News hook: InMobi Advertising announced on August 13 that it has placed a conversational AI agent inside Buyer Hub, its self-serve programmatic deal platform, while opening that platform to agencies and managed-service demand-side platforms. The release matters because it moves AI from campaign ideation into the operational layer where media teams discover inventory, build deals, forecast supply, and optimize live buying.

InMobi says the agent works through natural-language instructions and recommends curation and optimization moves grounded in campaign context. The trader still approves the decision. At the same time, Buyer Hub adds role-based permissions, inventory discovery, self-serve reporting, and a live Deals Forecasting panel. In other words, this is not simply another chatbot for marketers: it is an attempt to make the supply side of programmatic advertising easier for agencies to operate directly. (InMobi/Business Wire announcement)

1. The real product is workflow compression

The headline feature is the conversational agent, but the bigger change is the removal of coordination steps. Buyer Hub debuted in June 2025 with curators as its first users. The new agency opening lets trading desks reach InMobi inventory directly instead of relying on repeated email exchanges or a managed-service handoff for every deal adjustment.

That matters to agencies managing multiple clients, markets, and permissions. A natural-language interface can help a buyer find relevant supply or ask for a deal configuration, while role-based access controls determine who can view or edit each account. Self-serve reporting also puts delivery diagnostics closer to the people responsible for pacing and performance. The result is potentially faster execution, but only if the underlying inventory, identity, measurement, and approval data are reliable.

2. Forecasting before activation changes the buying decision

InMobi’s Deals Forecasting panel may be more strategically important than the chat interface. It provides a live estimate of available ad opportunities while a deal is being configured, before the campaign is activated. That shifts a common failure point in programmatic buying earlier in the process: a team can discover whether a target has enough addressable supply before committing budget and explaining underdelivery to a client.

This is the direction AI advertising infrastructure is taking. The useful system is not the one that produces the most fluent recommendation; it is the one that connects audience, supply, availability, and performance signals at the moment a decision is made. InMobi’s related partnership with Scope3 opened premium CTV and mobile in-app supply, including Glance, to AI-powered buying agents. Together, those moves suggest a strategy built around both human traders and machine-to-machine transactions. (Adweek coverage of the Scope3 partnership)

3. Human approval is becoming a competitive feature

Kunal Nagpal, InMobi Advertising’s chief business officer, said the company built the agent so traders do not have to choose between speed and control: “It recommends the curation and optimization moves, and the trader still makes the call.” That human-in-the-loop model is not a minor implementation detail. It is a response to the commercial risk of giving an agent authority over targeting, supply quality, brand safety, and live spend.

For agency leaders, the question is not whether an agent can write a recommendation. It is whether the recommendation can be audited, challenged, approved, and tied to a measurable business outcome. InMobi’s announcement does not disclose pricing, customer counts, or an independent performance lift for the agent. That means buyers should treat the launch as a workflow test, not as proof that autonomous optimization will outperform an experienced trader.

4. What agencies should do now

Agencies do not need to hand over campaign decisions to benefit from this shift. They should build a controlled pilot around one repeatable workflow:

  • Choose a bounded use case: Start with inventory discovery, deal setup, or delivery diagnostics rather than unrestricted budget changes.
  • Define approval gates: Require a human sign-off for targeting changes, new supply sources, bid or budget changes, and any action that affects brand safety.
  • Measure the workflow, not the novelty: Track time to launch, time spent on troubleshooting, forecast accuracy, deal win rate, delivery, CPM, conversion quality, and client-facing transparency.
  • Protect client boundaries: Use account-level permissions and document which data the agent can access, retain, or use to make recommendations.

The strategic lesson from InMobi is clear: AI is moving below the campaign dashboard and into the plumbing of advertising transactions. The agencies that win will not be the ones that automate the most steps. They will be the ones that connect automation to clean data, explicit guardrails, and accountable decision-making.

Ready to turn AI marketing tools into a measurable growth system? Real Internet Sales helps businesses and agencies build practical AI marketing, search, and automation strategies. Call 803-708-5514 or visit realinternetsales.com.