What are OLAP and OLTP and why should you care about them as a finance leader?

OLAP and OLTP are two fundamental database systems that encompass the majority of data that finance teams use.

OLTP is like a data tab in a spreadsheet, based on transactions (online transaction processing).

Examples of OLTP systems include:

- SQL and relational databases,

- CRM systems (Salesforce…)

- ERP systems (NetSuite, Intacct…)

- HR systems (Workday, ADP…)

- Operational systems (Stripe, Shopify…)

OLAP is like a pivot table, based on dimensions (online analytical processing).

Examples of OLAP systems include:

- Data warehouses (Redshift, Snowflake, BigQuery...)

- BI systems (Tableau, Looker, Power BI...)

- EPM systems (Anaplan, Pigment, Adaptive, Abacum…)

OLTP and OLAP systems are apples and oranges and not connecting them thoughtfully can lead to big problems:

❌ Aggregating transactions from OLTP -> OLAP is straightforward, but going the opposite direction can be like putting toothpaste back in the bottle. Your latest plan or outlook doesn't have transaction IDs!

❌ Over time, FP&A teams will often try to bring more and more OLTP data into their EPM system in an effort to get more accurate and real-time, effectively creating an OLTP database in the OLAP system. This mismatch leads to poor performance and high costs.

❌ The OLAP EPM system can get into a tug-of-war with the other OLAP system, the BI system. This leads to confusion and conflict on what numbers are correct, and where to go for information.

❌OLAP databases are also not ideal for unstructured data, which is becoming increasingly important for real-time signals and generative AI use cases.

❌ Many OLTP systems try to introduce more OLAP-like features in an effort to be more strategic, but they usually fall short, feeling inflexible and lacking vs. true OLAP systems.

For these reasons, our ‍ approach is designed as follows:

✅Sequences the flow of data and systems to effectively utilize OLTP, OLAP, and unstructured data concepts, providing access to ALL data for planning. The best next-generation planning solutions like Abacum and Pigment leverage a hybrid OLAP/OLTP approach, though the real bottlenecks are usually upstream from the EPM system and can only be band-aided by Finance.

✅Builds the 'planning source of truth' not siloed in the EPM system but in the same place as BI, avoiding an unsustainable situation with high costs and data volumes in the EPM software and unifying reporting and analytics across the organization. This relieves pressure on the EPM system and allows it to be more focused on modeling and collaboration.

✅ Supports the use of the best hybrid OLAP/OLTP modeling tool out there: spreadsheets. No other tool handles both transactional data and pivot tables (assuming data volume isn't an issue) with as much ease and flexibility as spreadsheets. This is a big reason why spreadsheets just feel so right despite their downsides.