The media agency business is getting harder to defend on the ground it once owned: the mechanics of buying media. As self-serve platforms steadily take over tasks that once required agency intervention, the industry is confronting an uncomfortable question: how much of media buying still needs an intermediary?
The pressure is also visible in the growing consolidation among media agency networks. Most recently Omnicom merged Hearts & Science and Mediahub to launch Hearts United, a new global media agency that operates across 40 markets, with $9.1 bn in 2025 billings. The move is part of a broader push by holding companies to bring capabilities, data, technology and buying power under fewer, larger structures. Interestingly media agencies are fortifying themselves with scale which will reflect in the buying process even as platforms are making provisions for the marketers to go the DIY way.
The question is can scale and strategic intelligence become the agency’s biggest USP in the years to come. As platforms making buying easier, what else will media agencies need, to make themselves indispensable for marketers?
The buying process
Earlier, a marketer would brief an agency, which would negotiate with media owners, compare rates, reserve inventory, manage contracts and secure placements. Every campaign passed through several hands before it appeared on television, in print or on a billboard. Digital advertising initially replaced some of that process with another layer of intermediaries: demand-side platforms (DSPs), supply-side platforms (SSPs), ad exchanges, verification partners, data platforms and measurement vendors. The machinery became more sophisticated, but the dependence on specialists remained.
Now, platforms are attempting to take even more of that machinery away from the media agency. Businesses can increasingly log into dashboards, select audiences, set budgets, buy inventory and monitor performance themselves. Google and Meta have long normalised this model, while newer self-serve offerings are pushing it into premium video, OTT, commerce, connected TV and other environments once dominated by agency-led buying.
The proposition is simple: fewer people, lower overheads, faster execution and greater control. For a neighbourhood business, a D2C brand or even a large advertiser testing a new channel, the ability to buy media directly can be compelling.
While self-serve advertising platforms pose a growing threat to digital media buying, other channels such as television and out-of-home (OOH) advertising still rely heavily on media agencies for planning, negotiation, inventory management and execution. However, the increasing dominance of digital could put the role of media planners under greater pressure. According to the Pitch Madison Advertising Report (PMAR) 2026, India’s advertising market is projected to reach `1.55 lakh crore, with digital already accounting for nearly 60% of the market. As more digital ad buying moves towards automated, self-serve platforms, the industry could see a gradual erosion of the manual planning and buying functions that have traditionally formed the core of media agencies’ value proposition.
Where does this leave media agencies then? Tejas Maha, Associate Director - Media, White Rivers Media, believes this is pushing agencies further upstream. “Agencies now function as strategic partners rather than simple media operators. Brands handle execution, so agencies focus on audience strategy, platform governance and creative testing,” he says. With digital channels increasingly at the centre of media planning, he argues that agencies are being pushed towards designing customer journeys, building attribution models and translating data into decisions. “This advisory role drives better business outcomes,” he stresses.
A campaign begins with a click
Imagine a furniture retailer in Jaipur launching its festive sale. A decade ago, the owner would probably have approached a local agency, discussed media plans over several meetings and waited days before seeing the campaign go live. Today, the same retailer can open an advertising dashboard. The budget is entered. The target audience is selected. Campaign objectives are defined. Artificial intelligence suggests creative formats, audience segments and bidding strategies. Within minutes, the campaign is ready to compete for attention alongside some of India’s largest brands.
That democratisation of advertising is precisely what companies such as JioStar, which launched their self-serve platform prior to the IPL seaon, believe is reshaping the industry. The broadcaster’s introduction of self-serve advertising reflects a broader effort to lower the barriers to premium video advertising. Initially, they started campaigning at `4,000 as a limited-time special offer during IPL, allowing small and medium businesses to access audiences that were once largely the reserve of national advertisers. But campaigning prices have risen to `50,000 since then.
That evolution extends beyond streaming. Professional platforms, messaging apps and retail media networks have embraced similar models, enabling businesses to launch campaigns independently while retaining visibility over spending, targeting and performance.
