Programmatic digital advertising agency for accounts where identity stitch and supply discipline decide outcomes.

The 2026 buyside is a four-platform race:

• DV360: default for GMP-stack buyers. Broadest single-DSP CTV access. Only DSP shipping Topics API + Protected Audience API targeting in live campaigns by default.
• The Trade Desk: strongest independent. Kokai UI + AI bidding default. UID2 majority adoption. OpenPath compresses take rates.
• Amazon DSP: forced consideration since Prime Video Ads scaled. Singapore, Australia, Japan, India live markets. Amazon shopper signals unlock cookieless targeting.
• Yahoo DSP: performance and CTV in US/EU.

Long tail: Adobe Advertising Cloud (captive AEC tool). MediaMath: bankruptcy 2023, deprecated.

Kokai is The Trade Desk's current UI and bidding engine, now the default. Brings forecasting plus AI bidding plus a redesigned campaign management workflow. UID2 adoption is layered in natively. OpenPath enables direct bidding into publishers without SSP intermediaries, materially compressing take rates on premium open-web.

Practitioner upgrade path from older Trade Desk workflows is straightforward, but AI-bidding settings need calibration against historical performance; do not enable Kokai's bid recommendations on a cold account without 30 days of conversion data to anchor against.

Google has effectively abandoned forced third-party cookie deprecation in Chrome. The Privacy Sandbox page now states some Sandbox technologies are being phased out, the inverse of the original posture.

Cookies in Chrome persist by user choice. Privacy Sandbox APIs (Topics, Protected Audience, Attribution Reporting) remain available and continue to mature, but adoption is now market-led not deadline-led.

The identity-graph investment thesis does NOT reverse: those identifiers were always needed for Safari, Firefox, in-app, and CTV. EU and Singapore regulatory direction still demands consented first-party identity.

• UID2 (Trade Desk-led): primary cross-DSP open-web identifier. Majority adoption among premium US publishers, growing in APAC.
• RampID (LiveRamp): offline-to-online matching, particularly strong with retail, financial services, and CPG.
• ID5: dominant European cookieless ID.
• LiveRamp ATS: publisher-side authentication.
• Hashed-email infrastructure: underlying authentication layer.
• Contextual targeting (IAB Content Taxonomy 3.0 plus GumGum, Seedtag, Captify LLM classifiers): cookieless complement.

Yes, where the DSP supports them and the audience signal justifies the layer. DV360 is the only DSP shipping Topics API and Protected Audience API targeting in live campaigns by default in 2026; Trade Desk and Amazon DSP have varying degrees of Sandbox API support.

The Sandbox reversal removed the forced-march deadline, not the underlying utility. Topics API is genuinely useful for cookieless contextual signal. Protected Audience API enables remarketing without third-party cookies.

Practitioner posture: layer Sandbox APIs where they add signal, do not over-commit infrastructure to a deadline that is no longer real.

IAB's 2026 outlook projects digital video ad spend exceeding $80 billion in 2026, with social, CTV, and commerce media driving 9.5% ad spend growth.

Premium-CTV inventory buyable programmatically:

• Netflix Ads: DV360, Microsoft, Trade Desk
• Disney+ + Hulu: DRAX programmatic platform, integrated with all major DSPs
• Amazon Prime Video Ads: Amazon DSP-first, third-party access via SSP partnerships
• Roku: OneView DSP plus open programmatic
• Samsung Ads: deep ACR-data targeting
• Warner Bros. Discovery / Max: programmatic on PMP basis

Premium CTV is increasingly biddable but PMP and programmatic guaranteed still dominate. Open-exchange CTV remains thin and largely lower-tier FAST channels.

Amazon Prime Video Ads launched 2024 and fully scaled across 2025. It is now the single largest premium-CTV pool outside YouTube.

The unique unlock versus other premium CTV is Amazon shopper signals: browse history, purchase data, and Prime Video streaming behaviour can target audiences without third-party cookies.

Live markets: Singapore, Australia, India, Japan. Case studies cite up to 20x ROAS using contextual plus shopper-signal targeting.

For any brand selling product or running performance media, Amazon DSP plus Prime Video is now a default consideration line on the 2026 plan.

MFA sites are publishers built primarily to arbitrage ad spend: high ad density, low content quality, often AI-generated "AI Slop" content, sometimes outright bot fraud. They look legitimate in dashboards but deliver minimal brand or performance value.

