Show recap — Tuesday, June 23, 2026

Another Orange Morning Recap: June 23, 2026

Tuesday morning, and the show was back with a full list after a quiet Monday. Yesterday's episode had no updates to run, so the time went to a deep dive with Joshua Oakes on conditional reporting and how to think about old build habits versus talking the same problem through with Breeze. Today the updates were waiting, including a count that did not quite agree across two screens. One side showed eight, the other showed nine. The list settled at nine, and the show got moving.

Here is what the episode covered: activity auto-associations and a detour into the data model, the sandbox cluster inching toward production parity, knowledge vault permissions and where context lives, and a long deep dive on everything new in the Notetaker.

The Show Opened on Associations, Then Drifted to the Data Model

Activity Auto-Associations for App Objects

The first update was one that reads quietly and runs deep. You can now configure activity auto-association settings for app objects from your connected apps, the same way you already do for standard CRM objects. The interesting turn in the conversation was less about app objects specifically and more about the thing they enable, because activities like calls and emails do not show up on a related record's timeline unless the association is configured to put them there.

That control is the part worth slowing down on. There is a real temptation to want everything automated and associated, and the more useful instinct is to confirm what should be associated to a given record and what should not. When you have multiple deals running against the same contact and company, you do not want every activity landing on every deal. App objects also came up as a milestone of their own. They used to require building custom objects; now apps can bring their own objects without forcing customers up to Enterprise, and with this update they behave like first-class citizens when it comes to activity management.

Why it matters: Associations are getting more powerful at the exact moment they matter most. As teams lean on AI to make sense of unstructured data, what that data is associated to, and what it is not, is what decides whether Breeze is working from context or from noise. This update hands the default behavior to the admin who actually understands the relationships.

For more details: Activity auto-associations for app objects live.

Centralized Data Model Management

The associations talk pulled the show naturally into the next one, since both now live closer together. Centralized Data Model Management is a unified table and graph view for managing objects, properties, associations, and conditional logic from one place, and it moved into public beta this round after a private-beta run earlier in the spring. For anyone who spends real time in the data model, this was the happy one of the morning.

The take from the room was that it is well designed, not just consolidated. It reaches just about everything except pipelines, and the aside there was that pipelines arguably belong in the workflows space rather than the data model. Everything else, associations, properties, conditional logic, is now reachable and visible inside one shared context.

Why it matters: Managing the data model used to mean jumping between settings pages with no consistent picture of how the pieces connect. Auditing a complex configuration or making a quick edit to a property group becomes a single-screen task. The move from private to public beta is the cue to get in if you have been waiting, and the bigger signal is that HubSpot now treats the data model as a surface you manage, not a settings detour you tolerate.

For more details: Centralized data model management public beta.

Sandboxes Keep Getting Closer to Production

A cluster of the morning's updates was all the same project from different angles: sandboxes getting better at copying what production actually contains. The framing on the show was that this has been a deliberate push, prompted in part by a recent post and a conversation with the sandbox team about how it was not always clear whether what you built in a sandbox could be pushed to production. The gaps are closing one asset type at a time.

The deep dive landed on sandbox creation now copying automated transactional emails, along with any workflows or segments that reference them, with Deploy to Production supported for new transactional emails too. That sent the conversation down a useful rabbit hole on what a transactional email even is. Order confirmations are the obvious case, but password resets and app login emails for SaaS portals fall in the same bucket. The nerdier point underneath it is that transactional email is not really a HubSpot term at all; it describes how email clients like Gmail and Outlook treat the message, which is why it is more likely to reach the inbox and why providers take abuse of it seriously. That also ties to the footer and unsubscribe rules, since nobody is unsubscribing from an order confirmation.

The other two in the cluster rounded out the same theme. Sandbox creation now copies playbooks from production, including their content and recommendation settings, and team-based and shared CRM views are now copied as well. Together they mean a new sandbox looks a lot more like the environment your team actually works in.

For more details: Sandbox copies automated transactional emails live.

