid message that someone placed through advertising. Very often, it is difficult for the “average” consumer to differentiate between the two.
3. Attitude and Preference Measurements
When it comes to seeking to measure the overall impact or effectiveness of a particular public relations program or activity, assessing individuals’ opinions, attitudes, and preferences become extremely important measures of possible outcomes.
It needs to be kept in mind that “opinion research” generally measures what people say about something; that is, their verbal expressions or spoken or written points of view. “Attitude research,” on the other hand, is far deeper and more complex. Usually, attitude research measures not only what people say about something, but also what they know and think (their mental or cognitive predispositions), what they feel (their emotions), and how they are inclined to act (their motivational or drive tendencies).
“Opinion research” is easier to do because one can usually obtain the information desired in a very direct fashion just by asking a few question. “Attitude research,” however, is far harder and often more expensive to carry out because the information desired often has to be collected in an indirect fashion. For example, one can easily measure people’s stated positions on racial and/or ethnic prejudice by simply asking one or several direct questions. However, actually determining whether someone is in actual fact racially and/or ethnically prejudiced usually would necessitate asking a series of indirect questions aimed at obtaining a better understanding of people’s cognitions, feelings, and motivational or drive tendencies regarding that topic or issue.
Preference implies that an individual is or will be making a choice, which means that preference measurement, more often than not, ought to include some alternatives, either competitive or perceived competitive products or organizations. To determine the impact of public relations preference outcomes usually necessitates some type of audience exposure to specific public relations outputs (e.g., an article, a white paper, a speech, or participation in an activity or event), with research then carried out to determine the overall likelihood of people preferring one product, service, or organization to another.
Usually, opinion, attitude, and preference measurement projects involve interviews not only with those in the public at large, but also with special target audience groups, such as those in the media, business leaders, academicians, security analysts, and portfolio managers, those in the health,medical, and scientific community, government officials, and representatives of civic, cultural, and service organizations. Opinion, attitude, and preference measurement research can be carried out in many different ways, through focus groups, through qualitative and quantitative surveys, and even through panels.
4. Behavior Measurements
The ultimate test of effectiveness—the highest outcome measure possible — is whether the behavior of the target audience has changed, at least to some degree, as a result of the public relations program or activity.
For most media relations programs, if you have changed the behavior of the editor and/or reporter so that what he or she writes primarily reflects an organization’s key messages, then that organization has achieved a measure of behavior change.
However, measuring behavior is hard because it is often difficult to prove cause-and-effect relationships. The more specific the desired outcome and the more focused the PR program or activity that relates to that hoped for end result, the easier it is to measure PR behavior change. For example, if the intent of a public relations program or activity is to raise more funds for a nonprofit institution and if one can show after the campaign has been concluded that there has, indeed, been increased funding, then one can begin to surmise that the PR activity had a role to play in the behavior change. Or, to give another example: for measuring the effectiveness of a public affairs or government relations program targeted at legislators or regulators, the desired outcome — more often than not — would not only be to get legislators or regulators to change their views, but more importantly to have those legislators and regulators either pass or implement a new set of laws or regulations that reflect the aims of the campaign. Behavior change requires someone to act differently than they have in the past.
More often than not, measuring behavior change requires a broad array of data collection tools and techniques, among them before-and after surveys, research utilizing ethnographic techniques (e.g., observation, participation, and role playing), the utilization of experimental and quasi-experimental research designs, and studies that rely on multivariate analyses and sophisticated statistical applications and processes).
What is crucial to bear in mind in connection with PR outcome behavior measurement studies is that measuring correlations — that is, the associations or relationships that might exist between two variables — is relatively easy. Measuring causation — that is, seeking to prove that X was the reason that Y happened — is extremely difficult. Often, there are too many intervening variables that need to be taken into consideration.
Those doing PR outcome behavior-measurement studies need to keep in mind these three requirements that need to exist in order to support or document that some activity or event caused something to happen: 1) cause must always precede the effect in time; 2) there needs to be a relationship between the two variables under study; and 3) the observed relationship between the two variables cannot be explained away as being due to the influence of some third variable that possibly caused both of them.
The key to effective behavior measurement is a sound, well thought out, reliable, and valid research concept and design. Researchers doing such studies need to make sure that study or test conditions or responses are relevant to the situation to which the findings are supposed to related, and also clearly demonstrate that the analysis and conclusions that are reached are indeed supported and documented by the fieldwork and data collection that was carried out.