Power BI

Episode 320 – Power BI February 2026 Feature Summary

John and Jason are fresh off a working trip together in Hawaii — and while the aloha spirit may have faded (especially with San Antonio already hitting 90°F before March even showed up), the February 2026 Power BI Feature Summary gave them plenty to dig into. This is a focused, quality-over-quantity month: smarter Copilot experiences, a long-requested slicer upgrade reaching GA, useful modeling additions, and a handful of deprecations worth paying attention to before they sneak up on you.

First, the Goodbyes — Deprecations to Know

Before getting to the new stuff, John and Jason walked through a few things waving farewell in 2026.

Power BI Scorecard Hierarchies (and the Heat Map) Are Gone

Scorecards themselves remain intact, but hierarchies — and the heat map view that came with them — are being discontinued. Jason wondered aloud how much longer scorecards stick around in general. John drew a direct line back to ProClarity (yes, that ProClarity) and the original balanced scorecard concept. If roll-ups and hierarchies never really got broad adoption, it’s hard to argue with their removal. The renaming saga from Goals → Metrics → Scorecards has been its own kind of journey, and we probably haven’t seen the last chapter yet.

Legacy Excel/CSV Semantic Model Experience Riding Off Into the Sunset

The entry point for the legacy “get data from Excel or CSV” semantic model experience disappears on May 31, 2026. Refresh support for models created through the legacy experience ends on July 31, 2026. John and Jason both acknowledged they stopped demoing this approach a year or two ago when the better Get Data experience arrived. Their advice: if you’ve got anyone still using this workflow, migrate now — because the people who haven’t been paying attention to the feature changelog are exactly the ones who are going to get caught flat-footed by this deadline.

SSRS, PBIRS, and AS Management Packs in SCOM Going Away

The management packs for SQL Server Reporting Services, Power BI Report Server, and Analysis Services in System Center Operations Manager (SCOM) will be deprecated effective January 2027, with no further updates planned. Jason had a good laugh — SCOM was one of his early areas of expertise before it became the thing nobody wanted to own in the data center. Azure-based alternatives are the suggested path forward. If this one applies to you, you probably already know who to call.

Simba Vertica ODBC Driver

The Simba Vertica ODBC driver deprecation process begins in February 2026. Customers using the Vertica connector should transition to the Vertica ODBC driver, which is generally available and will continue to receive investment. Brief mention on the show — if it applies to you, you know what to do.

Copilot & AI: From Tweet-Length to Genuinely Useful

Copilot Prompt Limits: From 500 Characters to 10,000

The Copilot prompt input limit across all surfaces — report pane, apps, mobile, embedded — has expanded from 500 characters to 10,000. As Jason put it, you’re no longer trying to cram your intent into something the size of an old tweet. You can now give Copilot actual context, specific instructions, and enough detail to get genuinely useful results. John’s comparison to Twitter expanding from 144 to 288 characters landed perfectly.

Fewer Steps to Insights in Power BI Apps + Copilot App Awareness

Copilot in Power BI apps is getting more conversational — it now shows up at the top of your app navigation and understands the context of the app itself. Ask it “what’s this app all about?” and it’ll actually tell you. It used to prompt you to select which items to query; now it figures it out. App authors also get a new control: you can manage Copilot visibility in workspace apps, which means you can turn it off if you’re not backed by a Copilot capacity. Jason is leaning more toward data agents these days than traditional apps, but acknowledged this is a meaningful improvement for those who are.

Reporting: The Slicer Gets Its Moment

Input Slicer Is Now Generally Available

This one was a long time coming. Formerly known as the “text slicer” during preview, the Input Slicer is now GA — and it’s a proper upgrade from what came before. Report consumers can now filter using contains any, contains all, starts with, ends with, is any, does not contain, is not any, and more. You can stack conditions. It’s flexible, it’s powerful, and it’s exactly what customers were asking for. Jason recalled pushing for something like this back when he was a customer, then hearing the same requests over and over from customers when he was at Microsoft. It was asked for a lot — and now it’s here.

Paste Multiple Selections Into Any Report Slicer

Any report slicer can now accept multiple selections pasted directly from the clipboard. Copy a list of values from a spreadsheet, paste it into a slicer, and the report filters accordingly — no more manually selecting each item while holding Ctrl. Jason called this one out as more than fit and finish — it’s genuinely new functionality. John agreed. It’s the kind of thing that sounds small until you’ve ever needed it.

