HubSpot Marketing Studio Beta and Google Analytics Source Platform Dimensions

On 17 November 2026 HubSpot posted Marketing Studio and linked a beta product page. On 19 November Google Analytics posted that cross-channel budgeting supports source platform dimensions, and the linked help page says those dimensions are launching.

PUBLISHED //
DURATION // 21 MIN READ
A dark desk at night with a wide blank canvas of cards and a separate glass panel of muted budget bars, no text
FIG.01 // VISUAL DOCUMENTATION A dark desk at night with a wide blank canvas of cards and a separate glass panel of muted budget bars, no text

HubSpot posted on 17 November 2026 that Marketing Studio is the place to run marketing when a new AI tool arrives every day. The status, time-stamped 00:26 UTC, is two sentences and one product link. The link expands to the product page Marketing Studio. The page title is Marketing Studio | HubSpot, and the page marks the product Beta. The status does not list prices, plan names, agent names, or connector names. Those sit on the page. A video is attached to the status. It is not a source for the product description below.

Two days later, Google Analytics posted that cross-channel budgeting now supports source platform dimensions. The status, time-stamped 19 November 2026 at 15:31 UTC, says the dimensions are for deeper visibility and granular media optimization across ad platforms. The only article link on that status expands to https://goo.gle/4l5wJ74, which resolves to the help page About cross-channel budgeting. The help page does not print a publication date in the text used here. The dated public post is the status. A photo is attached to the status. It has no accessibility text in the post record, so it is not a source for the feature description below.

The two pages are treated as two documents. The HubSpot page is a product page for a campaign workspace. The Google page is a help article for a planning feature that, in its own words, will not affect budgets or spend in connected accounts. A pipeline figure on one page is not a measure of a budgeting model on the other, and a projection chart on the help page is not a description of Marketing Studio.

What the HubSpot page says the studio is for

The page headline under the Beta label is Marketing Studio. The line under it says to grow and convert demand from a single conversation with marketing studio. A button offers a demo. The next block is headed “Grow pipeline. Close deals. Drive revenue.” and prints three figures. It says 31% more marketing deals created with Marketing Hub, 29% more deals closed with Marketing and Sales Hub, and 36% more revenue with Marketing + Sales Hub. Those lines name Marketing Hub and Sales Hub. They do not say the percentages were measured on Marketing Studio alone, and they do not name a sample, a date range, or a control group. A later site-wide line on the same fetched page, in the Why HubSpot navigation, says that after just one year HubSpot customers acquire 129% more leads, close 36% more deals, and see a 37% improvement in ticket closure rates. That sentence is not inside the studio feature copy. A trust line on the studio page says trusted by 306,000 customers worldwide. The page does not say those 306,000 customers are Marketing Studio users.

The feature copy is more specific than the percentages. Under “Run campaigns from plan to proof” and “Turn insights into action,” the page says marketing studio flags opportunities before you go looking, like a drop in AEO visibility. Breeze Assistant uses brand, audience, and past performance to propose a campaign plan, then builds it on the canvas once you approve. The same page’s navigation describes a separate item, AEO (Beta), as answer engine optimization tools that track and improve a brand’s visibility in AI results. The studio paragraph uses a drop in AEO visibility as an example of a flagged opportunity. The AEO product page is not a source here. The only AEO wording in the record is the navigation description and that one example, so the example cannot be expanded into a method.

The next block says Breeze Assistant builds campaign automations without leaving the canvas where assets live. It says a team can run multiple automations in one campaign, trigger off asset interactions, A/B test branches, and use nurture agent to personalize emails for each recipient. A collaboration block says a team maps campaigns, leaves comments, assigns asset owners, and tracks progress on one shared canvas, in real time, without hunting through email chains or wondering who is doing what. A measurement block says marketing studio connects campaign activity to deals, so the team is not piecing together separate reports. It lists attributed revenue, ROI, deal influence, and lifecycle impact, for a single campaign or a whole program. Those four measures are named. The page does not define the attribution window, the deal object, or the formula for ROI.