Vikas Iyer, Marketing Head, Lenexis Foodworks, says self-serve platforms have made experimentation, optimisation and speed significantly easier, but their biggest value is improving execution rather than replacing marketing strategy. “For a mass QSR brand like Chinese Wok, growth still comes from building broad mental availability through consistent reach and distinctive brand assets, and self-serve platforms help us execute faster without changing the fundamental job of making the brand easy to notice and easy to choose,” he says.
But self-serve does not benefit every advertiser in in the exact same way. For smaller brands, it solves an access problem. Platforms cannot economically provide dedicated sales or account-management resources to every advertiser spending a few lakh rupees in a single market, but a dashboard can. For large advertisers, however, the equation becomes more complicated. Their scale gives them access to relationships, customised packages, negotiated rates and strategic support that a standardised self-serve interface may not replicate.
Shekhar Saurabh, Head of Marketing, Tata AIG, describes it as a question of scale versus volume. “If someone is spending Rs 20 crore, Rs 30 crore or Rs 40 crore, it may not make sense to buy purely through a self-service platform because direct relationships can bring better packages, add-ons or commercial value,” he says, adding, “For a smaller advertiser or a new brand spending Rs 1 lakh or Rs 2 lakh in one geography, self-serve can be a very good deal. It would not be cost-effective for a platform to allocate dedicated manpower to such an advertiser.”
For Piramal Consumer Healthcare, meanwhile, a majority of media buying now happens through self-serve platforms in one form or another. Abhishek Kumar Srivastava, Vice President, Marketing, says performance-led channels such as e-commerce and quick commerce are largely managed in-house because they require agility, close coordination with business teams and faster optimisation cycles. For top and mid-funnel campaigns, the company continues to work closely with agency partners.
The dividing line, therefore, is not simply between brands that use agencies and brands that do not. It is increasingly between the kinds of work brands need agencies for and the kinds of work they can comfortably bring in-house.
Does doing it yourself actually save money?
Control and speed are among the most obvious attractions of self-serve advertising. A marketer does not have to wait for an intermediary to make a change. Results are visible immediately, budgets can be moved quickly and campaigns can be tested continuously.
Anmol Sahai Mathur, VP - Marketing, MARS Cosmetics, said in an interview, the brand uses self-serve platforms across different stages of the consumer journey, from awareness to conversion. “Self-serve platforms have given brands much more control, and we use them across planning, optimisation and measurement. This has helped us test faster, understand consumer behaviour in real time and make more informed, data-driven decisions efficiently,” he notes. The biggest advantage, he adds, is agility. “I can experiment, learn, refine communication and reach my audience much faster than before, when traditional marketing often came with significant setbacks.”
But self-serve does not automatically mean cheaper. It can remove layers of operational dependency, but it does not remove the need for expertise. The cost simply moves. A brand may save on agency fees but spend more on building an internal team, acquiring specialised talent and learning through experimentation. For advertisers without sufficient scale or internal capability, that trial and error cost could outweigh the savings.
Girish Hingorani, Vice President, Marketing Head (Unitary Cooling Products) & Corporate Communications, Blue Star, believes the ability to operate a platform does not necessarily mean a brand will get the best outcome from it. “Every rupee that Blue Star spends on advertising goes through our media planning agency. We have directly experimented with self-serving and realised it is not really worth it unless you have specialised in-house resources to monitor campaigns day in and day out,” he comments.
The calculation, he argues, is ultimately an in-house versus outsourcing debate. “If I need media planning activities in-house, I have to hire people with the expertise to do it. When I compare that cost with the agency commission for the same service, what is cheaper for me?” he asks. That makes the economics of self-serve more nuanced than simply comparing platform costs with agency fees. For a small advertiser that cannot justify an agency relationship, self-serve can be dramatically more efficient. For a large brand that already has specialist teams or agency partners working at scale, doing everything internally may create a different set of costs.
Following the advertisement
The technology behind a self-serve dashboard is considerably more complex than the interface suggests. It is important to distinguish between self-serve advertising and programmatic advertising. Programmatic buying has been reshaping digital media for more than a decade. The newer development is not the existence of automated buying, but who is increasingly operating it.
