India’s digital ad spend is expected to top Rs 69,000 crore by the end of 2026. At present, video formats take up more than half of display budget, and CTV ad spend is seeing a healthy 22% YoY growth. Retail media is actually growing even faster, at 24.2%. Formats like pre-rolls, mid-rolls, native in-feed, rewarded video, shoppable units, interactive overlays and short-form interstitials are enhancing advertisers’ ad stacks. The pipes have never been wider or more varied.
And yet, a typical digital ad barely grabs 2 seconds of someone’s attention. Two-thirds of consumers don’t notice banner ads (‘banner blindness’) and nearly one-third of internet users globally use ad blockers. More formats, more inventory, more sophistication in delivery and the viewer is still tuning out.
In the last ten years, our industry has really focused on getting good at figuring out who to show an ad to. We’ve barely even started to scratch the surface of “when” and “how” to show it.
The Attention Spectrum
There’s a concept that truly deserves more discussion in our media planning meetings than it gets and that’s the “attention spectrum.”
Not all impressions carry the same weight of human attention. A person watching a long-form documentary on CTV at 9 PM is in a fundamentally different neurological & emotional state than someone scrolling through a news feed on their phone during an office lunch break. A viewer browsing a quick commerce app clearly has intent, but absolutely zero patience. A reader halfway through a 2000-word article on a Sunday morning has plenty of time but will deeply resent being interrupted.
The ad tech ecosystem tends to treat all these moments as “inventory.” But they’re really not. They’re distinct states of human attention, ranging from a relaxed, immersive lean-back immersion, to an active, lean-forward intent or even just a passive scroll-through. And an ad format that might work wonderfully in one of these attention states can actually damage the experience in another.
A 30-second unskippable pre-roll makes perfect sense on CTV. The viewer is settled, committed, watching something they deliberately chose. But put that very same format before a 90-second news clip on a mobile phone at 7 AM during a morning commute. That’s not advertising; that’s punishment. The viewer will likely remember the frustration long after they’ve forgotten which brand caused it.
The Journey Mismatch Nobody Talks About
Here’s something the industry really needs to hear more clearly: the most crucial factor in whether an ad “works” isn’t necessarily how good the creative is or even how accurate the targeting is. It’s about whether the format was right for where the viewer was in their content journey at that exact moment.
Someone who just opened an app is often in a discovery mindset, quite open and receptive to new information. A well-placed native unit or an in-feed placement that’s truly relevant to the context can feel natural, almost like part of the content itself. A viewer who’s 15 minutes into a video is in immersion mode. A mid-roll, timed just right, can echo the TV ad breaks they grew up with and they’ll often accept it because the format respects the rhythm of their experience. But a viewer who’s three paragraphs deep into an article they truly care about, suddenly hit with a full-screen interstitial that takes four seconds to get rid of? They’re not being advertised to; they’re being taxed for reading.
We’ve seen the same ad creative, with identical copy and targeting parameters, achieve a 68% completion rate in one placement and only a 12% completion rate in another, all on the same publisher’s site. The audience wasn’t different. The difference was how the format met the moment. One placement honoured the viewer’s attention state; the other completely ignored it.
Most ad serving infrastructure today doesn’t account for this. The decisioning engines tend to optimize for audience match and bid price. Very few truly optimize for aligning with that attention state. That’s the real gap. And it’s proving to be a pretty costly one.
What Actually Works
Criticism without a clear viewpoint is just noise. So here is what the data and operational evidence across publisher environments in 2026 reveal about which format-moment combinations are working and the underlying pattern.
CTV mid-rolls in long-form content consistently show some of the best completion rates in the market. It’s not just because the screen is bigger but because the implicit attention agreement between the viewer and the content is much deeper. The viewer chose to be there. The format mirrors a familiar rhythm. The interruption feels proportionate to the commitment.
Rewarded video in gaming and utility apps tends to outperform almost every other mobile format. And it’s not only because the creative is inherently better, but because the viewer opted in. They made a conscious decision to trade their attention for something of value. That simple act of consent completely changes the dynamic. The ad is no longer an interruption; it becomes a transaction the viewer initiated.
Short-form interstitials on mobile apps can work, but only when they’re deployed once per session. Stack them and they absolutely destroy the user experience. Publisher A/B testing has repeatedly shown this: a single interstitial per session helps maintain user engagement and encourages them to return. Introduce two interstitials in the same session and you’ll see bounce rates spike and next-day retention visibly drop. The line between “that was fine” and “I’m deleting this app” is often just one extra ad.
In-feed native ads on editorial platforms are effective when the content is truly contextual. A financial services brand appearing within a market analysis article feels relevant. The same brand dropped into a cricket highlights page with zero editorial connection does not feel like content. It feels like a disguise. And viewers, it turns out, are much better at spotting those disguises than most media plans give them credit for.
The pattern is quite clear: ad formats that genuinely align with the viewer’s attention state and offer a reason for them to stick around will always do better than those that just rely on interruptions, sheer frequency or forcing exposure. The latter might generate impressions but they also generate resentment. And that resentment builds up, often invisibly, until it surfaces in metrics nobody likes to prioritize: declining session duration, fewer return visits, more people using ad blockers, and a noticeable drop in organic traffic.
The Real Backfire
Here’s the part that needs to be said plainly: when ad formats are deployed without any real awareness of the user’s attention state, the cost isn’t only borne by the advertiser but also the publisher. The advertiser can move their budget to the next platform. The viewer shifts their attention somewhere else. However, the publisher is left with dwindling audiences and a monetization engine that’s been optimized for a user base that’s quietly shrinking.
The pattern is quite predictable. Aggressive format deployment might bring in strong revenue in Q1. But then, audience erosion starts to show up in Q3. By Q4, the publisher is scrambling to hit the same revenue targets with a smaller, less engaged audience, which often means deploying even more aggressive formats. This just accelerates the erosion further. The whole cycle only breaks when someone starts asking a different question: not “how do we fill this inventory,” but “how do we make the viewer actually want to come back tomorrow”
The Decade Ahead
The ad tech industry spent the 2010s building precision in who to reach. The 2020s have been about scaling the infrastructure to reach them everywhere. The next phase, the one that will separate the companies that endure from those that plateau, will be about precision in how and when.
The ad tech industry spent the 2010s perfecting the precision of who to reach. The 2020s have largely been about building out the infrastructure to reach them everywhere. The next phase, the one that will truly distinguish the companies that thrive from those that just stagnate, will be all about precision in how and when.
The “how” refers to the format. The “when” is about the attention state. Getting both of these right is what makes an ad feel like part of the experience. Getting either one wrong is what gives someone a reason to leave.
In a market quickly approaching Rs 2 lakh crore, the true opportunity isn’t about serving more ads. It’s about matching the right format to the right moment, within the right state of human attention. The platforms, publishers and advertisers who build for this will earn something no targeting algorithm can ever buy: the viewer’s active choice to return.
That is the highest-performing ad format there is. And it does not have a CPM.