Agents named on the page

Three agents are named in a section headed “Your marketing tools, working together.” Campaign Agent starts from a goal and plans the entire campaign based on brand, audience, and past performance. Content Agent creates blog posts, social content, and landing pages grounded in brand voice, audience data, and competitor research. Nurture Agent rewrites every nurture email for the specific contact receiving it, using CRM data, behavior, and live company research. A related-resources line says HubSpot’s campaign, content, and nurture agents plan, create, and convert in the same workspace. Those are product claims on the page. They are not a published score table, and the page does not print a benchmark, a latency figure, or an error rate next to any agent.

The frequently asked questions repeat the workspace claim in operational language. Marketing studio connects business context and campaign performance in one AI workspace. Studio uses that data to surface insights about where reach or engagement is being lost, then helps the team act on them directly. Breeze Assistant proposes a connected plan from a goal, a brief, or a live signal, then helps generate the assets and personalize the outreach. The person reviews, approves, and publishes, without switching tools. The audience answer says the studio is designed for marketing teams of all sizes who want to streamline campaign creation and execution, and that it is particularly valuable for teams that collaborate across functions, manage multiple campaigns at once, or need to move quickly from concept to launch. The problem answer says the studio helps a team improve campaigns over time because business context and performance data inform the next move, and that it surfaces gaps, recommends how to close them, and gives an AI workspace in which to act. The page does not say the recommendation is binding, and the how-it-works answer already places review and approval before publish.

Access, channels, and the connectors

The cost answer is the access rule, and the HTML prints it as two clauses with a period between them: marketing teams need Marketing Hub Pro or Enterprise to access marketing studio, and content agent and nurture agent within Breeze Assistant are credit-consuming features. The plan names link to a HubSpot pricing URL. That pricing page is not a source here, and the studio page text does not print a dollar price, a credit price, or a credit balance. The plan names are the access gate the page does print. Campaign Agent is named on the page and is not named in that credit sentence. Content agent and nurture agent are the features the cost answer marks as credit-consuming.

A separate answer says campaigns can use channels that are not HubSpot channels. The page says a team can map social platforms, ad networks, and whatever channels it uses, and can add links to designs, documents, or other assets that live outside HubSpot directly onto canvas cards, so campaign URLs sit in one view. That is a planning claim. It does not say HubSpot buys the ads, posts to those networks, or imports their spend. The Google Analytics help page, read separately, is the document that talks about importing cost and linking ad accounts. The HubSpot page does not cite that help page, and nothing on either page says the canvas cards and the Analytics cost import are the same object.

The last FAQ says marketing studio connects to Claude, Gemini, and ChatGPT through HubSpot’s native connectors. The page says a person can build campaigns, create assets, and pull live performance data in the AI interface already in use, and that everything appears back in marketing studio. Work can continue in the model interface or be picked up in HubSpot. The page does not name a connector version, a model version, or a rate limit. Those identifiers are absent from the page, so they are absent from this article. The closing line says marketing studio spots the opportunity, builds the plan, and tracks results to revenue, so the team can move from idea to launch without switching tools.

Related resources are listed and are not sources here: an Academy item titled Creating a Campaign in HubSpot, the Loop Marketing Playbook, and a Build Demand note about the agents. Their text is not used below. The copyright line on the page is Copyright 2026 HubSpot, Inc.

What the Analytics help page says budgeting already did

The help page title is About cross-channel budgeting. Its first caveat is eligibility. The feature may not be available to a given Google Analytics property. The page says the Google Analytics team is working to expand the feature to more properties, and that questions about eligibility go to the support team. Availability is therefore a property-level fact the page does not settle for any named account.

