Sitecore just announced at its Symposium in Orlando, Sitecore MarketingIQ, a module in SitecoreAI that connects AI-search signals with approved brand knowledge, then proposes content and campaign work for marketers to review, edit and approve. The company's own example is the one web teams will recognize: an AI answer omits a brand or lists outdated pricing. Scrunch, a Sitecore company, flags the issue. MarketingIQ checks approved product information and existing content, then stages a proposed fix. Sitecore's wording is direct: "nothing leaves a staging area without human involvement."

That is a useful next step after measurement. Earlier this week we looked at ChatGPT's Business Hub, which lets a claimed business see aggregated mentions, website clicks and discovery topics. MarketingIQ is aimed at a different job: once you know an answer engine has left you out or gotten a fact wrong, how do you turn that gap into governed work in the CMS?

CMSWire's Dom Nicastro, writing from the same conference, frames the launch as Sitecore putting Scrunch to work. Sitecore acquired Scrunch in June 2026 to track how brands appear in AI-generated answers. MarketingIQ is the follow-through: findings become staged responses, and results are supposed to feed the next recommendation.

How A Gap Becomes Staged Work

Sitecore describes MarketingIQ as two capabilities working together: contextual intelligence, meaning an understanding of the brand and its business, and AI agent orchestration, which coordinates the work needed to act. According to the announcement, it connects activity in AI search platforms such as ChatGPT and Google Gemini with approved knowledge and content, business objectives, evidence about customers, and results from previous campaigns. From those, it decides which facts support a claim, which audiences and products they apply to, and what content is suitable to use.

Work can start two ways. A marketer can ask MarketingIQ to pursue a business goal, or the system can surface an opportunity from customer behavior, market changes or campaign results. In either case, Sitecore says it finds current, approved content, identifies what can be adapted, proposes changes and flags gaps that need new work. It can generate new content where required, always subject to review.

The product page puts the flow in five steps: a goal or signal, contextual intelligence, agent orchestration, an approved outcome, then learning from what performs. Agents prepare a brief and draft the content and experiences "people and their agents will see." The team guides and approves the work before it goes live. After approval, MarketingIQ is supposed to coordinate the people and agents needed to finish it. Teams can then measure whether visibility improved, whether qualified traffic followed and whether that traffic converted — Sitecore's claim about the loop, not a result we measured.

Sitecore also says MarketingIQ works across Sitecore and third-party CMS, digital experience and business systems, without replacing the underlying CMS, DAM, CRM or other applications. The product page adds that teams can use the full SitecoreAI platform or connect MarketingIQ to an existing setup through Model Context Protocol, with the team still setting direction and approvals. We have not tested that MCP path.

What Ships, And What Is Still Planned

The announcement lists five supporting capabilities. Four are described as part of the launch: access to supported Sitecore capabilities from ChatGPT, Claude, Microsoft Teams and Slack, with permissions and approvals carried through; connecting Scrunch findings to the response work, including pages designed for AI agents; guided experiences that start from an outcome; and closer ties between content operations and digital asset management, so people and agents can use approved assets, usage rights and production workflows.

The fifth item is explicitly planned, not shipping. Sitecore says future capabilities "would let customers authorize their AI agents to take actions on a brand's website, such as requesting a demonstration or registering for an event," subject to permissions and required confirmations. CMSWire's editor's note makes the same distinction. We are not treating those customer-agent site actions as available today.

Customer quotes in the release should be read as vendor and customer claims. IFS chief marketing officer Oliver Pilgerstorfer says the value is acting on visibility findings and turning that work into "qualified pipeline" and "revenue growth." Sitecore CEO Eric Stine says the company is "turning insight into action and that action into revenue." Perficient's Carrie Grapenthin, speaking about the firm's own SitecoreAI website, says it "launched in just under 10 weeks." Those are not results we verified.

Availability, Packaging And Open Questions

Sitecore's press release says MarketingIQ "will be available to all customers of SitecoreAI." The product page is only slightly more specific: existing SitecoreAI customers "will get access," and the organization still has to set up connections and permissions before a team starts using it. That is a planned rollout, not a claim that every tenant already has it switched on.

Packaging is clearer than a public price. The product page says MarketingIQ is a module within SitecoreAI products, is included with SitecoreAI CMS and SitecoreAI DAM, and "isn't sold on its own." An XM or XP license on its own does not include it; Sitecore tells those customers to talk to their account team. We did not find a public dollar price or a standalone SKU. CMSWire had flagged whether MarketingIQ would carry its own price as an open question from the announcement. The product page now answers the packaging part. It does not publish a rate card.

What This Means For Content Teams

The DevContentOps problem here is familiar even if the trigger is new. An outdated price in a ChatGPT or Gemini answer is often a content problem first: the canonical fact was missing, stale or hard for an answer engine to trust. A dashboard that only reports the miss is not a workflow. A draft that publishes itself is not a workflow either. The useful pattern is the one Sitecore is describing: detect the gap, compare it with approved source content, stage a change, and require a person to approve it before it goes live.

What Web Teams Should Do

1. Separate measurement from action. Use first-party tools, including ChatGPT Business Hub where you have a claimed profile, to see when an answer engine cites you. Treat a miss or a wrong fact as a ticket that needs a content owner, not only a visibility report.
2. Keep canonical facts in the CMS. Prices, policies, product names and claims should live in approved entries, not only in a slide deck or a prompt. If an AI draft is going to propose a fix, it should be checked against that source.
3. Require staging and human approval. If you evaluate MarketingIQ, or build a similar path on another stack, make sure nothing publishes, archives or overwrites a live page without a person signing off.
4. Confirm packaging before you budget. MarketingIQ is not sold standalone, is included with SitecoreAI CMS and DAM, and is not included with XM or XP alone. Ask what "will get access" means for your tenant, and what connections and permissions you must set up.
5. Do not plan on customer-agent site actions yet. Demo requests and event registration initiated by a visitor's AI agent remain a planned capability.
6. Measure the loop yourself. Sitecore says teams can track visibility, qualified traffic and conversion after a change. Define those metrics in your own analytics before you treat the loop as closed.

MarketingIQ will not, by itself, make ChatGPT or Gemini get your brand right. It is Sitecore's bid to turn an AI-answer gap into the kind of work content teams already know how to govern: a staged update, reviewed against approved facts, published on purpose. That is the part worth testing, when access arrives, against a real stale fact on your own site.