Excel is an amazing tool. It enables the creation and expression of an endless array of financial and business concepts that can be generally understood by anybody. And if not immediately understood, it can be used as a tool for ever deeper learning. This is the reason modern business still runs on Excel, in spite of a growing number of SaaS products that look to displace Excel. Some do this successfully when there is a very narrow context and the business logic is relatively static and well understood. But we all know these systems are often shadowed by Excel applications to fill in the inevitable gaps.
The primary problem with Excel arises when business users start storing core data in Excel that is necessary to operate the business. We all know that Excel lacks the integrity of a database, but the near term convenience is easily prioritized over the benefits of a more robust solution. There is also a natural process of refinement as business processes and data structures evolve that make it premature to invest in a robust solution while requirements are still fluid. So it becomes a catch-22, it’s a bad idea to use Excel as a data repository, but it’s also a bad idea to hard bake a solution with dynamic requirements.
With the problem identified, we then begin to look for solutions. Is there an abstraction level between SaaS-ificatuon and Excel? We think so, and the next investigatory step is to identify patterns between these spreadsheet applications. Our particular area of interest is in financial forecasting spreadsheets, though other areas are likely to present their own sets of unique patterns. In financial forecasting spreadsheets (and even non-financial forecasting spreadsheets) we observe the following:
- Periodicity: There are clear break points of periodicity that are of interest, with monthly often being the minimum and annual being the maximum.
- Table Encapsulation: Almost all forecasting data can be reduced to a series of tables that have a heading, rows of calculations and in some cases, a summary row. Where the atomic unit of a spreadsheet is a cell, tables can generally represent the atomic unit for forecasting applications.
With these two key insights, we now have the building blocks for a modular system to build a solution that is not as general as Excel, but is more general than most SaaS products.
Why does such a solution not exist already? The reason is not entirely clear, but here are a few possibilities:
- Initial Web Conversion: Until now, there has been a lot of wood to chop with respect to the numerous static processes in business, and this is a much harder problem to solve.
- VC Trend Following: VCs have been focused on business model disrupters, then crypto, and now AI.
- Impossible Task: It will be hard to pry corporate America off spreadsheets and it is a fool’s errand to undertake such an effort.
- Lacking the Right Problem Domain: The right solution will require starting with a really hard problem to get the abstraction layers correct.
This last point is where we think we have made strides. In our core business of advising renewable energy developers and owners on financing and developing highly complex structured finance models, we think we have observed most of the possible complexities in spreadsheet analysis, including:
- Circular references.
- Aggregating N number of projects into a single set of cashflows.
- Scenario analysis.
- Large variation in how inputs can be structured.
- The need for custom table formulas that cannot be determined a priori.
- Deeply nested dependency trees.
- Custom data.
- Tracking actuals versus forecasts.
This rich problem set makes a stand-alone domain specific SaaS application almost impossible given the need for constant model changes to reflect different deal structures. A more general approach is the only one that can be widely adopted and solve the catch-22 issue noted above.
We believe that the Structrz platform represents a paradigm shift that can enable most enterprises to get the benefit of data warehousing core business information while using a tool that provides just enough of the flexibility of Excel to meet the business needs in a highly flexible format. Key to this is maintaining a spreadsheet-like environment on the web, with essentially all of the benefits of address based formulas, with the added ability to export an identical Excel workbook containing live formulas. Time will tell if users will be open to making the trade-off, voluntarily or with the influence of management that struggles with the huge amount of opacity that is reflected in the current approach.