The definition sentence says Google Analytics cross-channel budgeting helps a team optimize paid channel investments and track performance across channels. It then splits the tool into two plan types. Projection plans help identify areas for improvement and support in-flight optimization decisions. A team can monitor how advertising channels are projected to perform against key performance indicators such as budget, conversions, and revenue. Scenario plans help analyze predicted return on investment at different budget levels and create media plans for future initiatives that the model ranks as the highest return for a defined budget. The page says to use Projection and Scenario plans together so planning and optimization stay in one loop. A benefits list names three jobs: evaluate cross-channel budgets with empirical evidence and data-driven insights, monitor in-flight pacing of spend and conversions weekly so in-quarter adjustments can follow overperformance or underperformance, and build annual or quarterly media plans by modeling allocations from historical data. The next sentence is a limit, not a benefit. These tools are for planning purposes only and will not affect budgets or spend in connected accounts. A recommendation on the page is not an instruction that Google Analytics sends to an ad account.

To view budgeting, the page says the Advertising section must be set up, including a link to Google Ads and conversions. The Advertising section is only available on the desktop version of Google Analytics. The path is the left menu, Advertising, then a report name in the Budgeting drop-down. The overview page shows saved plans and the import sources that power the models. It lists a table of the most recently created or modified plans, product links such as Google Ads, Display & Video 360, or any other product, and a cost import that shows usage, the date of the last upload, and a file match rate. That cost import is how campaign data for non-Google campaigns and channels is uploaded. Manual integrations, the page says, currently show static text that explains how to import ad events from third-party campaigns. In the future, that section will dynamically show which platforms have manual integrations available for budgeting models. The future tense is on the page. It is not a list of platforms, and this article does not supply one.

Projection, scenarios, and the dimension sentence

A projection plan monitors in-flight performance against a target key performance indicator. The page says it is meant to answer whether the plan budget will be spent, whether a target number of conversions will be driven, whether a target revenue will be reached, which channels are overperforming or underperforming, and where budget should shift. Creating one requires an eligible property and sufficient historical data. The form asks for a plan name, a planning period, a target indicator of Budget, Conversions, or Revenue, and a target value. The page’s own examples of those fields are a name such as My Q4 plan and a period such as Q4’25. Eligible conversions can be viewed from the historical data available. After creation, a chart uses a solid line for actual performance to date, a dotted line for how the plan is projected to perform, and a shaded area for model confidence bounds. If changes are not explained by seasonality or holiday effects, the intervals may be wide. The page reads a wide interval as a sign that performance patterns have changed and that the model may need more data before it adjusts. Recommendations at the top of the chart highlight the most and least efficient channels based on expected performance, using historical data and model estimates. The page says projection outputs are modeled estimates and should not be treated as guarantees of performance. If budgets change, it points to a Scenario plan for the remaining budget.

The page includes an example titled as an in-flight course correction. It is written as a problem, an insight, and a resolution. The problem is a 12-week media initiative with a $100,000 revenue goal, a projection plan, and under-pacing for both spend and revenue at five weeks. The insight says Channel A is the most efficient with 75 ROAS, while Channel B drives the lowest ROAS. The resolution shifts $5,000 from the Paid Social budget to Paid Search, and the projection model then shows the plan back on track for the revenue goal. Those figures sit inside an example on the help page. They are not a reported result from a named advertiser, and they are not used here as one. The example also shows the page’s own vocabulary for channels, Paid Social and Paid Search, which is separate from the source platform dimensions discussed later.

Scenario plans are for quarterly or annual allocation. They require data-driven attribution from the property, plus campaign data from linked integrations or data imports. Offline data is included when offline imports are set up. The target indicator choices here are Conversions or Revenue, which is a shorter list than the projection form, where Budget is also a target. The team selects a budget plan estimate and one conversion. The page says only one conversion can be selected at a time, from the conversions that have enough historical data to be eligible. The response curve shows the allocation the model treats as optimal for the defined budget, and the curve can be used to compare return at different budget levels. A separate view shows projected performance if no optimizations are made. Scenario plans can only start on future dates. If a plan is checked during or after its date range, the curve stays as it was when the plan was created, unless settings such as budget, date range, or target metrics are modified. The curve reflects market conditions at creation and does not update on its own. An updated curve requires a new plan. Once campaigns are live, the page sends the reader back to a projection plan for pacing. Scenario outputs are estimates based on the model, not guarantees.