The technology that once sat largely with agency trading teams and specialist media buyers is being placed directly in front of marketers.
Behind the simplicity of a self-serve dashboard lies one of advertising’s most sophisticated technological systems. The advertiser may click on ‘Launch Campaign’, but that action can set off thousands of automated decisions in milliseconds.
The demand-side platform or DSP, acts on behalf of the advertiser. Rather than purchasing fixed advertising slots, it evaluates available impressions across websites, apps, connected televisions and streaming platforms and decides which opportunity is most likely to achieve the campaign objective.
On the other side, sits the supply-side platform, representing publishers and media owners seeking to maximise the value of their inventory. Connecting the two is the ad exchange.
Every time a consumer opens an app, loads a webpage or starts a video, these systems can conduct an auction in a fraction of a second. Advertisers compete for an individual impression, with algorithms evaluating signals such as audience characteristics, contextual relevance and historical performance before determining the winning bid.
The consumer never notices. The advertisement simply appears. What once required extensive planning and negotiation can now happen automatically. The difference is that the machine is accessible to the advertiser itself.
But automation alone does not solve the bigger question of what happened after an impression was served. The value of connecting adtech, programmatic buying and audience data lies increasingly in what those systems can tell marketers about the path from exposure to conversion.
Maha says the three now work as parts of the same feedback loop. “Adtech captures performance data. Programmatic technology automates buying decisions in real time across channels. Data-driven targeting identifies audiences using first-party insights,” he explains. Together, these systems can create a clearer trail from exposure to conversion and support attribution beyond simple clicks. In the Indian market, he says, this is increasingly becoming a more disciplined approach to planning, measurement and optimisation.
According to Rohan Chincholi, Chief Digital Officer at Havas Media India, today’s advertising ecosystem has evolved into an interconnected technology network extending far beyond display banners. Modern programmatic platforms now span video, audio, connected television, digital out-of-home media and commerce touchpoints, enabling brands to optimise campaigns continuously rather than relying on static media plans. The result, he says, is advertising that is more data-driven, measurable and outcome-focused than previous generations.
Iyer sees a similar change in the way media is used by QSR brands. “Programmatic has shifted the conversation from buying audiences to buying moments and contexts. For QSR, relevance often comes from being present around consumption occasions rather than narrowing audiences too aggressively, so we increasingly use technology to improve contextual relevance and media efficiency while maintaining broad category reach,” he shares, adding, “The objective isn’t precision at the cost of scale; it’s combining both intelligently.”
Automation is becoming the default
Industry estimates suggest programmatic advertising already accounts for around 42 per cent of India’s digital advertising expenditure, a figure expected to climb to 44 per cent by 2026. Digital advertising now contributes almost 60 per cent of India’s `1.55 lakh crore advertising market, making automation less of a competitive advantage and more of an operational necessity.
According to Muralidhar T, Senior Vice President, Media Delivery and Operations at WPP Media India, agencies are no longer distinguished simply by their ability to execute campaigns. As self-serve platforms automate campaign setup, audience deployment, bidding and optimisation, agencies are moving towards helping brands navigate complex technology ecosystems, interpret data and translate it into business decisions. Execution, he argues, is becoming technology-led. Intelligence is becoming the differentiator. Rather than competing with automation, agencies are repositioning themselves as technology enablers, strategic consultants and growth partners.
That view is echoed at Schbang, where National Media Strategy Lead Deepshikha Bhardwaj believes buying media is no longer the industry’s defining skill. The real value lies in understanding consumer behaviour, asking better business questions and connecting media, creativity, commerce and technology. Technology may automate execution, she argues, but judgment and cultural understanding remain human strengths.
Will media agencies be replaced?
