Many companies have created customer insights teams to better understand their customers in a data driven approach. Despite the good intention, the results often vary hugely. Here are a few factors that impact what your customer insights team can deliver:
- Which function does your team align to? Often the customer insights team is closely aligned either with Marketing or Sales. For a marketing focused team, your research is probably more focused on understanding target customers, whereas for a sales focused team, you are more likely to focus on preventing churns
- What data is available? Most companies have some forms of customer data, which is very useful for conducting existing customer analysis. External data is more difficult to get, but if you are able to obtain it, not only you could enrich your customer analysis, you can also more accurately identify and track external market factors
- Do you have data science capability? In other words, can you deliver predictive analytics? With some data science capability, you could create forecast models that give an extra edge to improve your business performance
Wherever your team is in terms of capability, we have listed below the top 6 top customer insight analysis that you should strive to deliver.
Part 1 — For new business
Market Research: Traditional market research focuses on figures and statistics to look for trends in data. But such research often just scratches the surface, and ignores to find out “why” of a purchase. With social media on a surge, people are more open and comfortable to share their opinions public. Social media has become a great source to find out people’s true intention. Market research focused on “social listening” can help your team to understand what your customers are really thinking, and track trigger events that may change their behaviour. Here is an example of Zity, a mobility sharing startup cleverly used Twitter to build up its brand images.
Funnel Analysis: Do you know how your customers convert at each touch point with your company to eventually purchase your product? For e-commerce business, tracking customer footprint has become much easier now thanks to web analytics. You can easily gather event log data to understand how customers interact with your website. Analyzing your customer funnel could help you improve and optimize your website design to gain more business. Here is an example of how ING used funnel analysis to optimize their website navigation.
Customer Segmentation: From mass marketing to personalized marketing, one key obstacle is to create meaningful segmentation of customers that make sense for your product. The traditional segmentation based on demographic is no longer sufficient, as the world becomes more connected and information shared more widely. You may find a financial analyst who works in Wall Street in New York have much in common with a hardware engineer who lives in the suburb of Shenzhen in China. Behaviour based customer segmentation works much better to help you understand your customer. Here is an example of how a political polling company Metroscopia use customer segmentation to understand different types of voter behaviours.
Part 2 — For existing business:
Attrition Analysis: This is a classic analysis for any customer insights team. It is highly popular because the data is easily obtainable from your own CRM, and the business impact is significant, especially for subscription business. There are already a few very highly developed methodology to analyze customer churn. But whether your attrition analysis is useful or not depends greatly on if you have the right data science techniques to create prediction models. Ideally not only you can predict which customers are more likely to churn, you should also be able to explain the main drivers that caused the attrition, and create appropriate business strategies to react on it. Here is an an attrition analysis based on IBM internal data that well explained attrition reason.
Product Basket Analysis: You probably find Amazon product recommendation or Netflix movie recommendation very useful. What about applying the same analysis to find out what other products you should recommend for your customers? Product basket analysis is a perfect tool to help you identify cross-sell and up-sell opportunities from your existing customers, and design product bundle or promotion strategy. Here is an example of how Lola Market, an online supermarket identified their most purchased product baskets.
Customer Voice Analysis: Tracking direct customer feedback is critical to help you business improve your product and service. Customer feedback comes in many formats: NPS survey, customer call transcript, online review, etc. Finding a way to analyze this data (whether it is structured or unstructured) can help you quickly identify customer pain points and narrow down your focus. Here is an example of a product review analysis of Alexa from Amazon.
A capable customer insight team should be able to deliver majority of the analytics mentioned above. The key challenge has always been how to execute your analysis while balancing your investment. If you are not able to deliver these analysis yourself, you need to consider whether it makes sense for you to outsource some of these activities, or investment in technology to equip your team.
One easy solution is to outsource your analytics. You could find quite a few consultancies that specialize in data analytics. The negative side of using them is they tend to be expensive, and it doesn’t help you develop your own capability in the long term.
We believe that investing in technology to equip your own team is a better option. With the proliferation of SaaS product, there are a number of software tools to choose from that are competitively priced. By getting their hands dirty, your team will also be able to improve their capability to deliver more complex projects eventually.
One important factor in selecting the right tool is to make sure the software is easy to use. Unless you have a team that is fluent in data science programming languages such as R and Python, you would want a self-service software that is intuitive and self-explanatory. Following that, you also want a tool that is flexible enough so that you could use it to support majority of your analysis. After all, investing in a product that you will only use for few project a year is not a great return on investment.
If you want to know more about self-service data analytics software, check out graphext.com to learn more.
