State of Analytics In Customer Programs: Customer Loyalty Focus, Machine Learning Adoption and the Data Science Skill Gap

Buffer

A new customer analytics survey of 80+ companies provides a look into the state of analytics in customer programs. Only 32% of respondents are satisfied with their company’s use of analytics to create a competitive advantage. The use of multiple survey methods is the most common practice across companies (80% of companies). The use of machine learning was one of the least adopted practice in customer programs (38% of companies). Customer professionals lack proficiency in, or access to, three important data science skills: programming, mathematics, statistics. Customer professionals said their biggest roadblock was the inability of translating customer insights into business operations.

Download the free white paper, “State of Analytics in Customer Programs,” by clicking the image.

We studied 88 companies that have a formal customer program (e.g., customer experience (46%), customer success (30%), other (24%)) to determine the extent to which they incorporate analytics practices into their program. Most respondents were from the US (86%) and worked in a B2B (50%) or B2C/B2B (42%) company. About half (48%) of respondents were from companies with 5000 or more employees; 40% of respondents were from companies with 50,000 or more employees. Most respondents worked in the IT (29%), Healthcare and Medical (12%), Financial Services (8%) and Aerospace and Defense (8%) industries.

Data were collected from January to March 2017. Potential respondents were invited to complete a survey on customer analytics best practices via the researcher’s blog and social media profiles.

Creating a Competitive Advantage with Analytics

Companies continually look for ways to outperform their competitors. One way they are trying to get ahead is through the application of analytics on their data. Researchers have found that top-performing businesses were twice as likely to use analytics to guide future strategies and guide day-to-day operations compared to their low-performing counterparts.

Researchers from MIT and SAS showed that analytical-leading companies (those that use analytics to create a competitive advantage) adopted analytics practices to a greater degree than analytical-lagging companies.

In the current study, we asked respondents to indicate how satisfied they were with their company’s use of analytics to create a competitive advantage.

We found that 32% of respondents were satisfied with their company’s use of analytics to create a competitive advantage. 41% were neither satisfied nor dissatisfied about their company’s use of analytics while 27% were dissatisfied.

Analytics Primarily Used to Improve Loyalty and Reduce Costs

We asked respondents to indicate the primary areas in which their company uses analytics. On average, companies use analytics in three areas.

Figure 1. Primary areas in which company uses analytics. (click image to enlarge)

The top three areas in which companies use analytics include improving customer loyalty, reducing enterprise costs and improving resource allocation (see Figure 1).

The least popular areas in which companies use analytics include accelerating development of new products, making real-time decisions and identifying new markets.

Analytics Practices in Customer-Centric Programs

We wanted to understand the extent to which companies adopted 25 specific analytics practices across six customer program components (Strategy/Governance, Business Integration, Method, Reporting, Advanced Research).

Figure 2. Adoption rate of analytics practices in customer programs. (click image to enlarge)

The results revealed that, on average, companies adopted 12 analytics practices. For each of the 25 practices, adoption rate varied widely (see Figure 2). The most widely adopted analytics practices included using multiple survey methods (80% of respondents indicated their program adopted this practice), having an executive champion of the program (74%) and integrating the program into business technology and processes (74%).

The least adopted analytics practices included using machine learning for insights (38%), using social media to determine customer sentiment (35%) and incentivizing employees using customer metrics (33%).

It appears that fewer than half of customer programs adopt analytics practices related to advanced research activities (e.g., conduct in-depth studies and employ customer data platforms).

Data Science Skills Gap

Figure 3. Customer professionals proficiency in five data science skills. (click image to enlarge)

We asked customer professionals to indicate their proficiency in five data science skills as well as their access to teammates who are experts in those same five skills. While customer professionals were competent in two skills (i.e., Business knowledge and Technology), they lacked adequate proficiency in Programming, Mathematics and Statistics (see Figure 3).

Additionally, these customer professionals do not have teammates who can fill those quantitative skills (see Figure 4). Like the customer professionals’ themselves, their teammates lack expertise in Programming, Mathematics and Statistics. One positive finding is that 66% of these customer professionals said that they have access to a data scientists/analyst within the company to help them make sense of their data (see Figure 1).

Figure 4. Percent of Customer Professionals who have access to a team member who is an expert in 5 data science skills. (click image to enlarge)

Roadblocks in Customer Programs

I asked respondents to indicate their biggest roadblocks that hinder their customer program to improve customer loyalty (e.g., recommendations, up/cross-sell). Respondents indicated that, on average, they are experiencing two big roadblocks in their customer programs.

The most popular roadblock mentioned by these customer professionals was the lack of integration of customer insights into business operations (59%) (see Figure 5). Only 14% of respondents said their program does not provide customer insights. So, while companies are getting insights from their data, they are experiencing difficulty putting those insights into operations.

Figure 5. Biggest roadblocks that hinder customer programs to improve customer loyalty. (click image to enlarge)

Summary and Conclusions

We asked customer professionals about their analytics practices within their company. We found that adoption rates of customer analytics practices in customer programs varied widely. While some practices were adopted by over 70% of companies, a few were adopted by less than 40% of companies. The use of machine learning in customer programs was only adopted by 38% of the companies.

Customer programs generate a lot of data. Consider the customer data from customer surveys, CRM systems, Web analytics and support systems, to name a few. It’s clear that customer programs need to include efforts to improve how companies govern, collect and analyze that data. This study showed that customer programs appear to lack the analytical rigor needed to extract insights from their customer data. Less than half of the companies surveyed used a customer data platform to deliver automated insights and even fewer companies report using machine learning to gain insights.

Customer professionals lacked proficiency in quantitative skills, limiting their capability of extracting insights from their data. It’s important to note that, in our prior study of data professionals, proficiency in statistics skills were a top driver of analytics project success, even for job roles that were not primarily quantitative in nature. Specifically, we found that Business Managers who were highly proficient in statistics and statistical thinking were more satisfied with their work than Business Managers who were less proficient. I suspect that customer professionals would benefit from having deeper knowledge of statistics. Customer programs now process a lot of data, and I believe that customer pros need to keep pace and develop basic quantitative skills if they want to put that data to use. Customer pros who are better equipped at mining and visualizing their data will not only generate better insights, but will ask better questions.

We also found that the biggest roadblock in customer programs revolves around translating insights into operational practice. Fortunately, many service providers are addressing this problem by helping companies set up a seamless link between the service provider’s insight engine (typically machine learning capabilities) and their customers’ operational systems. For example, these machine learning service providers are able to integrate data silos and extract customer insights via machine learning. These insights (in quantitative form) can then be fed back into the company’s marketing automation or customer success systems to build workflows that can automatically trigger a targeted marketing campaigns or notify a customer success manager of an at-risk account.

A good step toward improving your customer program starts with understanding the specific practices that define your program and incorporating practices that will move your program and company forward. In our next paper to accompany this study, we compare three types of companies: analytical leaders, analytical practitioners and analytical laggards (identified earlier in this report) to understand if there are differences in how they adopt analytics into their customer programs. This type of insight will help us identify analytics practices that truly differentiate analytical leaders from laggards.