Reprinted from
Intellectual Property Today
August, 2003

10 Frequently Asked Questions
About Intellectual Property Litigation Surveys

By James T. Berger

James T. Berger is principal of Evanston (IL)-based Market Strategies, a strategic marketing consulting firm.  He also does extensive consulting and expert witness work for intellectual property attorneys.  Among his areas of expertise is both the development of surveys and the critiquing of adversarial surveys.  He is a faculty member at Roosevelt University he is also is a often published freelance magazine writer.  He can be contacted at: j-berger@iplsn.com

The use of a survey can be a powerful tool in proving or disproving intellectual property cases - especially those involving likelihood of confusion and trade dress issues.  However, doing a survey can be an expensive and time-consuming proposition.  Moreover, if not done properly it can do more harm than good in that an adversary can refute either or both the findings and methodology.

As one who has been retained in a number of cases to perform and refute surveys, I am continually asked a number of questions.  Following are 10 of the most frequently asked questions with respect to surveys:

How do I know if I need a survey?

If you are trying to prove or disprove confusion in the marketplace, there is a very good chance that a survey can be helpful to your case.  If properly executed and using the right procedures and methodologies, a judge and/or jury might very be persuaded by your findings.  

In many ways, a survey is very similar to a scientific experiment.  There is a hypothesis and the survey is the experiment devised to test that hypothesis. If an experiment can be created that shows the hypothesis either has or does not have support, and such an experiment can be replicated at different times and different places than the survey can probably provide powerful evidence.  Clearly, in litigation where one side performs a survey and the other side does not, the side that has performed the survey usually holds a major competitive  advantage.  

In a case in which I was involved where a U.S. power tool company believed one of its registered product designs was being infringed by a foreign competitor who entered the U.S. market with a similar looking product in terms of paint and appearance, a survey was all that was needed to win the case. Trades people were brought into a research facility where they were shown the competing product and asked who made it.  A significant number named the U.S. manufacturer thus proving the confusion.  

When should one NOT do a survey?

A survey is like any other form of market research in that it is a function of time and money.  Make sure you give yourself enough time to do what is needed and your client can afford the cost.  Sometimes surveys can be very costly.  One rule of thumb is determine the potential exposure if you were to lose the case and then try to figure the cost of the survey including the cost of retaining an expert and the fees the law firm will have to charge in working with the expert.  If all the costs related to doing the survey exceed the client's exposure from an unfavorable ruling, it is unwise to undertake the survey.  On the other hand, if the exposure is great and the survey cost is considerably less, it probably would pay to undertake the survey.  There also is the possibility that the outcome of the survey is not black or white but some shade of gray.  In this case where the results of the survey come out mixed, it might have some value but perhaps not much.

If the other side has done a survey, if nothing else the law firm should hire an expert to critique that survey.  This can be accomplished for a relatively low cost.  If the opponent's survey is poorly done, it then can  work in your favor.  Every survey - no matter how well done - is subject to criticism.

A law firm, which retained me, had a case involving a drug product and filed an action alleging confusion.  To test this hypothesis required a survey that would have had to be administered over the Internet.  Before the survey was executed, the opponent offered a very attractive settlement.  In fact, the cost of the settlement was less than cost of doing the survey.  Needless to say, the survey was never done even though it was ready to go.

How much does a survey cost?

There is no easy answer to this question.  However, the best way to estimate costs is to determine exactly what is needed and then to determine  the most cost-effective way of doing a survey that delivers what is needed.  Surveys can cost as little as a few thousand dollars but can also go into six figures.

In a likelihood of confusion case in which I was involved, an attorney  was seeking a temporary restraining order and presented his case before a federal judge.  The judge bought all the arguments but one.  The judge was not certain that the “channels of trade” were the same for each product and that the customers of the retailer in question also shopped at the same stores where the competing product was sold.  A very fast, inexpensive survey was executed.  A researcher stood outside the store and asked customers where else they shopped.  It turned out the customers shopped in a variety of stores and proved that both retailers were in the same channel of trade.   Once the judge heard testimony of these findings, the TRO was quickly granted.  The research was done in less than a week at a cost of around $4,000 including all the expert's fees.

