Nick Mountain | 15th June 2018

Key Steps in CRM Data Analysis

It is no secret nowadays, that data plays an important role in the decision-making process. The wider the use of CRM in a company, the more valuable insights it can provide. Any well-structured CRM analysis can answer:

1. Who are your most loyal customers?

2. What marketing channels are generating the most number of subscribers?

3. What are the conversion rates of the subscription process on the website?

4. What are the best/worst performing campaigns and how do they compare with historical trends?

4. What content is the most effective?

A CRM system stores a bunch of datasets that are invaluable sources of information regarding a company’s customers, sales and marketing operations. This data just needs to be appropriately analysed, but it is extremely important that these analyses are set up as correctly and optimally as possible from the beginning. Otherwise, you can be faced with critical data issues and rework. Therefore, I would like to share some high-level steps in CRM data analysis exercise.

Step 1: Investigating CRM Data

When you gain access to a CRM system, firstly, you should understand what is stored in each dataset, each worksheet and the definition of each field, then evaluate the coverage of important data fields (the percentage of records that are filled in). Another important step is to make sure you understand the pitfalls of the data: the number and causes of corrupted records, missing values, duplicates, and why it happens. Because CRM data is mostly manually entered, believe me, unfortunately, it will contain errors.

Step 2: Merging the Data

Group the data together to gain easy access in one place (for example, build a Single Customer View). You must structure the data around the unique identifier, e.g. customer account ID, individual email address, segment level etc. When you define that, it will be your key to link all the different data sources together to get any information you want to analyse (your imagination is only limited by the available data).

Step 3: Set-up Objectives

As I’ve already said, your imagination is only limited by the available data. However, if your CRM dataset consists of hundreds of different fields, it is important that you limit your focus strictly to the analyses that meet your objectives. Therefore, you should begin firstly by defining the types of conclusions that you want to get. Try to stay focused on the evidence that you are trying to get and the hypothesis you are trying to prove or disprove, so that you can evaluate the results and avoid getting bogged down in analysis.

Step 4: Visual Presentation

Nowadays, there is no lack of BI tools for visual data analysis to help you get your point across, but they should be consistent with what you are going to achieve and the tone you want to project. Try to consider your audience — what assumptions do they have and what approach do you want to take in delivering the story to them. We use one of the leading BI software platforms, Tableau. It provides great capabilities that go beyond MS Reporting Services and Excel-based reporting, allowing users to analyse data extensively and easily. The product does let users create interactive dashboards “when any visual display or report is selected and placed in it via the drag and drop interface,” company officials explain, “letting users create ad hoc pipeline analyses, sales trends, won/loss reports and others, as well as standard business performance summaries and dashboards”.

With Tableau, you will be able to do the following:

• Customer segmentation (for example, the people who are most and least likely to repurchase a product).

• Profitability analysis (which customers are the most loyal or brought the highest profit).

• Pareto analysis (what percent of your buyers generated x% of revenue).

• Event monitoring (for example, when a customer reaches a goal of open x many campaigns).

• Market basket analysis (for example, buyers who bought one product are more likely to buy x product).

• Predictive modelling (forecast based on historical customer behaviour).

• Data mapping (in case your customer records have geo information, you can easily plot them onto the map to see where they live, how far they are from a specific point etc.).

At Winners, we are specialists in CRM data analysis. If you’d like to know more about Winners’ work in this area, or have any questions concerning CRM data analysis, please contact us.

Key Steps in CRM Data Analysis

Share this article: