When we launched Ant10, the idea was simple: media planning shouldn’t require days of spreadsheets, research and back-and-forth before a marketer can see a workable plan.
Ant10 was built to turn a campaign brief into a structured media plan — helping marketers move from a campaign objective to media recommendations, budget allocation and a plan they can review and refine.
If you’re new to Ant10, we’ve previously explored how AI can be used for media planning and how Ant10 approaches AI-powered media planning in India.
But building an AI media planning tool doesn’t end at launch.
Once people start using a product for real campaign briefs, they find things that aren’t always obvious during development.
Some recommendations need refinement. Some workflows need to be easier. Some assumptions that look reasonable in theory don’t hold up in actual use.
That’s exactly what happened with Ant10.
The latest version includes four significant improvements:
- A better mobile experience
- More accurate media recommendations
- Better validation of active media
- Smarter editing and context retention
Here’s what changed — and why it matters.
1. Ant10 is now built for mobile users too
One of the first things we noticed after launch was something we hadn’t expected to be as significant: nearly 20% of Ant10 users were accessing the tool from mobile devices.
The earlier experience wasn’t good enough.
A responsive website isn’t necessarily a good mobile product. Media planning involves budgets, tables, recommendations and multiple pieces of information that need to be easy to read and interact with.
So we improved the experience with mobile usage in mind.
The result is a mobile-optimised Ant10 experience, making it easier to access and work with media plans when you’re not sitting in front of a laptop.
This matters because media planning doesn’t always happen at a desk.
A marketer may want to:
- Check a recommendation before a meeting
- Review a media mix while travelling
- Make a quick change to a campaign brief
- Revisit a plan from a phone
- Explore media options before building a detailed plan
The goal is simple:
Ant10 should work where marketers work.
2. More accurate media recommendations
AI can produce an answer very quickly.
But speed isn’t particularly useful if the recommendation isn’t accurate.
Some early Ant10 users pointed out that the quantities of certain media recommended in a plan weren’t always accurate enough.
This is an important distinction in media planning.
Recommending Outdoor is one thing.
Recommending a specific quantity of outdoor inventory that makes sense for the campaign’s geography, budget and objective is another.
A useful AI media planning tool has to understand the difference.
We’ve therefore refined the underlying recommendation logic to improve how Ant10 evaluates media quantities and builds its recommended plans.
The objective isn’t simply to generate more recommendations.
It is to generate recommendations that are more useful as a starting point for an actual media plan.
And this is where Ant10’s underlying media intelligence matters.
The Media Ant has built a database containing 3.5 lakh+ advertising options, giving Ant10 a much more specific media universe to work with than a general-purpose AI system that is simply asked to suggest advertising channels.
A general AI tool can tell you that outdoor advertising might work for a campaign.
A purpose-built media planning tool needs to help answer a much more practical question:
Which media opportunities actually make sense for this brief?
3. Fewer AI hallucinations — and better validation of active media
This is perhaps the most important change.
One of the challenges of using generic AI for media planning is that an answer can sound completely plausible while still being impossible to execute.
We encountered exactly this problem in Ant10.
Some recommendations included inactive media — including examples such as IPL 2026 inventory after the tournament had already ended.
For a media planner, that’s not a minor error.
A media recommendation is useful only if the opportunity actually exists and can be considered for execution.
So we’ve improved Ant10’s validation process to check whether media opportunities are active before they are included in recommendations.
This is an important part of building an AI media planning tool for the real world.
AI hallucination in media planning is different from a wrong answer
If a chatbot gives you an incorrect fact in a general conversation, you can correct it.
In media planning, an incorrect recommendation can have practical consequences.
It can lead to:
- Time spent researching unavailable inventory
- Incorrect assumptions during budget planning
- Unnecessary back-and-forth with vendors
- Plans that need to be rebuilt
- Loss of confidence in the AI-generated recommendation
That’s why freshness and validation matter alongside intelligence.
The goal isn’t to make Ant10 sound smarter.
The goal is to make it more trustworthy as a planning starting point.
This builds on a distinction we’ve discussed in our earlier article on AI media planners and how Ant10 approaches media planning: a useful AI media planner needs more than generic marketing knowledge. It needs media-specific planning logic and relevant information about actual advertising opportunities.
4. Smarter editing: because the first plan is rarely the final plan
No media planner expects the first version of a plan to be the final version.
Budgets change.
Geographies change.
A client asks for more TV.
