Predicting consumer behavior is the key to business success
Businesses can grow and remain competitive if they can predict what consumers are most interested in. For example, by checking how their customers react to certain offers or incentives. Predicting consumer behavior is a core responsibility for modern marketers. Market research can help reveal customer intentions, but penetrating the veil that protects actual motivations from close scrutiny can be a tall order for even the best research project. With the advent of big data, marketers are accessing increasingly more sophisticated predictive and analytic tools to forecast consumer behavior. However, certain old school predictive techniques can be equally effective. Let’s examine them.
Problems have solutions
Scott Anthony identified three solutions-oriented steps in predicting consumer behavior in The Little Black Book of Innovation: How It Works, How to Use It. First, marketers should get to the context — the hints that dictate how customers respond to questions.
Second, marketers should be alert about how customers use their products. A.G. Lafley, CEO Emeritus of Procter and Gamble, told the story of women consistently telling P&G researchers how much they liked the new Tide detergent packaging but used sharp objects to open the box because they did not want to damage their nails.
Third, marketers should focus on non-buyers. Your product may not provide a solution to the problems many non-buyers have, which keeps them out of your franchise. Providing a solution could open up a whole new market segment.
Business workarounds
Big data analytics are likely to be a permanent tool in the toolkit of marketers with deep pockets. As the cost of technology continues its downward spiral, this tool will become increasingly more accessible to small-business operators for predicting consumer behavior. Several available low-tech tools for predicting consumer behavior are easy to execute and could likely produce some very useful data.
Data to predict consumer behavior
Data is providing a competitive edge to digital marketing efforts through the following:
Helps understand target audiences: Since AI analyses massive amounts of complex and deep consumer data, it predicts consumer behavior with seamless ease. These may include users’ interests, focus, demography, price limits etc.
Improves user experience: Customer experience is the most important aspect of any marketing strategy or campaign. AI ensures this by collecting data and deciding which content is the most applicable based on factors like trends, location, historical data, and past behavior. This creates an impression among users that the brand is suited to their needs and demands.
Efficient marketing: Apart from personalizing the customer experience and eliminating guesswork in digital marketing, AI also predicts behavior for new and existing users. As data management platforms gather second- and third-party data, AI helps to acquire information from users across the web rather than from a mere session on the company website. This understanding helps target potential leads instead of focussing on users who are unlikely to convert.
Boosts productivity: Algorithms help automate various draining and repetitive tasks. This enables a company’s human resource to delve into other more important areas of business. This invariably increases productivity while saving on both time and money.
The applications of algorithms have eased company processes in various ways in the past few years which are only expected to better with further advancements. Some such applications already revolutionizing digital marketing include the following.
Examples of customer prediction initiatives:
Biometrics predicting campaigns?
The difference between content that goes viral and content that fails to find an audience depends on a single, critical moment: a person seeing the share button and deciding whether or not to click.
To predict what will happen at this moment would be similar to discovering the holy grail of marketing research. Simply asking people what kind of content they would share doesn’t do a great job of anticipating actual outcomes. However, researchers can utilize physiological markers to measure emotional responses to content. This exists not just in the mind, but also in the body, to better understand what makes someone click share
In a recent study, people were shown a mix of popular and unpopular content. HBR asked participants the usual follow-up questions, such as: “Were you engaged in this content? Do you think you would share this?” and so on. By recording an electrophysiological signal called galvanic skin response (GSR) — a response that is constantly changing in an individual, though rarely noticed — during the study, HBR was able to predict the viral outcome of a piece of content significantly better than was possible via any of the usual survey measures.
Watson marketing
IBM Watson Marketing Insights evaluates customer behaviors to deliver critical customer insights that marketers can use to drive more effective customer interactions.
Watson Marketing Insights analyzes current and past customer behavior based on the customer data that you provide. After you have uploaded your customer data, Watson Marketing Insights applies cognitive analytics and models to develop insights that reveal critical issues that are likely to affect business now and predict changes in how your customers will interact with your brand in the near future.
Food outlets using predictive analytics
A different example is with pizza. A company like Dominos Pizza is using location-based services to activate consumer loyalty offers. This is based on artificial intelligence following and analyzing the habits of customers. The company is much a technology company that happens to distribute pizza as it is a pizza delivery company.
Dominos Pizza, The Daily Sabah reports, also use data to analyze different outlets, reviewing factors like average service time, comparative turnover, monthly sales of branches and whether customers order from the website or mobile applications.
Predictive analytics also work for charities as well or those who specialize in environmental products. Research by Michele Laroche and colleagues has found, using various statistical analyses, the demographic, psychological and behavioral profiles of consumers who are willing to pay more for environmentally friendly products. For instance, this research found that the segment of customers more likely to purchase products marketed as ‘green’ or ‘environmentally friendly’ were females, married and with at least one child living at home. This information is used by marketers to undertake targeted advertising.
Changing marketing strategies based on predictive consumer behavior
A number of dominant and emergent trends help pinpoint how strategy is likely to change in the next few years:
- ROMI: Justifying a return on marketing investment is now often a prerequisite for any campaign.
- Predictive analytics: The ability of data scientists to forecast the likely outcomes of marketing activities is a fast-emerging discipline.
- Behavioral economics: Relatively lower costs of creating prototypes and beta versions are enabling people to implement more test and learn experiments.
- Performance-based remuneration: Paid-for advertising is changing the way that clients remunerate many of their agencies, based on pre-promised outcomes.
Going forward the quality and effectiveness of the strategy will be evaluated on the forecasted consequences, not the retrospective results. Thus the strategy must show improvements.
These changes present a new challenge for the marketing industry. How will we be able to forecast the outcomes of work before it’s even been launched? It will require a different mentality, focused on experimentation, optimization, and analysis.