Live actuals to rolling forecast in SAC: the hybrid pattern (Live Versions + Data Actions)

The most useful recent change in SAP Analytics Cloud planning isn't a new chart — it's that your actuals can now stay live while you plan on top of them. Introduced with the QRC4 2025 release, "live versions for planning" enable a hybrid pattern that removes the old import-and-reconcile step from a rolling forecast. Here's the pattern, precisely, with its real boundaries.

The pattern in one line

Actuals stay in the source as a Live Version; a Data Action copies them into an SAC plan version; you build the rolling forecast on that plan version. No data replication into SAC, and no write-back to the source.

What "Live Version" means — and where it works

A live version links a planning model to live data so the source is the actual baseline, without importing it. Concretely, you can display live version data, reference it in model and story calculations and in data actions (including advanced formulas), and copy it into planning versions via table copy/paste, data actions and version management. What you cannot do: write back to the source — plan data always lives in SAC's internal versions.

The boundary that matters: this applies to SQL live sources — SAP Datasphere (seamless planning), Snowflake (from 2026) and Google BigQuery, with more SQL sources coming. It does not apply to a classic BW/4HANA or S/4HANA live connection, which connects to a query or BW object under a different paradigm. If your actuals sit in classic BW or S/4HANA live, you're still on the import route for planning — see live connection vs import model.

The accelerator, step by step

— 1. Add the live version as the actual baseline. In model details, under Data, use External Live Version Data Sources to link the source view (Snowflake/BigQuery/Datasphere). SAC flags any mismatch between the live data and your master data so you can add missing dimension members.
— 2. Copy actuals into a plan version with a Data Action: add a Copy step from the live actual version to a plan version, optionally with a factor or seasonality profile (e.g. 1.03 for a flat uplift). The pattern is the same Actual→Plan copy covered in our data actions guide.
— 3. Build the rolling forecast on the plan version — driver assumptions, allocations, a rolling window that always looks N months ahead. Version design matters here; see the Version dimension and annual budget vs rolling forecast.
— 4. Report forecast vs budget vs prior year in one story, refreshed against the live actuals underneath.

Why this beats the old way

Until now you had to physically bring actuals into the model with import jobs, uploads or cross-model data actions — with the latency and reconciliation work that implies. Live versions pull actuals (and external drivers) directly, without replication, so the forecast is built on the latest governed facts from a single source of truth. Rolling forecasts and variance analysis become immediate, and live versions can even feed SAC predictive scenarios.

The honest limits

— SQL sources only (Datasphere seamless planning, Snowflake, BigQuery) — not classic BW/S4 live.
— No write-back: the source stays read-only; plan data is stored in SAC versions.
— Not all data types are supported, and Snowflake models remain native SAC models (they can't be deployed to Datasphere like seamless-planning models).
— Performance depends on co-location: with Snowflake, put SAC and Snowflake in the same region or expect a hit.

Building it as an accelerator

The rolling-forecast model structure (dimensions, measures, versions) is the reusable, packageable part — the template catalog has a free rolling-forecast skeleton to start from. The live wiring is source-specific by nature, so it comes as a blueprint (the steps above), not a magic file. Put together, that's a real "live actuals → rolling forecast" accelerator: a clean model to receive the data, plus the pattern to connect it.

Sources: SAP Community — "Live versions" for SAP Analytics Cloud planning (QRC4 2025 / fast-track wave 20.2025), seamless planning in SAP Business Data Cloud, live versions from SQL sources (BigQuery, Snowflake); SAP Help — data actions and version copy. Availability and supported sources change each release — confirm against your version.