How fast can a survey be done?

This all depends on what needs to be proved.  In some cases, telephone surveys can be done very quickly and are a function of how many callers are used as well as the likelihood of getting qualified and cooperating respondents. .  In mall intercept surveys where is not necessary to select any specific demographic, this also can be done relatively quickly depending on the numbers and the time it takes to administer the survey.  Surveys that tend to take longer are those that require recruiting specific types of people usually by occupation.  These people need to be scheduled to come into a research facility. In addition, there is usually a cost connected with each person who comes in - no matter how short the interview is.  

A good rule of thumb for most surveys is a month to six weeks.

Is there any such thing as an irrefutable survey?

NO. There is no such thing as perfect research and there is no such thing as totally useless research.  Every survey lies somewhere in between.  No matter how well a survey is done; no matter how solid is the methodology, the survey and its findings are ALWAYS subject to some form of criticism.

In a seminar before the Minnesota State Bar Assoc., I was given a survey that won a case and asked to critique it.  It was splendid research and proved beyond any doubt its hypothesis.  Yet, as good as it was, I was able to find a number of points that subjected this survey to question.

Should one ALWAYS try to critique an opponent's survey?

YES.  Since there is no such thing as a perfect survey, any survey no matter how well done is subject to criticism.  In fact, some surveys are so poorly done that they serve as a liability rather than an asset.  It's like a counter puncher in boxing.  Such a boxer can do more damage by responding to his opponent's attacks than by initiating the attack himself.

In a case where I was involved, the opponent produced a survey that I was asked to critique.  The survey was so poorly done in terms of the way the exhibits were produced, the methodology and other aspects, the mere production of such a survey hurt the opponent's case.  When we researched the research company who performed the survey, we discovered that  previous work produced by the research firm had been repudiated in court rulings. This served to strengthen our case even more.     

Are two survey experts working together better than one expert?

ABSOLUTELY NOT.    Not only will it cost you more money, it might damage your case.  A good attorney might be able to get each expert to contradict the other.  Because of the nature of survey work with so many gray areas, an attorney can play one expert against the other.  It's just a bad idea and should be avoided.

Can a survey be done over the Internet?

It's uncharted ground but it certainly can be done.  The Internet is simply an interactive communications device that enables questions to be asked and answered on-line and the completed questionnaire to be submitted electronically.  It can be especially good if those persons that need to be interviewed are not readily available.  Like any other kind of survey research, make sure the methodology is correct and that results are reliable and can be validated.

In a case in which I was involved, it was necessary to interview thoracic heart surgeons.  These highly paid specialists are not about to come to a research facility for $100 to answer questions.  They were reachable by e-mail, however.  So, we created a special Web site questionnaire that not only had the questions but had the exhibits from which they could compare the products in conflict.  Their incentive was a $100 donation in their name to a favorite charity.  The questionnaire was easy to fill out and they could do the questionnaire at their convenience.  Once completed all they would have had to do is click submit.  It should have worked but we never got a chance to try it because the case settled.

Why is it so important to pretest a questionnaire?

A survey should be pretested for the same reasons that a packaged goods manufacturer would test market a product.  Before one wastes valuable time and money, it is wise to make sure those who are being surveyed understand the questions and provide the answers you seek.  For example, on a sunny day with the temperatures in the mid-70s, if you asked people?  “What kind of a day is it?”  If the answer you sought was “a sunny day,” some people might say, “a nice day.”  Others may say,  “a pleasant day,” “a warm day,”  “a cool day,” etc.  The correct question would have been?  “Is the sun shining today?”  While this is an oversimplification, these are some of the problems a researcher will encounter in a survey and testing the questionnaire enables those problems to be eliminated on the front end.

Should one try to find an expert who promises to produce the results your client needs?

No reputable expert can guarantee results.  What you should seek is an expert who understands what you want the survey to address and who can create a correct and honest survey to address that issue. The last thing you want or need is an expert who manipulates the results.  Such a practice will undoubtedly be revealed in deposition and/or another expert will find the flaws when doing a critique.  Before you hire an expert, check references and review any cases in which that expert has testified.  If that survey expert has been repudiated, it may reflect badly on your case.