Unlock your business potential with a full-funnel Digital Strategy

The working paid media stack now spans seven categories: paid search (Google, Bing), paid social (Meta, LinkedIn, TikTok, Snap, Pinterest, X), video (YouTube, connected TV platforms like Hotstar, JioCinema, Netflix Ads, Disney+), retail media (Amazon Ads, Flipkart, Myntra, Blinkit, Instamart, BigBasket Now), programmatic display and video through DSPs (DV360, The Trade Desk, Amazon DSP), audio (Spotify Ads, Apple Podcasts), and increasingly, AI search surfaces (Perplexity, ChatGPT, Bing Chat). Most brands run four to six of these actively at any time. The mistake is treating each platform as a separate campaign rather than as part of a single audience journey – the platforms a buyer encounters in week one shape the conversion that happens in week three on a different platform.

Demand-side platforms like DV360, The Trade Desk, and Amazon DSP let brands buy ad inventory across display, video, audio, and CTV automatically, bidding on individual impressions in real time. That means showing up for the right audience across millions of sites and apps, not just the ones you have a direct relationship with. Programmatic complements paid social and search rather than replacing them: search captures active demand, social builds and converts demand within feed environments, and programmatic builds reach and prospects audiences across the long tail of the internet. The most-asked question is whether programmatic still works in a post-cookie world – the honest answer is yes, but the practice has shifted heavily towards contextual targeting, retail media DSP integrations, and first-party data activation rather than third-party cookie-based segments.

Retail media is advertising on retailer-owned platforms – sponsored product listings on Amazon, Flipkart, Myntra, Blinkit, Instamart, BigBasket Now – that target shoppers at the point of purchase intent. It is the fastest-growing channel in performance marketing because it captures buyers who have already entered the consideration window, with shorter paths to conversion than any other paid surface. The category has grown from a niche on Amazon to a $150B+ global market because every major retailer now monetises its first-party data through an ad network. The strategic shift for brands: retail media is increasingly the budget battle that determines category share, not just a sales acceleration channel – if a brand isn’t visible on the retailer surfaces where its category is bought, no upper-funnel investment compensates.

CTV (connected TV) and OTT (over-the-top) advertising runs across streaming platforms like Hotstar, JioCinema, Netflix Ads, Disney+, YouTube on TV, and the increasing range of ad-supported tiers across SVOD services. The buying happens either directly through platform self-serve interfaces (YouTube, JioCinema, Hotstar) or programmatically through DSPs with CTV inventory. Brands should invest in CTV when they need reach with brand-building creative at TV-quality contexts, when their audience has shifted away from linear TV, or when they need premium video inventory to complement performance-led social and search. Treating CTV as a performance channel is a category error – it works as measurable upper-funnel video, but expecting last-click ROAS from a 15-second pre-roll on streaming is the wrong evaluation frame.

Publishers, creators, and comparison sites earn commission on conversions they drive. That is the affiliate model, and it tends to work well for D2C brands with clear conversion paths and meaningful basket sizes. It is rarely a primary acquisition channel, but it complements paid search and social effectively through comparison content and creator-led recommendations that drive consideration. Affiliate earns its place in the mix when partner quality is strong and tracking is clean – and rarely otherwise.

Attribution hasn’t fully broken, but it has fragmented – iOS signal loss, cookie deprecation, and walled gardens make any single attribution model an incomplete view. The modern measurement approach combines three layers: in-platform metrics for short-term campaign optimisation, incrementality testing for causal lift, and media mix modelling (MMM) for cross-channel budget allocation. The L&F Weavr platform was built specifically to unify these signals into a single view, pulling data from every paid platform a brand runs and reconciling it against business outcomes. The honest framing brands need to accept: no system delivers perfect attribution, and the most expensive mistake is spending years trying to build one rather than operating well with imperfect information.

Media mix modelling is a statistical technique that measures the impact of marketing channels on business outcomes using historical data – typically 2 to 3 years of weekly spend and outcome data across paid media, organic channels, pricing, distribution, and external factors. Multi-touch attribution (MTA) works at the user level, tracking individual journeys across touchpoints and assigning conversion credit. The two answer different questions: MTA tells you which ad in a user’s journey deserved credit; MMM tells you how much each channel actually drove revenue at a portfolio level. With signal loss eroding MTA accuracy, MMM has returned to prominence – particularly modern MMM tools like Meta’s Robyn, Google’s Meridian, and commercial platforms like Recast and Aryma that compress MMM cycles from quarterly to weekly.