Card Visual Cross-Filtering Fix

If you’ve been using categories in the card visual and noticed that clicking a category didn’t apply that as a filter to the rest of the report — that was a bug, and it’s now fixed. Clicking a card category will now cross-filter other visuals just as you’d expect. Jason mentioned he’d been trying to use exactly this behavior just two weeks before the update dropped. Timing.

Conditional Formatting: Data-Driven Colors

The conditional formatting dialogue has been updated to support data-driven colors. Rather than just setting static threshold-based colors, you can now bind the color output to a column in your data — whether that’s a color name or a hex code stored in the model. It’s the kind of precision that report builders have wanted for a while, and John gave a clean explanation of how it works: the column value drives the format.

Improved Error Dialogues in Power BI Desktop

When a visual renders with one of those dreaded red X error states, Power BI Desktop will now give you more useful information about what went wrong. Jason hasn’t had a chance to intentionally break something to test it yet — but given his track record, that’s probably coming. There’s also a direct link to troubleshoot visualizations documentation in the “learn more” section, which is a nice touch.

Azure Maps Updates: Performance + New Geographies

The Azure Maps visual continues to get attention this month, primarily with performance improvements. If you’re rendering pie charts on a map — yes, hundreds of pie charts on a map — they now render significantly faster. More importantly for international deployments: Azure Maps has now lit up in Korea and Brazil, addressing data sovereignty requirements that previously blocked usage in those regions.

Preview Visuals Get Better Labeling + Restore Default Option

Preview visuals will now clearly surface their preview status in brackets when you hover — which matters for accessibility and screen readers. Eventually, all preview visuals will appear below the divider line in the visual picker rather than mixed in with production visuals. And if you’ve customized your visual panel and need a clean slate, there’s now a “restore default visuals” option in the ellipsis menu. Jason noted that the accessibility angle is an underrated part of this change.

Font Rendering Fix for Non-Windows Devices

Some fonts — including Segoe UI — weren’t rendering correctly on non-Windows devices. That’s fixed. Short mention on the show, meaningful fix for anyone building cross-platform reports.

Modeling: New DAX Functions Come Out of the Cold

TABLEOF and NAMEOF DAX Functions Are Now Official

Two DAX functions — TABLEOF and NAMEOF — have apparently been lurking in the engine undocumented for a while. They’re now first-class, documented citizens. NAMEOF returns the source name of an object; TABLEOF returns the source table of a measure. Together, they reduce the need to be explicit in naming scenarios and make renaming operations smoother. They’re useful in field parameters and calculation groups, but Jason raised the obvious question about which use case actually drove the engineering investment. His vote: User-Defined Functions (UDFs). John agreed. Either way, Power BI benefits from the Fabric spillover — and that’s a win.

Admin & Governance: One Setting Flip

Fabric Copilot Capacity Designation Now On by Default

Beginning February 8, 2026, the tenant setting Capacities can be designated as Fabric Copilot capacities is enabled by default for all tenants. To be clear — as John carefully noted — this does not change any existing capacity configurations. It doesn’t magically assign a capacity as a Copilot capacity. It just turns on the ability to do so. If your organization hasn’t designated any Copilot capacities, nothing in your environment changes. Security admins may still want to review this one carefully. The ability to opt out exists in the Fabric admin portal.

Org Apps: Persistent Filters (Finally)

Org Apps — the Fabric-level app structure built on the same fundamentals as Power BI apps — now persist filters not just across pages within a session, but across visits to the app entirely. If you left a filter applied last Tuesday, it’ll still be applied when you come back. This is the kind of behavior most users probably assumed was already there.

Column Sizing Reminder: Content-Based Is Back as Default

A quick reminder carried over from January: table and matrix columns now default to content-based sizing again in new reports, following user feedback after the brief change. The January release rolled this out as the new default. If you need to adjust manually or use “grow to fit,” that’s still available — but the default is back to what felt natural.

What’s Coming Next?

The episode closed with a bit of Kreskin-style prediction about when the March updates would drop. John guessed the week of March 16th. Jason nudged toward Wednesday specifically — Keynote Day at FabCon. As Jason noted with a grin, that would make it a birthday present. John is heading to Atlanta for FabCon, then MVP Summit. After that, the two are reuniting in Cologne for the co-located European Collaboration Summit. More on that as it gets closer.

The March update — and whatever FabCon announcements come with it — should be interesting. Stay tuned.

Links

Microsoft Blog Posts Referenced in This Episode

Previous Episodes

Events

Subscribe: SoundCloud | iTunes | Spotify | TuneIn | Amazon Music

Tags: , , , , , , , , , , , , , , , , , , , ,

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

AvePoint