bell signifying sophisticated customer analytics

How sophisticated is your approach to customer analytics?

Conjura has worked with all types and sizes of consumer business. We’ve worked with businesses who are just starting on their customer analytics journey as well as businesses who have large analytics teams. We’ve seen a huge range of sophistication when it comes to approaching how data is collected, used and acted upon.

Our experience has allowed us to identify and surface 16 questions across 4 categories that can signal your level of sophistication in customer analytics. In this article, we’ve outlined these questions and how they can help you identify the level of sophistication of your approach.

Data Structure for Customer Analytics

Customer analytics follows the Garbage In, Garbage Out (GIGO) principle, so understanding how you are collecting and structuring your data is critical for success. Ask yourself these questions to identify the data gaps in your organization.

1. Have steps been taken to improve the quality of data you are collecting?

It doesn’t matter how smart your analysts are, if they work with bad data, they will get bad insights which results in wasted time and resources.

To ensure data analysts are using reliable data, businesses must focus on data collection from all sources including marketing platforms, payment tools, internal databases and tools like Google Analytics. This means using standardized methods for creating and naming campaigns in marketing platforms or reducing the amount of human-inputted forms that are in use (could you use a drop-down menu instead, for example?).

Check out this article outlining some issues that can affect the quality of data in Google Analytics, as well as some simple solutions to make sure your data is high quality.

2. Can you extract data from relevant tools in an automated fashion?

All analytics projects require data. In customer analytics, it is likely that relevant data is stored in a cloud-based platform like Google Analytics, Google Ads or Facebook.

In all cases, data must be extracted from these platforms in order to be used by the analyst. Most tools have a manual export functionality. But this is usually very time consuming and often the tool will limit the amount of data that can be exported in one go.

Sophisticated businesses will extract this sort of data via an API. APIs usually have fewer restrictive limits on data extraction and can be automated, but to do this generally requires knowledge of a coding language like Python.

3. Has a data warehouse been created that includes all datasets related to the customer?

To get a comprehensive and objective view of performance or customer behaviour, it is not appropriate to use only one data set from one source. Effective and insightful customer analytics uses multiple data sets.

Conjura believes that the best insights are derived from combining data from marketing platforms (such as Facebook or Google Ads), web analytics platforms (like Google Analytics), customer data and payments information.

The most sophisticated businesses will ingest and store this data in a single, easy-to-use warehouse. The result is that analysts will spend less time data wrangling and more time generating the insights that can change your business.

4. Does a foreign key exist to join data sets together?

Foreign keys are unique identifiers that exist across different data sets that enable them to be integrated as part of an analytics project. They are fundamental to an analyst’s ability to combine multiple data sets and deliver comprehensive and insightful analyses.

5. Have you automated the integration of these data sets?

Not only is it important to use and store multiple data sets in a single, easy-to-use location, the most sophisticated businesses also automate the integration of these data sets.

Integration refers to how the data can be used in conjunction or “joined together“. This is by far the hardest part when it comes to supercharging your customer analytics capability. Most data sources are not designed to talk to one another, so it takes a lot of data manipulation and processing before data can be joined.

This is also the part of an analytics project that takes so long. But by automating this process, a huge amount of time can be saved for every analytics project, thereby increasing the throughput of your analytics team and increasing the volume of insights generated. It also reduces the chances of human error affecting the accuracy of analytics.

6. Do you integrate data at different granularities?

Not only is it important to integrate all your data sets, in order to speed up insight generation and ensure future proofing of your infrastructure, data must be integrated at different granularities.

Granularity refers to the level of detail that your data contains. For example, Google Ads will allow you to extract data at a keyword level, a country level or a campaign level. Each level may be used by a different analysis, but it is important to have the flexibility built into your data infrastructure to handle them all.

This is a very important, but also very difficult step to get right. Given that your business may by using dozens of data sources, there may be hundreds of levels of granularity required to give your infrastructure the flexibility its needs to support highly sophisticated customer analytics.

