Databricks acquired Row Zero on 24 September 2026, a cloud spreadsheet that connects straight to a data warehouse instead of to a CSV someone downloaded. For a finance director whose month-end runs on exports from Business Central and a CRM, the point of that deal is simple. The spreadsheet survives, and the export that feeds it does not.
That is the direction, and it isn't the delivery. Nothing about the Genie integration shipped on the day of the announcement, and the parts of it that matter to you are described in the press release in future tense. Row Zero as a product exists today and you can buy it today. The native, fully governed spreadsheet inside Genie is a plan Databricks has stated publicly, and plans move.
What Row Zero actually is
It is a spreadsheet. Not a BI tool with a grid bolted on, and not a notebook pretending to be friendly. Row Zero has 400 plus Excel-compatible functions, pivot tables, charts, keyboard shortcuts and real-time collaboration, per its pricing page. A controller who has spent fifteen years in Excel doesn't retrain.
The difference is where the cells get their numbers. Row Zero has a built-in Databricks connector that authenticates with a personal access token or with Databricks OAuth, and it works on AWS, Azure and GCP. The OAuth path is the one that matters for governance, because queries made over an OAuth connection respect the database permissions of the person running them, including row-level security and role-based access control. The regional sales manager opens the same workbook as the CFO and sees their own region, because Unity Catalog decided that, and no tab had to be remembered and deleted.
Scale is the other half. Databricks says Row Zero's processing engine lets business users work with billions of rows at interactive speeds, which is the number that explains why anyone bothered. Excel stops at roughly a million rows. A full year of order lines for a mid-sized distributor clears that before you have joined anything to it.
Why Databricks calls your exports a pipeline
The press release uses a word worth repeating in your next systems meeting, and that word is spreadmarts. Databricks' argument is that ungoverned spreadsheet files, used to store data or to pass it between people and agents, add up to arguably the largest set of ungoverned data pipelines in the enterprise.
That lands hard in a manufacturer or distributor, because the pipeline is usually a person. Someone exports open orders from Business Central on Tuesday morning. Someone else pulls pipeline out of the CRM. A third workbook VLOOKUPs them together on a customer name that is spelled two ways, and the margin number that goes to the board comes out of a process that nobody owns, that carries no lineage and that nothing tests.
The agent angle is what made this worth an acquisition instead of a feature request. A human who opens last Tuesday's export usually notices when the totals look wrong. An agent handed the same file has nothing to compare it against and will act on it anyway. Databricks makes exactly that point, since companies deploying agents that analyse, model and act on business data can't afford those agents working on sensitive data of unknown lineage and quality. I wrote a margin note on the governance argument in the Databricks Field Guide that goes further into the engineering side of it.
Shipped today, announced for later
Keep these two columns separate when you plan.
Row Zero ships all of this today, including the Databricks connector with PAT or OAuth authentication, import of XLSX, CSV, Parquet and JSONL, pivots and charts, and real-time collaboration. Automated data refresh and write-back to the warehouse start on the Business plan at $20 per user per month. Restricting data export, SSO, SCIM, workspaces, Private Link and the billion-plus row tier are Enterprise, priced on request. The free tier gives you one workbook and tens of millions of rows, which is enough to test whether your people will actually use it.
Also available now, and underrated: Row Zero can discover the Genie Spaces available to a Databricks connection and use a selected space to generate SQL. Setting that up needs admin access to the Databricks account and workspace, Genie enabled, and at least one Genie Space whose underlying Unity Catalog tables the user has SELECT on. The docs spell out the entitlements on both sides.
Still waiting to ship is the native spreadsheet experience inside Genie itself, on web, desktop and mobile, built on Genie Ontology, Unity Catalog and Unity Gateway. Databricks says Row Zero will be available to all Databricks customers across all major clouds and will keep supporting data sources beyond Databricks. No date was given, no pricing for the integrated version was given, and terms of the deal weren't disclosed. Do not put the integrated product in a 2027 budget line yet.
How it fits with Genie and Genie One
The naming shifted this year, so the current shape is worth laying out from Databricks' own documentation. Genie is the umbrella. Genie One is the simplified interface for business users, previously called Databricks One, where someone with only the Consumer access entitlement lands after signing in. From there they can browse dashboards, ask questions in natural language and run apps. Genie Agents are the domain-specific environments your data team configures with trusted tables, metrics and business rules. Genie Code is the developer-facing assistant.