Incrementality testing measures the causal lift a paid media channel delivers – the conversions that would not have happened without it – by comparing outcomes between exposed and unexposed audiences. ROAS measures attributed revenue divided by spend, but treats every attributed conversion as if the ad caused it. The gap between the two is often significant: incrementality studies regularly show that 30% to 50% of ‘ROAS-attributed’ conversions on lower-funnel paid search and remarketing would have happened anyway. This matters because brands optimising on ROAS without incrementality checks tend to over-invest in channels that capture demand rather than create it – which is why most performance-led brands eventually hit a growth ceiling that more spend doesn’t fix.

iOS 14.5’s App Tracking Transparency and the gradual deprecation of third-party cookies in Chrome have together removed the user-level identifiers that drove targeted advertising and conversion tracking for over a decade. The modern response operates on three fronts: server-side conversion tracking (Conversions API on Meta, Enhanced Conversions on Google, Stape and similar GTM server containers) to recover lost signal, first-party data activation through customer data platforms (CDPs) and CRM integrations, and measurement approaches that don’t depend on user-level tracking – incrementality and MMM. Brands waiting for signal to come back are operating on a flawed premise: signal loss is permanent, and the operational discipline of working with it is now the differentiator between strong and weak media practices.

The Conversions API is Meta’s server-side method for sending conversion events directly from a brand’s servers to Meta – bypassing the browser-based Pixel which is increasingly degraded by iOS tracking restrictions, ad blockers, and consent management. Google has an equivalent in Enhanced Conversions, and most major platforms now offer similar server-side APIs. Implementing CAPI typically recovers 20% to 40% of conversion signal that the browser-side pixel alone misses, with the recovery rate higher for iOS-heavy audiences. The implementation cost is moderate – 1 to 3 weeks of engineering depending on stack complexity – and the ROI is usually clear within 30 days through improved attribution and algorithm performance. Skipping CAPI in 2026 is operating with a measurable handicap.

There is no universal answer, but the working framework for most consumer brands starts with category and funnel logic, then layers in business stage. For D2C brands with strong product-market fit, a 50-30-20 split between paid social (Meta, TikTok), paid search (Google, Bing), and reach-building channels (CTV, YouTube, programmatic) is a reasonable baseline. Retail-heavy brands should shift 20% to 40% of budget to retail media (Amazon, Flipkart, Blinkit, Instamart). Brands in awareness-building stages skew towards reach channels; brands in efficiency stages skew towards demand capture. The dangerous version of channel allocation is letting last-click attribution drive the split – it concentrates spend on demand-capture channels and starves the channels that create demand in the first place.

The honest answer most agencies won’t give: in-housing makes sense for brands with the scale, talent budget, and operational depth to run media as a permanent capability. The threshold is usually around $5M+ in annual digital media spend, a dedicated head of growth or media, and the willingness to invest in measurement infrastructure beyond what comes free with the platforms. Below that threshold, agencies typically deliver better outcomes because the same brand cannot afford to hire specialists across every channel – the depth penalty of generalist in-house teams costs more than the agency fee they save. The hybrid model – in-house strategy and decisioning with agency execution and specialist channels – works well for brands transitioning between the two.

Most agencies price in one of three ways. Some charge a percentage of media spend, typically 8% to 15%, with the rate coming down at higher spend tiers. Others work on a monthly retainer, usually somewhere between $3K and $30K depending on scope. The third model is outcome-based, a base retainer with a performance bonus tied to specific KPIs. Under $50K a month in paid media spend, a flat retainer tends to make the most sense. Between $50K and $500K, percentage-of-spend or hybrid structures are more common. Above $500K, brands usually negotiate fixed-fee arrangements with defined service-level agreements. The cheapest agency is rarely the most economical – lost performance from weak optimisation costs more than fee differentials in most engagements.

AI search and AI Overviews are absorbing high-intent queries that previously drove paid search clicks – some categories now see 20% to 40% of informational queries resolved without a click-through. The paid media response operates on three fronts: shifting budget from generic informational keywords (which AI absorbs) towards branded and high-intent transactional terms (which AI still routes to advertisers), investing in retail media surfaces where AI Overviews don’t yet compete, and treating brand search visibility as a paid media outcome rather than only an organic concern. The deeper shift: paid media is increasingly being asked to compensate for organic visibility loss to AI – and brands building GEO authority simultaneously protect both their organic surface and their paid efficiency.