In 2026 Made-for-Advertising filtering is a standard practice, not an emerging issue. Every plan goes through MFA filters (DoubleVerify's, IAS's, Jounce Media's) before launch.

Supply Path Optimisation compresses inventory to 5-8 named SSP paths per buyer (down from 30+ in 2022). PMPs and curated marketplace deals capture share from open exchange.

The 2026 threat vectors are AI-generated "AI Slop" content schemes and evolving bot-fraud rings.

SPO is the discipline of compressing the number of SSP and reseller paths a buyer uses to reach the same publisher inventory. The 2022 norm was 30+ supply paths per inventory source; the 2026 operator default is 5-8 named paths.

Compression reduces duplicate bidding, eliminates resellers who add no value, lowers take rates on premium open-web, and concentrates spend on publishers with direct relationships.

The Trade Desk's OpenPath bids directly into publishers without an SSP intermediary, materially compressing take rates.

SPO is the single highest-leverage hygiene move on most legacy programmatic accounts.

The 2026 stack:

• Comscore: cross-screen reach + frequency + attribution (mainstream measurement standard post Nielsen's currency stumbles)
• Comscore: household-level CTV measurement
• IRIS_ID: universal content ID enabling consistent measurement across walled gardens
• ACR data: Samsung Ads, Vizio Inscape, LG Alphonse. ACR-fuelled household graphs now dominant cross-CTV identity
• DoubleVerify + IAS: viewability, IVT, brand-safety with attention frameworks
• MMM + incrementality testing: IAB-codified supplements to last-touch attribution

Nielsen lost sole-source TV currency status across 2024-25; the market is now multi-currency by default.

Practitioner-level bands for premium PMP inventory at scale (; open-exchange runs lower, walled-garden CTV runs higher):

• Open-web display PMP: SG $4-8, MY $2-5, AU $6-10, US $8-15, CA $7-12
• Premium CTV (Netflix, Disney, Prime): SG $35-55, MY $25-40, AU $45-65, US $50-80, CA $45-70
• Mid-tier CTV (Roku, FAST): SG $20-35, MY $15-25, AU $25-40, US $25-45, CA $25-40
• pDOOH: SG $10-25, MY $8-18, AU $15-30, US $12-25, CA $12-22
• Online video (open-web pre-roll): SG $8-15, MY $5-10, AU $10-18, US $12-22, CA $10-20

Treat as planning ranges; rebaseline from account history within 30 days of launch.

pDOOH is the programmatic buying layer for digital out-of-home inventory: digital billboards, street furniture, transit screens, shopping mall displays.

• Vistar Media: global pDOOH DSP/SSP/ad-server stack, broadest geographic footprint
• Place Exchange (acquired by Broadsign): integrates with DV360 and major DSPs, ships PerView measurement
• DV360, Trade Desk, Yahoo DSP all support pDOOH inventory directly

In 2026 pDOOH is a planned line on most major plans for Australia and the United States; Singapore and Malaysia pDOOH inventory is developing but inventory is genuinely buyable.

Agentic AI ad-buying tools are now a real category in 2026, not pilots. Buyers are deploying guardrails and contracts now require explicit human-control clauses.

Our posture: AI does the heavy allocation and bidding work inside the DSP; senior practitioner judgement sets the strategy, approves the creative, locks the brand-safety, and reads the incrementality.

For regulated clients, no agentic AI step ships creative or makes budget allocation decisions without a senior-practitioner review gate in the loop. The 2026 default is human-in-the-loop, not human-out-of-the-loop.

The regulated entity is responsible for every ad served regardless of who placed it. Operational implications:

• Brand-safety exclusion lists: mandatory and audited (DoubleVerify + IAS deployed pre-launch, not post-incident)
• Creative-rotation logs: retained per record-keeping obligations
• Suitability targeting locked: no inferred-income or vulnerable-audience targeting without explicit policy clearance
• CTV PMPs preferred over open exchange: premium-publication context guarantee

For Singapore FI clients: MAS FAA-N03 effective March 25 2026. For Australia: ASIC RG 234. Healthcare brand-safety adds HIPAA-equivalent restrictions in US and additional MOH/HSA review in SG; programmatic for pharma is operationally CTV-PMP and contextual-display only.

Look for three signals:

• Regulated-sector programmatic experience. Most Singapore agencies have e-commerce programmatic experience but limited fluency with MAS FAA-N03 and the brand-safety discipline FI clients need.
• Technical depth on the 2026 identity-graph stack (UID2, RampID, ID5, LiveRamp ATS) plus MFA filtering plus SPO discipline plus CTV PMP relationships.
• Senior practitioner involvement on the account.