Underneath all of them sits Genie Ontology, currently in Public Preview, the context layer that holds what your business words mean. It combines Unity Catalog semantics you define and certify, such as metric views, with context Genie infers from dashboards, SQL queries and existing agents. Snippets carry an authority score and are gated by Unity Catalog permissions, so Genie only uses definitions drawn from assets you are allowed to see.
This is the part to act on, and you can act on it today. If your definition of net revenue, on-time delivery or gross margin by product family only exists in the head of the person who maintains the workbook, a spreadsheet interface over Genie inherits that gap and presents it more convincingly. Get the definitions into Unity Catalog as metric views, once, where every tool reads the same one. That work pays whether the Genie spreadsheet ships in March or never.
The chat surface and the grid answer different questions, too. Genie One chat is good for "why did margin drop in the Midwest last month" and returns an answer, a chart and the query. A grid is for the thing after that, the scenario where you push three price changes through a model and watch the contribution line move.
Getting Business Central and your CRM in there first
None of this helps if the governed data isn't governed data yet. A live spreadsheet pointed at a Databricks workspace that holds nothing from your ERP is a faster way to look at an empty table.
For most mid-sized manufacturers and distributors, the sequence runs like this. Land Business Central and CRM data in Unity Catalog on a schedule you trust, model it into tables finance recognises, define the metrics once, then put a spreadsheet on top. The first step is the unglamorous one. Worth knowing: Lakeflow Connect's Dynamics 365 connector reads Dataverse, and Business Central doesn't sit in Dataverse, so getting BC data into the lakehouse is a separate piece of work. We cover the options in data integration, and the reporting trade-offs specifically for BC in an earlier post.
Where a spreadsheet is still the wrong tool
Live data doesn't make a workbook into a system of record. Four places I would hold the line.
Recurring transformations belong in pipelines. If the same seventeen steps run every month to turn raw order lines into the revenue table, that is ETL, and it should be versioned code with tests and a schedule instead of formulas that one person understands. Row Zero can write back to the warehouse, which makes it easy to put a hand-built number into a governed table. Easy and wise are different.
Statutory and audited reporting belongs where the controls are. Version history on a workbook isn't a close process.
Anything a hundred people read should be a dashboard. A grid serves the person modelling, and the audience needs something else.
And the definition of a metric belongs in the semantic layer, well away from cell B7. The moment two workbooks disagree about what a qualified lead is, you are back in the reconciliation meeting the live connection was supposed to end.
What I would do about it this quarter
Nothing structural. The integrated product isn't here, and building around an announcement is how teams end up with a migration they didn't need.
Do the two things that pay regardless. First, write down which spreadsheets currently carry a number that leaves the building, and who rebuilds each one by hand. That list is your export inventory and it will be longer than you expect. Second, move the metric definitions on that list into Unity Catalog so Genie, your dashboards and whatever spreadsheet arrives all resolve the same words the same way.
If your finance team is already pulling CSVs out of a warehouse into a shared workbook every month, that is precisely the workload this acquisition is aimed at. Knowing where the platform is heading is a good reason to leave the replacement unbuilt.
Frequently asked questions
What is Row Zero?
Row Zero is a cloud spreadsheet that connects directly to live data warehouse tables instead of to exported files, founded by former AWS and Tableau engineers and acquired by Databricks on 24 September 2026. It offers 400 plus Excel-compatible functions, pivot tables and charts, and Databricks says its processing engine handles billions of rows at interactive speeds.
Is the Row Zero spreadsheet inside Genie available now?
No. Databricks has announced that Genie will gain a native, fully governed spreadsheet experience built on Genie Ontology, Unity Catalog and Unity Gateway, but no release date was given and nothing shipped on the announcement date. Row Zero as a standalone product with its own Databricks connector is available today.
How much does Row Zero cost?
Row Zero's published pricing starts with a free tier giving you one workbook. Pro is $12 per user per month for unlimited workbooks and version history, Business is $20 per user per month adding automated refresh and write-back to the warehouse, and Enterprise is custom-priced for the billion-plus row tier, SSO, SCIM and export restrictions. Pricing for the planned Genie integration has not been announced.
Can Row Zero connect to Business Central?
Row Zero has no Business Central connector. It connects to data platforms such as Databricks, so the practical route is to land Business Central data in the lakehouse first, and the Lakeflow Connect Dynamics 365 connector reads Dataverse instead of Business Central, which makes ERP ingestion a separate piece of work.