Skip to main content

Trade Show Forecasting – Using Analytics to Decode the Zeitgeist

Trade Show Forecasting – Using Analytics to Decode the Zeitgeist

These days I can’t read exhibition industry websites, blogs or magazines without coming across the word trend. Used alone, the word trend does nothing more that refer to a general movement in a particular direction. But hot trends, top trends, evolving trends…these do much more – they seek to communicate a powerful sense of future, hinting at an author’s ability to accurately conceptualize the future buying power, purchase plans and attendance habits of trade show attendees.

Indeed, it doesn’t take much thought to appreciate the business value of an event manager conceptualizing a future exhibit design and budget today based on knowing what next year’s attendees will want to experience and be able to spend. In this sense, trends don’t just convey helpful information on general movements – they become prerequisites for business success.

When viewed in isolation, a trend does nothing more that capture the zeitgeist of the current moment. So how do event managers know when an exhibition company’s thought leadership article on trends has accurately evolved into a tool that decodes future uncertainty in a manner that will help them correctly exercise strategic judgement and have confidence that their current business decisions for next year’s show are well informed? Similarly, how do business executives know when to trust a trade show trend analysis? When should we trust someone’s attempt to decode the exhibition zeitgeist?

The answer is that the trend must be analyzed using the forecasting analytical toolset. The key is to move beyond what is happening today and towards how the present will accurately shape the future.

I’m always skeptical when a trade show trend analysis can’t describe the analytical approach (unless the subject matter is obvious). I think most people would expect that the above trend usage examples are based on trend forecasting – a powerful analytical tool to map out future uncertainty and predict how present actions will influence the future. A good forecast influences business decisions by showing a clear map of alternative paths to the possible future: how much to budget, how much to advertise, how to design your booth, where to place your booth, and how to draw in those leads.

To go one step further, how the trend forecast data collection process is structured is just as important as the future prediction the trend forecast provides. Here are a few straightforward questions to ask to determine if a trend forecast can be trusted, and to help you can gain confidence that your strategic decisions today will move you towards the future trend prediction.

Can the trend forecaster clearly explain the method of data capture?

The goal of survey design is to minimize chance errors and increase the likelihood of a reliable analysis. The steps in survey design are straightforward: specify objectives; design and draft the survey; develop instructions; pre-test the survey; edit the survey; develop the statistical sampling plan; execute the survey; collect the data; analyze the data; and report the results.

Underlying this design process is the concept of interview interaction. It is always important to examine the method of interaction the trend forecaster had with survey respondents and the method of interview support the trend forecaster used. Method of interaction looks at the type of interview that was conducted. Did the trend forecaster conduct face-to-face interviews, telephone interviews, or use a self-administered (interviewee administered) questionnaire? Method of interview support looks at whether the survey was paper or computer assisted. Was this is a face-to-face computer assisted interview? Was this a snail mail paper assisted survey? Was this a self-administered web survey? There are many variations here.

While computer surveys are simple and cost effective to implement, the more expensive face-to-face interviews still provide for the best data quality because they allow for the development of extensive interaction with survey participants. The more trend forecasters move away from face-to-face interviews, the more they suppress interaction with survey participants, thus reducing the possibility of obtaining quality survey data. The trend forecaster must be able to explain how the method of data capture was structured to acquire the most reliable survey data.

Can the trend forecaster explain the questions that were asked and how those questions where structured?

Surveys are composed of numerous questions and statements. Survey questions are generally free response or response selection (i.e. check the box, the circle the letter). Response selection questions are attractive because they are easy for respondents to answer and simple for researchers to collect and analyze. Free response questions require more processing to categorize, score, and code. Further, survey participants may think it’s too much work to answer free response questions and may provide short responses which really don’t answer the questions. A trend forecaster relying on a survey full of free responses may not be making the most reliable predictions.

Statements rely on response selection to determine the extent to which survey respondents agree or disagree with a certain perspective (such as the Likert scale). Market researchers know that survey responses contain an element of bias. This can be seen on survey statements where respondents may be reluctant to select the extreme statements and their answers end up reflecting a central tendency.

Further, on both question and statement responses, survey respondents may simply choose to provide what they believe is the “correct” answer or they may be unconsciously providing an answer about a state they aspire to be rather than an accurate answer about the state they are actually in. For example, with the predictive questions and statements that form the basis of trend forecasts there is a risk that survey respondents will introduce a “status quo” answer when faced with a trend that would somehow cast their organizations in a negative light. This “status quo” answer is actually invalid data for the predictive question. Further, the survey respondent may be expressing a personal wish or gut feeling versus truly making a predictive exhibition industry-related opinion (i.e. based on facts and data).

The trend forecaster must be able to speak to how the survey questions were structured to avoid these biases and enhance the capture of reliable information. For example, the bias presented by statements can be minimized by expressing some statements in positive form while other statements are expressed in negative form.

Can the trend forecaster explain how the sample size (including confidence level and margin of error) was determined?

Market researchers are intuitively aware that simply making a survey bigger doesn’t make it better. It doesn’t make sense to ask all trade show attendees (i.e. your leads and contacts) how they feel. It’s a waste of time because adding all those answers to a database won’t make the database more accurate.

It is also true that some surveys are so fraught with mistakes that creating a larger sample will not produce better precision and will only add more bad data. However, that’s not an indictment of the need to determine an appropriate sample size and is only an admission that a large sample size can’t rescue a poorly designed survey with bad questions.

If you had a box of 10,000 widgets in four colors, would you dump out the entire contents to determine how the colors are distributed? Or would it make more sense to sample 300, 500 or 1,000? The key understanding here is the necessity to make a random selection from a coherent group, so you can call meaningful differences “significantly different.” That’s where the trend forecaster needs to be able to clearly communicate the confidence level and margin of error in relation to the results.

Can the trend forecaster clearly explain the type of forecasting approach used?

Historically, the fashion industry has led the way on integrating trend analytics with business strategy. The key analytical tool fashion analysts use to move from mood boards to recommendations on adjusting retail supply chains for “What’s in for Spring” is the regression analysis. Similarly, the regression analysis allows the exhibition trend forecaster to examine relationships between variables and is the key method to determine the trend worthy usefulness of the data. The trend forecaster must be able to explain the regression methodology. What type of regression analysis was used? Did the analysis incorporate time series data? When analyzing time series data, the value for the previous time is normally a good predictor of the value for the current period. However, some regression models are appropriate for time series data, some aren’t. Does the trend forecaster even know this?

Future-Forward the Right Way

All companies want to be future-forward. But exhibition executives should never accept trend forecasts that can’t clearly answer the above questions. Trend forecasts which cannot clearly explain their survey design approach or statistical reasoning won’t get things right and are nothing more than subjective guesses. That’s not the way to make investments in your company’s trade show future.

Trend Tracker Guide

About the Author

Amy is the Global Digital Content Marketing Editor at GES. With a strong background in content marketing, social media, and communications, she is a passionate writer and self-confessed word geek. She is also the founder of a non-profit and a health and wellness online community.

Profile Photo of Amy Kelley