Someone wants to remove outdoor.
The campaign duration changes.
The target audience gets narrower.
A new business requirement appears halfway through the conversation.
This is where conversational AI should have a real advantage over traditional planning workflows.
But that only works if the AI understands what has already been decided.
We’ve therefore improved Ant10’s edit logic, making it easier to refine a plan while retaining the context of the existing conversation and recommendations.
Instead of treating every edit as a completely new request, Ant10 is better at understanding that the user is modifying an existing plan.
That’s an important step towards making AI useful beyond the initial generation of a media plan.
Because the real workflow isn’t:
Brief → Plan → Done
It’s:
Brief → Plan → Question → Edit → Compare → Refine → Finalise
And that’s how we want Ant10 to work.
What these four updates tell us about building an AI media planning tool
These changes might look like four separate product updates.
But together, they point to something bigger.
Building an AI media planner isn’t just about putting a chatbot interface on top of media data.
There are at least four different problems to solve.
1. The AI needs to understand the brief
What is the objective?
Who is the audience?
Where are they?
What is the budget?
What does the campaign need to achieve?
2. The AI needs access to relevant media intelligence
A useful media plan needs more than generic marketing knowledge.
It needs access to real media opportunities, along with information such as audience, geography, pricing and availability.
This is one of the fundamental differences between a general-purpose AI chatbot and a purpose-built AI media planning tool.
A chatbot can suggest that a brand consider television, outdoor, digital or radio.
A media planning tool needs to take the next step: evaluate those options in the context of the actual campaign.
3. The recommendations need to be executable
A theoretically suitable channel isn’t enough.
The media needs to be relevant.
The opportunity needs to be active.
The recommendation needs to make sense for the budget and campaign.
This is particularly important when AI is being used for a task that eventually leads to real-world spending.
4. The AI needs to work with the planner, not just answer the planner
The first answer is only the beginning.
The ability to ask questions, change assumptions and refine the plan without starting again is what turns an AI generator into a planning assistant.
What we learned from putting Ant10 in the hands of users
One of the things we’ve learned since launching Ant10 is that product development doesn’t happen entirely inside a product roadmap.
It happens when someone says:
“This recommendation doesn’t look right.”
Or:
“Can I edit this?”
Or:
“I’m trying to use this on my phone.”
Or simply:
“Why is it recommending something that isn’t available anymore?”
Those questions are valuable.
They tell us where the gap is between what an AI tool can generate and what a marketer actually needs.
The latest Ant10 improvements came directly from that gap.
And that’s an important part of building AI products: the first version gives you a product; real users help you build a better one.
From generating plans to making planning better
The first version of Ant10 was about solving the blank-page problem.
Give the tool a campaign brief and get a structured starting point.
The latest updates take the product a step further.
It’s not just about generating a plan anymore. It’s about making that plan:
More accessible — with a better mobile experience.
More accurate — with improved recommendation logic.
More reliable — with better validation of active media.
More flexible — with smarter editing and context retention.
That distinction matters.
Because the goal of AI in media planning isn’t simply to produce an answer faster.
It’s to make the planning process itself easier to navigate.
What’s next for Ant10?
We’re treating this as an ongoing process rather than a finished product.
As more marketers, media planners, agencies, founders and businesses use Ant10, their feedback will continue to shape how the product evolves.
The questions we receive are often as useful as the answers we build.
Because ultimately, we’re not trying to build an AI that simply knows about media planning.
We’re trying to build one that can help people actually plan media.
And that means continuing to improve the details that matter in the real world — accuracy, availability, context, usability and the ability to turn a recommendation into something a marketer can actually work with.
Try the updated Ant10
If you haven’t tried Ant10 yet, this is a good time to explore it.
Describe your campaign brief, including your objective, target audience, geography and budget, and see how Ant10 turns it into a structured media plan.
And if you’ve already used Ant10, keep the feedback coming.
The latest improvements — from better mobile usability and recommendation accuracy to active-media validation and smarter editing — have been shaped by how people are actually using the tool.
Try Ant10 →
More on Ant10
New to AI-powered media planning?
- Ant10: The AI Media Planner Powered by The Media Ant’s Planning Expertise — Understand how AI can be applied specifically to media planning.
- Ant10: The AI Media Planning Tool Built for How India Actually Advertises — Explore the thinking, data and planning expertise behind Ant10.
- Try Ant10 — Put your campaign brief into the updated Ant10 experience.