The answer emerging from the industry is neither a straightforward yes or no. The pressure is greatest in digital-first businesses where the entire customer journey is already visible to the brand. Pranay Rao, Chief Marketing Officer, Sebamed India, sees self-serve as a definite threat, particularly for D2C brands. “It’s definitely a threat with D2C brands who, thanks to having visibility of the entire journey, don’t see value in external planners when they have their own teams. Eventually, a brand with the right setup and infrastructure can run its entire digital spends in-house. TV, print and outdoor would still need planners,” he observes.
But even Rao expects a hybrid model rather than an immediate elimination of agencies. “We will, like most organisations our size, shift to a hybrid model initially. Agencies will need to exhibit knowledge gained through running multiple campaigns and also a capability to work across the funnel,” he adds.
Where can the agency beat the machine?
If everyone has access to broadly similar dashboards, technology alone becomes less of a differentiator. The agency’s advantage therefore has to move towards areas where automation is weaker: strategy, cultural understanding, creativity, interpretation and the ability to connect media decisions to wider business objectives.
A machine can optimise towards an objective. It cannot independently decide whether that objective is the right one. It can identify patterns in consumer behaviour. It cannot always explain why a particular cultural moment matters. It can test thousands of creative variations. It cannot reliably determine which unexpected idea will become memorable before the audience has seen it. It can report a return on advertising spend. It cannot necessarily tell a brand whether the campaign built long-term demand.
For Deepshikha Bhardwaj, this is where human judgment remains central. Media planning increasingly involves understanding the consumer, asking the right business questions and connecting media with creativity and commerce rather than simply finding the cheapest impression.
At Social Beat, Co-founder Vikas Chawla points to another change. Optimisation has moved inside platforms through tools such as Performance Max and Advantage+. Marketers now spend less time adjusting bids and more time improving creative, first-party data and business inputs.
This means the human advantage is increasingly found in deciding what the machine should optimise towards in the first place.
When creativity becomes the competitive advantage
As media buying becomes automated, creativity is quietly becoming more valuable. For years, agencies were judged partly by their ability to negotiate rates and execute campaigns. Self-serve advertising has commoditised much of that work. The competitive edge is moving elsewhere.
For Aalap Desai, Co-founder and Chief Creative Officer at tgthr, optimisation and creative thinking belong to two different worlds. Technology can identify patterns and improve execution, but its logic is built on averages. Creative ideas, by contrast, often emerge from breaking those averages. Self-serve platforms may give everyone similar access to media, but they cannot replicate experience, intuition or the ability to produce an idea that surprises people. Today’s data-rich environment is also changing how campaigns are built. Rather than developing one universal campaign, agencies can create multiple creative assets tailored to different audience segments, interests and behaviours.
For Namrata Chawla, Co-founder of mktgstack, technology has made advertising dramatically easier to access. Almost anyone can launch campaigns, analyse dashboards and optimise performance. What it cannot provide is taste, cultural understanding or human insight. The risk, she argues, is that brands begin evaluating every campaign through the same performance lens.
Advertising has never had a single purpose. A performance campaign may exist to generate immediate conversions. A brand film may exist to build memory. An influencer collaboration may strengthen credibility. Measuring all of them against identical return-on-investment expectations can create misleading conclusions.
The agency’s role, therefore, begins long before a campaign launches. It involves helping brands define realistic objectives, establish meaningful success metrics and understand that not every valuable outcome appears immediately on a dashboard.
What happens when self-serve becomes AI-serve?
The next phase of self-serve advertising could require even less hands-on campaign management. Today, a marketer operates the machine. The marketer chooses the objective, sets the budget, selects or approves audiences, uploads creative and monitors performance.
The next version could look very different. A marketer may simply define the business objective, provide the relevant customer, product and creative inputs, and allow an AI agent to decide how the campaign should be executed. The system could select audiences, allocate budgets, adjust bids, test creative variations, move spend between platforms and optimise towards the business outcome.
In other words, self-serve could evolve from marketers operating advertising platforms to marketers briefing AI systems that operate them.
Rao believes agentic AI and the emerging Model Context Protocol (MCP) could make this increasingly possible. “Through agentic AI and the new MCP, we can at least run digital campaigns independently. At the end of the day, it will remain an ROI and ROAS discussion. Whichever model gives better value will be adopted,” he shares.
