Businesses large and small are scurrying to collect customer data. But to what end? Data collection is a costly exercise. And data is only useful if it can be harnessed to provide actionable or informative insights that improve the business. If analysed correctly, customer data presents the potential to uncover sales trends, prescribe marketing strategies and predict long-term outcomes. Whether you’re an analyst at the coalface or an executive decision maker, these strategies will help you to get the most from customer data and communicate findings within your organisation.

Form an Aim and Hypothesis – What Question Are You Trying to Answer?

Any data collection effort should begin by deciding the aim of the investigation and suggest a hypothesis about the results. 

For example, when looking at sales data, your aim could answer the question, ‘what is the dominant demographic of customers that purchase product A?’ You could hypothesise that the answer would be ‘first-home buyers between the ages of 20 and 30.’

Your findings will inform an action or decision, for example, increase marketing spend towards a more relevant market that will generate better return than other demographics.

With any dataset, spend some time up front to decide what kind of result you would like to see. This will help to avoid unnecessary and potentially fruitless analysis. The results will then usually suggest an action to take, depending on the validity of the hypothesis.

Expand Your Scope of Investigation By Considering the Bigger Picture

Always take into account the wider context surrounding your investigation. If you are an analyst, someone else may be asking the question without communicating their aim. If you are a decision maker, you may be looking to test something particular while having a broader question in mind. Issues may arise where the decision maker is not the analyst, and so clear communication of the scope and purpose of the analysis is crucial.

Consider the following scenario: a decision maker wants to compare the performance of product A and B, but only asks the analyst for data on product A. The analyst obliges. The decision maker comes back and asks the same for product B, and the analyst repeats accordingly. The decision maker then compares the two on their own. Had the decision maker stated their original aim, or had the analyst expanded the analysis themselves, the latter could have analysed all the information at once and compared the two products. This would produce a higher quality result to inform the decision maker on what they truly needed to know.

In practice, the decision maker is often not as familiar with the data as the analyst. Decision makers can get more meaningful results by taking the following steps. 

Clarify the purpose.

This helps the analyst understand the true aim of the investigation and conduct the analysis in a way that provides a more useful answer. It also gives analysts the autonomy to decide the best way to analyse the data and perhaps use data that the decision maker would not have thought to use.

Seek knowledge and perspective from other team members.

Data analysts understand the limitations and intricacies of the data; frontline operational staff are ‘on-the-ground’ and deal with the practical aspects of data every day, and team leaders understand their team members’ interests best. Drawing on the strengths of other teams members can help better realise the potential of the data.

Ask for a broad set of results.

This avoids unnecessary to-and-fro. For example, if you are looking to identify the ‘top 10 clients’, consider making a list of all clients.

Expanding the scope of the investigation reduces the risk of tangential analyses, helps to anticipate and address downstream questions and more effectively responds to the intent of the investigation.

Use Benchmarks to Provide Context for Results

A number without context is meaningless. Benchmarking provides a frame of reference for evaluating a number’s meaning and for determining the best course of action to pursue as a result. Common benchmarks include:

  • related items or groups (Product A versus Product B),
  • previous time periods (2016 versus 2017),
  • competitors (my product versus my competitor’s similar product), and
  • industry standards (my business’s revenue versus my sector’s average revenue).

Segmenting data is also useful for comparing groups, as long as the segmentation is meaningful to your business. For example, a wholesaler planning its marketing strategy would find it more useful to segment customers by industry than demographics.

Benchmarking your data helps decision makers make sense of the information and decide what type of responsive action to take.

Automate Repeated Analysis and Reports

Automation is a sustainable and scalable solution that frees up time and reduces errors. It also makes the reports more accessible to decision makers.

First, identify pieces of analysis or reports that you need to update regularly. Common examples are performance reports, end-of-quarter bonus calculations and monthly revenue. Next, invest to build a framework that can be easily updated. Here are some common methods I use (with links to further information):

  • If you regularly pull data from databases into Excel, connect those databases directly with Excel (instructions). This way, you can refresh data within Excel instead of having to pull it from the database.
  • If you have a dataset where new information comes in all the time, such as a list of purchases, use Excel ‘tables’ (instructions). Tables save a lot of time since formulas that rely on them dynamically account for extra rows of data.
  • If you use external data sources such as Google AdWords or WordPress, you can use their APIs. This is fairly technical and may be more suited to developers, but allows you to create live dashboards and reports. If possible, consult a developer who is familiar with your databases.

Communicate with Decision Makers in a Top-Down Manner

Presenting the information a decision maker needs in a top-down structure is best, for example: 

  • Start with the conclusion or recommendation.
  • Follow by summarising the key results.
  • Append details, footnotes, caveats and the method of analysis at the end.

When communicating with a decision maker, it is good practice for the analyst to suggest a course of action. The decision maker can then simply agree, disagree or open a discussion. Suggesting a decision is a natural next step that also helps the decision maker keep decision fatigue at bay.

In summarising and presenting results, graphs and other visualisations should be the tool of choice for numerical data.

They are easier, quicker and more meaningful to consume than raw numbers and show trends and anomalies immediately. Flag all anomalies, explain them if possible, and suggest explanations for other issues the decision maker may be likely to point out.

Key Takeaways

Data is a powerful driver of business improvement. And effective data analysis is always focused on finding practical ways to improve the business. 

It’s important to clarify the aim of the analysis and have an idea of the output to achieve this goal. To make results valuable to decision makers, analysts should consider the full intent of the investigation and provide their opinion on what the results suggest. Benchmarking results and automating reports are also ways of making analytics more useful. 

How do you capitalise on your business’ data? Let us know on Twitter or LinkedIn.   

Jonathan Ling
If you would like further information on any of the topics mentioned in this article, please get in touch using the form on this page.
Would you like to get in touch with Jonathan about this topic, or ask us any other question? Please fill out the form below to send Jonathan a message!