A note on the same page says the budgeting models consider all click-through conversions, YouTube engaged-view conversions, and all cost data. It says to include impressions in a data import when available, to prepare for future updates. Impressions are therefore named as input the current note asks teams to keep, not as a result the current models already score in the text of this article.

The dimension sentence sits in the scenario section, after that note and before the scenario example. It says that, in order to improve the ease of use and actionability of budgeting plans, new source grouping dimensions, source platform dimensions, and alpha-only source platform by campaign type dimensions are launching. The page says these changes support analysis, optimization, and more standard reporting for both Google and other platforms. The status posted on 19 November says cross-channel budgeting now supports source platform dimensions, for deeper visibility and granular media optimization across ad platforms. The help page’s verb is launching, and it names two sibling dimensions the status text does not name: source grouping dimensions, and source platform crossed with campaign type, the last of those marked alpha-only. The status and the help page agree that source platform dimensions are part of the change. They do not use the same verb, and they do not list the same set. The help navigation also lists a page titled About the Source Platform dimension. That page is not a source here, so this article does not define the dimension beyond the launch sentence and the status.

The scenario example that follows is also an example. A new budget of $50,000 for the next year’s first quarter, described as $1,500 more than the prior quarter, is allocated across Search, Display, and Video. The recommendation in the example drives 3% more conversions by increasing Search by $1,000 and Video by $800 and decreasing Display by $300. The same rule applies as for the projection example. The numbers are the help page’s illustration of how a recommendation banner and a data table might read. They are not an audited media plan, and the page does not attach them to a customer.

Model names the help page does print

The methodologies section says that, depending on whether the property is Google Analytics or Google Analytics 360, plans are powered by either a DDA-based model or Google’s open-source Meridian model. The DDA-based budgeting model is described on this page as currently in beta and available to Google Analytics and Google Analytics 360 customers. The Meridian-based budgeting model is described as currently in alpha and only available to select Google Analytics 360 customers. The page points to further methodology pages. Those pages are not sources here, so the only model facts used below are the names, the beta and alpha labels, and the availability lines printed on the budgeting article itself. The page does not say which model produces the source platform dimensions, and it does not say that the dimension launch changes the beta or alpha status of either model.

Two Semrush posts between the two dates

On 18 November 2026, between the HubSpot status and the Google Analytics status, Semrush posted twice. The first, at 13:33 UTC, says ChatGPT has officially announced ads, calls that a major shift in the business model of the world’s leading AI assistant, and says the assistant is used by 800M+ people every week. It points to preparation steps and a tool. The only link expands to https://social.semrush.com/3X7mMw4. That page is not a source in this article, so the preparation steps and the tool are not described. The post record does include accessibility text for the attached image. The text says the infographic compares Google Ads and ChatGPT Ads in two columns, and that the differences it draws include search queries versus user prompts, keyword signals versus research prompts, and ad placement. It says Google Ads uses structured signals and direct ads, while ChatGPT Ads use conversational prompts and integrated content, and that the image highlights a need to understand audience needs, capture attention, test, and use creative strategies. The text says the infographic is from SEMrush. Those sentences describe the image Semrush attached. They are not an OpenAI product page, and no OpenAI page is in the record used here. The 800M+ figure is a figure in the Semrush post. It is reported as that post’s wording.

