Report: 84% of marketing leaders use predictive analytics, but struggle with data-driven decisions

Artificial intelligence (AI) holds great promise for today’s businesses, especially for marketing teams that need to anticipate customer interests and behaviors to achieve their goals. Despite the growing availability of AI-powered technologies, many marketers are still in the early stages of formulating their AI strategies.

There is a lot of interest in the potential of AI-based predictive analytics, but marketing teams face several challenges in fully adopting this technology. With no universal roadmap for integrating data science into marketing, several approaches have been developed, with varying degrees of success.

Pecan AIs The report Predictive Analytics in Marketing Survey reflects this complex situation and provides important insights for marketing teams and business leaders who are tackling challenges with AI, no matter where they are on the adoption curve.

Key Findings — Integrating Predictive AI Analytics

While many companies emphasize the importance of consumer data in various areas, from predicting future purchases to customer churn, the reality is that more than 4 out of 5 marketing executives report problems making data-driven decisions despite all the consumer data they have at their disposal. . The same number of respondents (84%) say their ability to predict consumer behavior feels like guesswork.

An overwhelming majority (95%) of companies are now integrating AI-powered predictive analytics into their marketing strategy, including 44% who said they have fully integrated it into their strategy. Of companies that have fully integrated predictive AI analytics into their marketing strategy, 90% report that it is difficult for them to make day-to-day data-driven decisions.

Marketing and data science face unique challenges when trying to work together. As a result, data projects get stuck. The study provides insights into their struggles, including:

• 38% of respondents say data isn’t updated fast enough to be valuable.
• 35% say it takes too long to build the models.
• 42% say data scientists are overwhelmed and don’t have time to fulfill requests.
• 40% say those building the models don’t understand the marketing goals.
• 37% of respondents indicate that wrong or partial data is used to build models.

Methodology

The Pecan Predictive Analytics in Marketing Survey was conducted by Wakefield Research on 250 US marketing executives with a minimum seniority of director. These executives work at B2C companies that use predictive analytics and have a minimum annual revenue of $100 million. Participants responded to an email invitation and an online survey between September 13 and 21, 2022.

Read the full report from pecan.