He also believes the expertise required from marketers could change as platforms increasingly use creative signals to determine targeting. If machines become better at deciding which consumers are likely to respond to a particular piece of creative, the marketer’s job could shift from manually defining narrowed audiences towards producing and selecting the right creative inputs.
This is already visible in the way brands are using AI across the media cycle. Audience discovery, bid optimisation, budget allocation, forecasting, creative testing and reporting are increasingly automated.
At Lenexis Foodworks, Iyer says less than 20 per cent of the company’s media spend currently goes into programmatic platforms, but AI-led optimisation has improved the speed of experimentation significantly, with CPMs improving by 20 to 30 per cent. “We typically test anywhere between five to 20 creatives for each campaign in the initial phase and then invest in the top 20 per cent performing ones, with AI playing a critical role in identifying which assets are working best,” he says.
As Chawla puts it, AI has already won the mechanical battle. No human can outperform a machine at bidding, audience segmentation or real-time optimisation. The unresolved challenge lies elsewhere. Algorithms optimise towards whatever signal they are fed. If that signal is flawed, AI can simply scale the wrong outcome with extraordinary confidence. The marketer’s role, therefore, could become less about operating tools and more about deciding what success should actually look like.
The last unsolved problem
If automation has made targeting and optimisation increasingly efficient, attribution remains advertising’s biggest unanswered question. How do brands know which touchpoint deserves credit for a sale?
Today’s customer journey rarely follows a straight line. Someone may discover a product through an influencer, later encounter a Connected TV advertisement, receive a retail media promotion on Blinkit, search for reviews on Google and finally complete the purchase after seeing a retargeted Instagram advert.
Which advertisement deserves the credit? There is still no universally accepted answer. The problem becomes even more complicated as more of the advertising journey takes place inside walled gardens such as Google, Meta, Amazon, Blinkit and Zepto, where the same platforms control audience, inventory, targeting and measurement. As Chawla points out, optimisation has increasingly moved inside these platforms. Marketers may spend less time manually adjusting bids, but they still have to determine whether the metrics being reported represent genuine business growth.
Privacy regulation and signal loss make the challenge harder. The technology may be getting better at deciding how to deliver an advertisement, while the harder question remains: how accurately are we measuring what happened afterwards? As Gozoop Creative’s Country Head Amyn Ghadiali observes, data identifies the audience, programmatic determines how and when to buy impressions, while adtech connects the system and reports what happened afterwards. The strength lies in the feedback loop created by these technologies, where every campaign generates new learning.
Perfect attribution, however, remains elusive. Anyone claiming to have solved it completely, he suggests, is probably selling something. The industry’s objective, therefore, is shifting from finding a perfect number to identifying a clear direction. And this is where the agency’s role could become more important, not less. If platforms are becoming capable of executing and reporting on their own environments, brands still need someone capable of connecting those signals across channels, benchmarking performance and asking whether the reported outcome actually reflects the business outcome.
The future belongs to judgment
The story of self-serve advertising is often framed as one of automation replacing people. The evidence suggests something more complicated. Technology has democratised access.
For media agencies, however, the threat is real, but more specific than outright replacement. The routine executional layer of media buying is becoming increasingly automated. The agency’s survival depends on proving its value beyond the buttons. That value lies in deciding what the machine should optimise, understanding consumers beyond the data points, creating ideas that algorithms cannot derive from averages, connecting fragmented media environments and determining whether a campaign achieved something meaningful for the business.
The next generation of self-serve advertising may not remove the human from the process. It may simply move the human farther away from the mechanics. The marketer may eventually brief the machine rather than operate it. The machine may buy the media, allocate the budget, optimise the campaign and even generate much of the creative.
What remains is the harder question: what should it be trying to achieve? As Ghadiali puts it, self-serve platforms have democratised the buttons, but not the judgment. In many ways, advertising has come full circle. The advertisement still begins with an idea; it simply travels a far more intelligent route before reaching its audience.