The second Semrush post, at 13:43 UTC the same day, says that last week a coalition of brands and industry leaders said AI search needs a measurement standard. It does not name the coalition, the date of the statement, or the standard they asked for. It then names a Semrush metric. AI-influenced traffic estimates the visits a brand earns from being mentioned or cited in AI answers. The post ends with the phrase “Not just the clicks.” The post record has no article URL. A photo is attached and has no accessibility text in the record, so the photo is not a source. A product version, a report name, and a method for counting later branded visits are not in the post text, and they are not used here. The metric is a name and a one-sentence definition, placed in public on the day between the HubSpot workspace post and the Google Analytics dimension post.

What a team can actually line up

The four posts and two opened pages can be lined up by date, and they cannot be lined up by a shared identifier. HubSpot’s page, dated in public by a status on 17 November, describes a beta workspace that proposes a plan from a goal, a brief, or a live signal, builds assets after approval, and says it connects campaign activity to deals. Access, on that page, requires Marketing Hub Pro or Enterprise, and content agent and nurture agent inside Breeze Assistant are marked as credit-consuming. Google’s help page, pointed to by a status on 19 November, describes planning models that estimate pacing and allocation and that explicitly do not change spend in connected accounts. It says source platform dimensions are among the dimensions being launched so Google and other platforms can be reported in a more standard way, and it keeps a separate alpha label on the cross of source platform and campaign type. Semrush’s posts on 18 November name a metric for visits from AI mentions and a claim that ChatGPT has announced ads, with an image whose accessibility text contrasts prompts and keywords.

None of the opened pages says it ingests the other vendor’s metric. None of the posts prints an account ID, a conversion name, or a field mapping that would join Marketing Studio’s deal influence to a Google Analytics scenario plan, or join either of those to Semrush’s AI-influenced traffic. The operational reading is that campaign assembly, paid-budget planning, and AI-answer measurement were each given a public product sentence in the same four-day window, and that a team still has to decide, outside these documents, which number is allowed to move a budget. That reading is an interpretation of the gap. It is not a feature any of the four posts announces. The help page already says its own numbers are estimates and that the planning tools will not spend money. The HubSpot page already says a person reviews and approves before publish. Those two limits are in the documents. A third limit is the absence of a join key. An absence is not a product, and it should not be briefed as one.

What these documents do not settle

The HubSpot page does not say when the beta ends, which duties sit only in Marketing Hub Pro and which sit only in Enterprise, or how many credits a content agent run or a nurture agent run consumes. It does not say that the 31%, 29%, and 36% figures are Marketing Studio results. It says they are results with Marketing Hub and with Marketing Hub plus Sales Hub. The Google page does not say which properties already have source platform dimensions, and it does say the broader budgeting feature may be unavailable on a given property. It does not define source platform in the article that carries the launch sentence. That definition is the job of a help title that is not in this record. It does not say the projection example’s 75 ROAS, or the scenario example’s 3% conversion lift, is a benchmark a buyer should expect. The Semrush metric post does not define the estimate, the lookback, or the set of AI answers it counts, and the ads post does not, in the text used here, list a price, a placement, or an eligibility rule for those ads.

A reader who needs those definitions has to leave this set of documents. What this set does settle is narrower, and it is enough to brief a team without borrowing numbers from outside it. Marketing Studio, as of the page linked on 17 November 2026, is a beta workspace gated by Marketing Hub Pro or Enterprise, with Campaign Agent, Content Agent, and Nurture Agent named, with connectors named for Claude, Gemini, and ChatGPT, and with a canvas that can hold links to channels outside HubSpot. Cross-channel budgeting, as of the help page linked on 19 November 2026, is a desktop Advertising report for eligible properties, with projection and scenario plans that do not spend money, with a DDA-based model in beta for Google Analytics and Google Analytics 360, and with a Meridian-based model in alpha for select Analytics 360 customers. The same page says source grouping dimensions, source platform dimensions, and an alpha-only cross of source platform and campaign type are launching. The public posts that point at those pages are the sources for the dates. The measurement sentence Semrush placed between them is a named metric and a short claim about ads, not a method note, and it should be briefed at that length.