Cornerstone analytics: your options compared

Ask five people how to do analytics on Cornerstone OnDemand data and you get five answers: "just use the reports", "connect Power BI", "put it in the warehouse", "buy a people analytics platform", and – from us – "use octo.taxi". All five can be right. They solve different problems, and the expensive mistake is picking one for a problem it was not built for.

This is our honest map. We build on Cornerstone every day, we make one of the options, and we will say where each one shines and where it stops – with a source for every claim we make about someone else's product.

Option 1: Cornerstone's own reporting

Cornerstone ships a full reporting family: canned standard reports, the Reporting 2.0 builder for your own reports, and Dashboard Reports – pages of small report widgets, up to nine per dashboard, that users can personalize and share. All of it reads from Cornerstone's Real-Time Data Warehouse (RTDW), a read-only store refreshed about every 15 minutes in production.

Strengths. It is already in your licence, it knows every Cornerstone object, it respects Cornerstone's own permissions, and nobody has to move data anywhere. For an ad-hoc list or a standard question, it is the fastest answer there is.

Limits. Cornerstone itself notes that "some data is not available as a dashboard" and points to standard reports and Reporting 2.0 for those cases. Reporting 2.0 downloads stop at 200,000 records by default and one million at most, and the output is built for a person reading a report, not for a system that needs the data continuously. More in the Reporting 2.0 limits.

Option 2: Cornerstone's Power BI utility

Cornerstone publishes the CSOD Power BI Integration Utility on GitHub: an HTML helper that generates the query string for a direct Power BI (or Excel Power Query) connection to Cornerstone's Edge OData API.

Strengths. Free, from Cornerstone, no middleware, and it brings Cornerstone data into the BI tool many companies already standardize on.

Limits. Per its README it "currently only supports Basic 2-Factor Authentication" and "signature based authentication cannot be used at this time". You model the raw objects yourself in Power BI: joins, definitions and paging are yours. We compare it with the other Power BI routes in Cornerstone data in Power BI.

Option 3: your own pipeline or an ETL connector

Build a job on the Data Exporter API (DEAPI) or the Reporting API into your own warehouse – or buy the plumbing from an ETL vendor. Fivetran's Cornerstone connector, for example, is listed as a "Lite" connector and lands tables such as users, OUs, training, transcripts, certifications, goals and requisitions in your warehouse.

Strengths. Cornerstone data sits next to every other source in the warehouse; you control retention and modelling; with a connector, somebody else maintains the extraction.

Limits. The warehouse, the model and the definitions on top are still your job – a table of transcripts is not yet a compliance rate. A self-built pipeline also inherits the API throttle limits and the RTDW lag, and needs an owner for every Cornerstone schema change.

Option 4: a people analytics platform

Dedicated HR analytics platforms combine data from many HR systems into one analytical model, with benchmarks, planning and their own metric libraries.

• SplashBI announced in May 2025 that it joined Cornerstone as a technology partner, so Cornerstone customers can analyze Cornerstone data alongside other HCM, applicant tracking and finance systems in one place.
• Visier is one of the best-known people analytics platforms. Its published list of data connectors covers HR systems such as Workday, SAP SuccessFactors and UKG Pro and warehouses such as Snowflake, BigQuery and Redshift; Visier also takes data as files transferred over SFTP. If Cornerstone is your main source, ask how its data would reach the platform and how often.

Strengths. The broadest cross-system view, built for organizations whose question is "the whole workforce across all systems".

Limits. These are platform projects: the data has to be fed in and mapped, and Cornerstone is one source among many rather than the centre. Check for each platform how Cornerstone data gets in and how current it will be.

Option 5: octo.taxi

We make this one, so weigh this section accordingly.

octo.taxi replicates your Cornerstone OnDemand data through the Data Exporter API with delta sync, in near-realtime – our own added data age stays at or below 15 minutes, on top of the RTDW's lag – and turns it into ten built-in dashboards, a search that answers, Ask octo for plain-English questions, alerts, an org chart, a Cornerstone-compatible REST API and an OData feed for Power BI. Data is stored in Frankfurt, minimum group sizes are on by default, and every view shows its data age. Details: What is octo.taxi.

Strengths. Ready after the first sync, no warehouse needed, Cornerstone's own definitions of "required" and "overdue" applied the same way everywhere, built for works councils and data protection officers from the start.

Limits. Cornerstone first: other HR systems are connected on request, not out of the box, so it is not a cross-system analytics platform today. It shows the objects it syncs, not every corner of Cornerstone. And it is read-only: it never changes anything in Cornerstone.

Side by side

Which one when

• One list, one question, today: Cornerstone's own reporting.
• A Power BI enthusiast and a few reports: Cornerstone's Power BI utility.
• A data team with a warehouse full of other HR sources: your own pipeline or an ETL connector.
• The whole workforce across many systems, with benchmarks and planning: a people analytics platform.
• Current Cornerstone dashboards for HR, L&D and managers, safe to show a works council, without a warehouse: octo.taxi – and its OData feed if Power BI should sit on top.

The honest bit

Most organizations end up with two of these, not one: Cornerstone's reporting for the daily lists, plus one option for the continuous picture. The trap is not choosing the "wrong" tool. It is letting each option invent its own definition of "overdue", so the number in the board deck and the number on the L&D screen never quite match.

Pick the option whose maintenance you can actually carry, and agree the definitions once.

Your data should move freely, from A to B, in style.

---

Sources. Cornerstone: Dashboard Reports overview · Cornerstone: Real-Time Data Warehouse overview · Cornerstone: Reporting 2.0 – Create Report · Cornerstone API throttling (csod.dev) · CSOD Power BI Integration Utility (GitHub) · Fivetran: Cornerstone connector · SplashBI: technology partnership with Cornerstone (May 2025) · Visier: data connectors.