This article is an update for customers who previously used Topics in the platform.
Drivers vs. Topics
Drivers is a new categorization model — cleaner and more consistent than Topics. Rather than adjusting the existing Topics structure, we rebuilt the model from the ground up.
In practice, that means:
Fewer, clearer categories — no more near-identical labels that made you second-guess whether two categories meant the same thing, and no more digging through hundreds of seemingly similar Topics
A visible relationship between Themes and Drivers — you can see exactly which Drivers make up a Theme and how much each one contributes, via new views: the Top 20 Themes tree map, and the Themes and Top Drivers stacked horizontal bar
Categorization you can trust — when something is grouped a certain way, it’s easier to see why
The result: reporting that’s easier to interpret at a glance, and trend-spotting you can act on with confidence.
Going forward, Drivers is the default across your reporting experience — every new report, dashboard, and filter uses it automatically.
KBo runs on Drivers data only. Any KBo answer or insight that references categorization reflects conversations from July 1, 2026 forward — KBo doesn’t draw on Legacy Topics data.
What happens to my historical data and reports?
Your Topics-based reports and historical data aren’t going away.
Because Drivers is a new and improved way of categorizing conversations, we’ve drawn a clean line at July 1, 2026:
Conversations before July 1, 2026 are categorized using Topics and Themes from Topics (legacy).
Conversations from July 1, 2026 forward are categorized using Drivers and Themes from Drivers (new).
In the platform, you’ll see two versions of some reports and filters — a Legacy version covering the older, Topics-based data, and a Current version covering new, Drivers-based data. Both are fully accessible at any time.
Applying date ranges
Legacy Topics-based reports only show data from before July 1, 2026. Apply a later date range and you won’t see data — you’ll see embedded guidance pointing you to the current version instead.
Current Drivers-based reports only show data from July 1, 2026 forward. Apply an earlier date range and you’ll see the same kind of embedded guidance pointing you to Legacy Reports.
If your date range spans both periods, each report shows only its own side of the split, with an info icon informing you of what data is available, so you’re never looking at blended or misleading numbers.
Want to view both periods together? Add a Legacy report and its Current equivalent to the same custom dashboard, then set each card’s date range independently using chart-level filtering — the Legacy card scoped to before 7/1/26, the Current card scoped to on/after 7/1/26. You’ll get both periods side-by-side on one dashboard; there’s still no single blended trend line across the cutover.
Need Drivers-based categorization for conversations before July 1, 2026? Backfill can be requested — reach out to your Customer Success Manager.
Two reports — Tree Map and Themes & Top Drivers — are new with Drivers and don’t have a legacy equivalent, since they didn’t exist before this upgrade.
Where do I find things now?
Existing dashboards and reports you've already built continue working exactly as they do — there's nothing you need to do.
1. Your Themes and Drivers dashboard is right where your Themes and Topics dashboard used to be.
The reports you know and love got an upgrade. The new Themes and Drivers dashboard has Drivers on the Rise, Top 20 Themes, Themes and Top Drivers, Trending Drivers, Primary Themes, Primary Themes Over Time, and Themes and Eddy Effect — all powered by Drivers data from 7/1/26 onwards.
2. Need historical Topics/Themes data? Add a Legacy report to a custom dashboard from the Add Cards menu.
Every Topics-based report — including Themes reports built on Topics — is still there, just grouped under Legacy Reports. Each one carries a Deprecated tag so you know at a glance it’s historical.
3. Building a filter? Look for the tag.
New, Drivers-based filters (like Primary Driver and Primary Theme) sit in the Core AI Models section of your filter list with a New tag. Your existing Topics-based filters are still available as filters too, grouped further down under Legacy Filters and tagged Deprecated.
4. If a report looks empty, check the date range.
If you apply a date range that doesn’t match a report’s data (for example, a post–July 1 range on a Legacy report), you’ll see an in-product message explaining why, with a link to more detail — not just a blank screen.
Scenario | Legacy (Topics) | Current (Drivers) |
Covers data from | Before July 1, 2026 | July 1, 2026 forward |
Where to find reports | Legacy Reports section in the Add Cards modal (when adding reports to custom dashboards) | Themes and Drivers standard dashboard; Reports section in the Add Cards modal (when adding reports to custom dashboards) |
Where to find filters | Primary Topic filter and Primary Theme filter live in the Legacy Filters section at the bottom of your filters list | Primary Driver and (new) Primary Theme filter live in the Core AI models section of your filter list |
How it’s marked | “Deprecated” tag | “New” tag |
FAQs
Is there a mapping between my old Topics and the new Drivers?
No. Drivers isn't a renamed or reorganized version of Topics — it's a new categorization model built independently, so there's no 1:1 or many-to-1 crosswalk between the two. A Driver may cover conversation ground that used to span several Topics, less than one Topic, or a mix that doesn't map cleanly at all. This is also why Legacy and Current reports are kept fully separate rather than shown as a continuous trend line.
Why doesn't Authenticx provide a Topics-to-Drivers mapping?
Because the models were built independently, any mapping we produced would be an approximation, not a precise crosswalk and could be misleading if used to explain shifts in trends. We'd rather be upfront that the two models aren't directly comparable than hand you a mapping that implies more precision than exists. Your CSM has already taken time to rebuild filters where appropriate with Drivers instead of Topics.
My Driver-based numbers look different from my old Topics-based numbers for a similar-sounding category. What happened?
This is expected, and isn't a data quality issue. Since Drivers wasn't derived from Topics, differences in volume or grouping between a Topic and a similarly named Driver reflect the two models categorizing conversations differently — not a change in your underlying conversation data.
Questions?
Your Customer Success Manager is happy to walk through this with you — reach out anytime.