The Ad Network Inside AI Chatbots Nobody Has Built Yet

01 /The Concept

Kontext is building what it calls "the world's largest advertising network" for AI chatbots – a single platform through which advertisers can reach audiences across the broadest possible range of AI chat applications, with maximum combined reach and impression volume.

The appeal of this ad format is that placements appear directly inside chatbot responses rather than as banner ads surrounding the chat window. A comparison in the startup's deck shows the difference vividly: traditional banner-style ads embedded in a chat interface (left) look jarring and cheap; Kontext-style in-response ads (right) feel natural – almost like a recommendation.

The scale of the opportunity is significant. AI chatbots already support roughly one billion ad impressions per day, and that number is growing at 20% per month.

Kontext starts at $3 CPM. The startup projects click-through rates of 3–5%, with cost-per-action running 2–4x lower than comparable placements on traditional websites or social platforms. One of the subtle advantages not spelled out explicitly: when ads are embedded in conversational AI responses, the AI can reformulate ad copy on the fly to better match the conversation's context – which should push CTR higher still.

In one early case study, a casual game developer ran campaigns through Kontext and saw CPM drop to $2.50 versus $7.35 on Facebook, while CTR jumped to 1.58% compared to 0.59% on Facebook.

For chatbot developers themselves, adding an ad layer meaningfully improves unit economics. Testing showed a 66% increase in average revenue per user once advertising was integrated.

Kontext's business model takes a 25–30% commission on ad spend, or alternatively a 20–30% revenue share calculated on the developer's customer LTV.

The startup sees itself becoming one of the dominant platforms for in-chatbot advertising – and has set its sights on the $100B+ scale. When its deck targeted a $5M seed round, the market interest was strong enough that the final close came in at $10M.

02 /Why It Matters

Kontext makes a case that mirrors the history of internet advertising: every dominant content format eventually spawned its own dominant ad platform.

- In the 2000s: websites and platforms like AdSense ($300B) and The Trade Desk ($70B).

- In the 2010s: mobile apps and platforms like AppLovin ($110B) and AppsFlyer ($2B).

- In the 2020s: AI chatbots are becoming the dominant interface – and the ad platform layer is still, for the most part, unoccupied. Kontext has drawn itself in on that map.

Regardless of whether Kontext specifically wins that position, the underlying argument is sound: advertising is coming to AI applications. The unit economics of AI apps almost make it inevitable.

Every AI application carries high variable costs – either paying for API calls to foundation models like OpenAI, or running proprietary models that require serious compute. The problem is that these costs don't scale linearly with subscriptions. A single power user with complex, frequent queries can make a subscription unprofitable all by themselves.

Set subscription prices too low, and heavy users will destroy the margins. Set them higher, and you thin the user base dramatically. Introduce usage tiers, and the complexity itself scares away mass-market users who feel constrained.

Consumers are now conditioned to expect AI tools to be free or very cheap. Extracting real revenue through subscriptions alone is extremely difficult. An advertising model may not be optional for many chatbot developers – it may be the mechanism that keeps the lights on.

What's particularly clever about the ad model for AI apps is that impressions scale with queries, not just with user count. A sudden spike in usage doesn't blow the margin – it proportionally increases ad revenue to offset the increased compute cost.

KOAH ([covered here](/review/nakonec-to-ii-prilozhenie-mozhno-prevratit-v-masshtabiruemyj-biznes)), covered in late 2024, is working on the same problem – a platform for in-chatbot advertising – though Kontext lists it as a competitor without much visible momentum on their end.

03 /Opportunities

The overall conclusion: surviving on subscription revenue alone is genuinely hard for most AI application developers.

That said, positioning yourself to become "the largest" platform for in-chatbot advertising globally is an extremely ambitious and risky goal – especially as major AI players like OpenAI and Perplexity are already moving to introduce their own native ad systems, which may eventually open to third-party developers as well.

A more realistic path for AI app developers: start experimenting with advertising revenue directly, rather than waiting for a platform like Kontext to do it for them.

Most AI applications are vertical – they serve a specific user base with a specific set of needs. That naturally limits the pool of relevant advertisers, which is actually an advantage: it means the audience is highly targeted and engagement rates should be strong.

So what's stopping the developers of specialized AI applications from building their own list of potential advertisers and approaching them directly?

The technical implementation isn't especially complex. The harder part is the decision to do it and the operational work of building a proper advertising sales process.

And the economics can be significantly better than routing through a platform like Kontext – no commission to pay, plus the ability to build custom ad formats that leverage the app's specific context to command premium rates.

Even today, with all the reach of programmatic platforms like Google AdSense, plenty of publishers still manage their own ad inventory directly – because the economics are simply better when you own the relationship. Why shouldn't AI app developers do the same?

The exercise is worth doing: map the user profile of your AI app, then list the brands with budget to reach that profile. The formats and price points follow from the conversation – and the economics can be significantly better than routing through a platform like Kontext, with no commission to pay and full control over the relationship.