Knowledge Vault Permissions, and Where Context Lives

Knowledge Vault Permissions went live across all hubs and tiers, letting customers manage access using Manage and Use permissions at both the portal and the vault level, with granular control over who can use or edit each vault. The show connected it straight back to yesterday's conversation with Oakes about user context, the idea that every person, and every AI, develops context about how to communicate and what to be careful with.

A story made the point land. One host described onboarding a client earlier this year who was told, very directly, that a particular person did not like AI, so it was worth treading lightly on the topic. Saving that kind of note in HubSpot is what keeps the system from later suggesting you walk that person straight into a Breeze Assistant demo. The client came around over the project as the value showed itself, but the note mattered the whole way through. That is the kind of context that usually lives in personal inboxes and desk notes, and vaults are becoming where it belongs instead.

Why it matters: As organizations accumulate more vaults, not all of them are relevant to every user, and vault-level permissions keep workspaces organized while reducing the risk of accidental exposure as teams scale. The wish out loud was the obvious next step: a Breeze Assistant that notices when something relates to a vault and offers to update it. That is the work people are doing outside HubSpot right now.

For more details: Knowledge vault permissions live.

The Notetaker Deep Dive

The longest stretch of the morning went to everything new in the Notetaker, an update that keeps growing under the Smart Deal Progression name. Rather than read the why-it-matters out loud, the show went hands-on with the two pieces that are new this round: external sharing and translation.

External sharing lets you send conversation recordings, summaries, and transcripts outside HubSpot through secure links, with expiration dates, the ability to revoke access, and no login required for the recipient. The live walk-through had its honest moment of hunting for the share button, which on the standard theme is the big orange one in the upper right, and a big black one for anyone on the new theme. The preview of the shared page was the payoff. All the hard parts, secure links, flexible controls, delivery options, hosted replayable video, are exactly why this took time to ship rather than landing on day one.

Translation came next, with summaries and next steps available in each user's preferred language, switchable on demand without altering the original record. The demo flipped a summary into French in a click. The conversation around it was a fair reminder that good localization has never been only about swapping words. One host recalled taking a product into Latin America years ago, where one of the three core words used to describe it everywhere, rugged, did not translate into Spanish the way anyone intended, landing somewhere closer to strong-man than durable. Per-user translation does not erase that kind of nuance, but for multilingual teams it removes a cost that used to keep the whole capability out of reach.

The segment also covered the logging change with a date on it. Starting July 31, 2026, meetings recorded through Notetaker, Zoom, Google Meet, and Microsoft Teams will create only a meeting record instead of both a call and a meeting record. Chris's read was that a lot of teams have been quietly working off inflated call numbers, and the funnier version of that is the manager who sees call volume drop while revenue holds steady and asks what happened. The answer is that a duplicate record went away. If you run call-triggered workflows meant for meetings, or reports filtered on call objects, those need updating before the deadline, and HubSpot is already adding the key properties to the meeting record so you can rebuild ahead of time. The idea of an end-of-July workshop to get everyone through it together came up, and it is a good one. This change first came up on the June 5 show; that earlier recap has more.

The caveat is real but small: teams already used to the meeting Notetaker living in the call object will have to adjust. The honest read on the show was that the meeting Notetaker is still underutilized, so for most teams this is something to start using, not something to worry about changing. The segment closed on the conversation review improvements, including a new suggested follow-ups card that drops post-engagement recommendations right onto the deal record, the kind of native CRM card people have been building outside HubSpot or with custom agents to get inside it.

Why it matters: The admin work that bookends customer conversations is exactly what makes CRM data go stale, and this update keeps collapsing it into the meeting itself. External sharing and translation are the headline additions, but the July 31 logging change is the one to act on now.

For more details: Notetaker with Smart Deal Progression public beta.

The Rest of June 22's Updates

Two updates from the release did not get pulled forward on the show: a private beta that lets visitors leave Live Chat messages when your team is unavailable, and a public beta that lets help desk users close a live chat while keeping the ticket open. Full breakdowns of all nine updates from June 22 are on the updates blog: HubSpot Updates: June 22, 2026.

The Sign-Off

You probably already own the value you are looking for in HubSpot. You just need to wake up to it. We will see you tomorrow morning.