← Back

Spreadsheets for Last Mile Analytics

2024-02-09 // Tom Ward, Software and Analytics Consultant

Go to any data analytics or enterprise software conference and you’ll hear all kinds of disparaging remarks about spreadsheets. They’re still using spreadsheets, Our new BI tool replaces spreadsheets, etc. Spreadsheets are one of the most maligned applications in all of enterprise computing.

This is not a new phenomenon. Fifteen years ago, I worked for a supply chain planning software company and our top ad was centered around the slogan “ditch the spreadsheet.”

However, this reputation is not entirely well deserved. The fact is, spreadsheets have long had a place in businesses for data analysis for good reason; they provide a natural format for data and a flexible way to view and analyze data. In many cases, spreadsheets are the tool most effective at allowing business users to turn data insights into real value - the so-called “last mile analytics” problem. Continue reading or skip to specific sections with the table of contents below:

  1. What is last mile analytics?
  2. Why is last mile analytics challenging with Dashboards and static reports?
  3. A brief history of spreadsheets
  4. The modern data stack
  5. Spreadsheets and the modern data stack
  6. Solving the last mile analytics problem with spreadsheets

What is Last Mile Analytics?

You may have heard the term “last mile analytics” to describe the gap between the output of “frontend” data analytics tools (dashboards, reports, etc.) and actual business insight or value that can be derived from that data.

The term attempts to put a fancy name to the problem of modern business intelligence and reporting tools failing to deliver the expected return on investment of large, expensive data initiatives and failing to deliver meaningful business value to organizations.

Why is Last Mile Analytics Challenging with Dashboards and Static Reports?

Data requirements, like any software requirements are often difficult to define, always changing, and rife with edge cases and caveats. It’s tough for front-line data consumers, like financial analysts, supply chain analysts, marketing analysts, etc. to fully know every aspect of data they will need to make business decisions. This makes rigid tools, like static reports and predefined dashboards, less effective at empowering these users to make decisions based on data.

The static nature of dashboards and reports contributes to the last mile analytics problem because data consumers in analytical roles aren’t empowered to drive their own insights from data and conduct their own analyses.

Spreadsheets are precisely the tool to allow these types of data consumers to turn data into action and business value. Many of these types of data consumers spend a lot of their working day in spreadsheets and have an extraordinary level of comfort viewing and manipulating data in spreadsheets.

A Brief History of Spreadsheets

Spreadsheets are such a natural means of viewing and communicating data that they’ve essentially been around since humans started recording data. Early Babylonian tablets from around 1800 BCE were used to record accounting ledgers and astronomical records. Scientists, accountants, and other data-centric professions have long kept paper records in tabular formats.

If you look at the history of enterprise computing, spreadsheets were commonly one of the first applications built on every major computing platform. Spreadsheet programs written in FORTRAN were used on mainframes starting in the early 1960s. VisiCalc, the Apple II spreadsheet application, was considered the first “killer application” that was so useful that people would purchase a particular computer just to use it. Lotus 1-2-3 was a spreadsheet application that drove the widespread adoption of the first IBM personal computers and Excel helped drive the success of the early Microsoft Windows programs in businesses.

Basically spreadsheet applications have always been a major part of enterprise computing because, for certain types of data, spreadsheets simply provide the best means of viewing and analyzing data. However, in recent years, advances across the data analytics ecosystem have largely left out the spreadsheet.

The Modern Data Stack

The data analysis application ecosystem has undergone a tremendous amount of innovation in the past 10-15 years. The modern data stack now includes simple, reliable extract, transform, and load (ETL) applications and orchestration tools, cloud databases specifically architected for analytics and cheap storage of large amounts of data, and cloud managed services that simplify managing the infrastructure that underlies analytics applications.

However, the “frontend” of the modern data stack - the applications that actually face business users tasked with using data to derive insights and take action - has seen relatively little change in that same time span. There is still a focus on providing static reports and connecting data visualization applications like Tableau, Power BI, and Looker to data sources as a means of presenting data to business users. These types of BI tools were developed when data set sizes outgrew the capabilities of spreadsheets and new software was needed to enable the average business user interact with big data sets. BI tools sold themselves with the promise of "self-service analytics," claiming dashboards enable everyone within an organization to perform big data analysis. The reality of what has happened is the vast majority of business users don't know how to use BI tools and those that do still want data in a more flexible tool that enables them to perform their own analysis.

Spreadsheets and The Modern Data Stack

Data visualization tools, like tableau, PowerBI, and Looker are great for creating dashboards and presenting data in easy to digest ways. However, many more analytical data consumers, like financial analysts, supply chain analysts, marketing analysts, etc. prefer data presented in a way that allows them to perform their own analysis and in a format that feels native to them, like a spreadsheet.

The problem is, most current spreadsheet applications like Microsoft Excel and Google Sheets don’t fit well with the rest of the modern data stack for 3 main reasons:

  1. Maintaining real-time connections to data for many users in these applications is difficult.
  2. The data in locally run spreadsheets is difficult to share with wider audiences, creating data governance issues.
  3. There are performance limitations with Excel and Google Sheets on large data sets, which are becoming the norm in every job function.

There is a big opportunity for modern, connected spreadsheet applications like Row Zero to provide a frontend data tool that allows business users access to data in a flexible, familiar tool, while integrating nicely with the rest of the modern data stack through easy connections to analytical cloud databases like Snowflake and Amazon Redshift and seamlessly handling large data sets.

Solving the Last Mile Analytics Problem with Spreadsheets

Spreadsheets have always had a place in data analysis and have always provided business users with a way to conduct flexible data analysis in a familiar format. Recent advancements in the modern data stack have made data more accessible to these users, but failed to give them a true frontend application that really meets their needs. Combining the familiarity and flexibility of the spreadsheet with the advancements of the modern data stack can empower businesses to truly derive insight and value from their data and solve the last mile analytics problem.

New spreadsheet tools like Row Zero present the best of both worlds: the ease and speed of the modern data stack in a flexible, familiar data analysis application like the spreadsheet.