7. Is there a data dictionary for each source in the schema?

Businesses are in constant flux, whether from a data or a personnel perspective. With that in mind, clear and comprehensive documentation helps to ensure the longevity and usefulness of your data infrastructure. The documentation should cover each aspect we have outlined above to reduce key man risk and to speed the upskilling of new hires.

Using Data for Customer Analytics

Data and analytics can, and should, support decision making across the whole organization. In this section, we have outlined some key questions to ask to understand whether your business is getting the most from your data.

8. Is your business performance reporting automated?

The most effective companies can make data-driven decisions very quickly. These organisations can assess performance, spot opportunities and act accordingly. This is supported by automated business reporting that is always fresh, accurate and available for use.

9. Can your business quickly and easily calculate cost per acquisition, customer lifetime value and retention rate?

These are the foundational KPIs for any consumer business. Sophisticated companies calculate these daily and will report them to the entire business.

10. Does your analytics capability support customer acquisition and the monetization of existing customers?

Given the data available, a sophisticated analytical capability will support the development of strategies to both acquire new customers and expand or retain existing ones. By failing to utilise analytics for these purposes, companies are at a disadvantage to their more analytically savvy competitors.

11.Does your analytical team focus solely on historical insights?

Conjura endorses data foresight in addition to data insight! Generating insights on historical trends is very valuable. However, a sophisticated analytics capability will also use their data to make predictions on the future value of their customers, on their stock requirements, on churn likelihood and so on. This sort of analytics enables a business to be proactive, rather than reactive.

12. Are your predictive models deployed into production?

A predictive model can be very informative at a single point in time. However, the insights gained from single-use models become out-of-date very quickly as data and customer behaviours change. The most informative predictive models are deployed into production, meaning they are automated to run every day and they are regularly maintained and tweaked to ensure that their insights remain relevant over time

Technology for Customer Analytics

13. Is your organization using the most innovative data infrastructure tools and processes?

The amount of data available to companies is massive and is growing rapidly. Legacy systems are not designed to deal with the scale of the data available and will limit your analysts from unlocking the value of your data.

By using modern data infrastructure such as those offered by Amazon, Microsoft or Google, companies are putting themselves in a good place to future proof their capability to continue to generate value from their data. The open source movement has also transformed data analytics. With the availability of libraries in Python more and more functions are possible with less code provided you are using the best coding language. Similarly with data transformation the emergence of frameworks such as DBT has massively decreased the reliance on third party software tools.

14. Do your BI tools enable business users to investigate data on the fly?

In addition to analytics teams, many organisations now have business users who are comfortable with data and are happy to perform minor investigations into data. There are several tools that enable users to do this (Tableau, Looker, PowerBI) as well as serving the analytics team and hosting business reporting.

These tools enable the whole business to make data-driven decisions and ensure data is being effectively used.

People & Organisation

15. Does your organization provide an attractive career for analytics talent?

Data analysts are highly sought after in today’s business environment, and for good reason. Data analytics, if used properly, has the power to transform your decision making and therefore the business. The best analytics talent wants to work on interesting and important business problems and wants to work with the best technology.

16. Are there no significant skill gaps within the analytics professionals in your organisation?

A data analyst requires a range of skills to be effective. These are not just technical skills like programming or statistics, analysts must have the ability to apply these skills to real business problems. Additionally, analysts must be able to communicate complex ideas and insights to non-technical business users. Without the correct mix of these skills, it is extremely difficult to effectively use customer analytics in your organisation.

Asking these questions of your organization

We’ve outlined how these questions can help you determine your level of sophistication in customer analytics. They should give you a very clear idea of the gaps in your organization and why it is important to plug those gaps.

If you would like to discuss how Conjura can help you improve your level of sophistication with customer analytics, please don’t hesitate to reach out to foresight@conjura.com.