Planning > Data Writeback in a BI or data platform

BI / OLAP Reporting systems have been in use for over 40+ years and isolated from your ERP/transactional systems (OLTP) for read vs write performance reasons. Planning systems on the other hand require a hybrid approach, as they require fast reads as well as concurrent writes (of budgets, forecasts and scenarios entered by users).  

As database technologies have matured, Hybrid Transactional / Analytical Processing (HTAP, a combination of OLAP+ OLTP) will be a differentiator for database and data platform vendors.  

For over two decades, BI Vendors have been offering basic write-back capabilities that capture user inputs in table/forms and save them back to a database. But none of them have matured to the level of a planning solution, something that is validated by the absence of BI software in the Gartner Magic Quadrant for planning.  

The foundation of data writeback is the provision of data input templates in the BI software for capturing user inputs. 

Enterprises today achieve this functionality through laborious workarounds with varying levels of success in each BI software. For example, Power BI users use code-based PowerApps solutions, but it can handle only simple data inputs and commentary capture. 

Do input tables with write-back make an enterprise-class planning solution?  

Business expectations for planning can become quite exhaustive. For example, here are a few considerations: 

• The table needs to support data inputs with time hierarchy in columns - like a pivot table matrix  

• Users need the ability to see prior period actuals next to budget input columns 

• Users must be able to create multiple versions of budgets and forecasts 

• The solution must support scenarios (with writeback) 

• You must be able to spread and distribute forecasts to granular items 

• Enterprises must be able to audit changes so that they can track who made the changes & when 

• The solution must inherit platform/database-level security (e.g., row level security in Power BI) 

• Managers must be able to approve or reject budgets entered by subordinates 

• When you use rolling forecasts, any period close must trigger a copy from ‘actuals’ to the ‘forecast’ for the closed period 

On top of this, your data/architecture team might not want users to save data to their on-prem SQL server or Oracle Database - but would rather see the solution support writeback to the data platform (Fabric OneLake, Snowflake, Databricks or BigQuery) in the cloud. Not all solutions support this today. 

Due to the above reasons, planning is simply much more than write-back. Having a strong writeback capability does not ensure that your planning needs will be met seamlessly. 

Lumel’s Enterprise Performance Management (EPM) Suite delivers a robust planning solution that meets all the advanced needs of enterprises. Lumel was voted the best new vendor for EPM in 2024.

Reach out to us to learn more.