Our leadership team brings over 50 combined years:

• Siddharth Surana, Founder & CEO, 18+ years (Programmatic Pioneer APAC 2011, ex-Havas Regional CDO, ex-Media360 COO)
• Sundeep Surana, Managing Director, 16+ years
• Ratnakar Nemani, Operations Director, 11+ years, Google Ads Certified
• Nitesh Sanghvi, Search & Social Director, 11+ years, Google Ads & Google Analytics certified

We work with Travel Guard Singapore and have anonymised proof: 6x digital sales growth for a regional banking institution across six APAC markets over seven quarters.

Every engagement is scoped to the data, industry, and market. We don't publish standard rates because no two engagements are the same. Talk to us about your specific challenge and we'll come back with a scoped proposal.

Written findings covering DSP setup health, identity-graph integration (UID2/RampID/ID5/LiveRamp ATS), MFA exposure assessment, SPO path inventory, CTV PMP deal inventory, brand-safety stack configuration, creative inventory + rotation cadence, attribution model assessment, cross-screen measurement integration (Comscore/IRIS_ID/ACR), regulated-sector compliance assessment (MAS FAA-N03 for SG FIs), 90-day execution plan. 2-3 weeks. fixed. Document yours regardless. Founder participates.

Three patterns: audit-only, embedded strategist, or full takeover. We do not bid against the incumbent in pitches or steal accounts mid-contract.

Yes, when the incumbent contract is up. The handover process: 1-2 week diagnostic on account state, formal handover of DSP and SSP access, parallel-run period, full operational responsibility at the agreed transition date. We do not poach accounts mid-contract.

A senior practitioner from the leadership team is on every account. Strategic direction is set by Siddharth Surana (Programmatic Pioneer APAC 2011), Sundeep Surana, Ratnakar Nemani, or Nitesh Sanghvi. Team based in Singapore and India. One team, two time zones, no white-labelling.

Monthly: cost per outcome by channel (display, online video, CTV, audio, pDOOH); viewability and brand-safety rate (DV + IAS); MFA exposure rate; SPO path concentration; cross-screen unduplicated reach (Comscore + IRIS_ID); CTV attention metrics.

Quarterly: incrementality testing where volume supports it, MMM refresh, competitive share-of-voice analysis on premium CTV inventory.

We do not report on vanity metrics (raw impressions, raw clicks) without context.

Three triggers.

1. The platform stack has consolidated under you (the 4-Demand-Side-Platform race after MediaMath's bankruptcy mid-2023, Amazon Prime Video Ads at scale since 2025, the Privacy Sandbox plot reversal of April 2025 where Google abandoned forced cookie deprecation) and the team has not refreshed the DSP-mix decision against current reality.
2. Made-for-Advertising exposure is climbing without your team catching it. Made-for-Advertising sites are low-quality, ad-stuffed publishers that drain spend without delivering business outcomes. Senior-grade discipline now is 5-8 named supply-side paths via Supply Path Optimisation, with Made-for-Advertising filters from DoubleVerify, Integral Ad Science, and Jounce Media deployed pre-launch, not post-incident.
3. Connected TV is now meaningful on your plan but your team has never run cross-screen attribution via Comscore plus IRIS_ID plus Automatic Content Recognition household graphs.

The audit reads which of the three is actually breaking your performance. Holding-company agencies often have the third issue (Connected TV measurement) regardless of in-house presence.

Programmatic's strategic edge is scaled reach across the open web, premium Connected TV, programmatic audio, and programmatic Out-of-Home with cross-screen attribution that paid social and search cannot match.

Programmatic deserves the next dollar when:

• Your bottleneck is brand reach against specific audience segments at scale where Meta and Google have hit a frequency ceiling
• Your bottleneck is Connected TV exposure on streaming (Netflix Ads, Disney+, Amazon Prime Video, Roku)
• Your bottleneck is regulated-sector brand-safety guarantee at scale (use Private Marketplace deals first plus DoubleVerify and Integral Ad Science filtering)

If your bottleneck is bottom-funnel performance with strong intent signal, search wins. If pure direct-response retail, Meta retargeting beats programmatic display.

Three layers.

1. Business outcomes. Cost per outcome by channel (display, online video, Connected TV, audio, programmatic Out-of-Home), brand-lift score where Brand Lift Studies ran, contribution to pipeline or revenue via a Marketing Mix Modelling line item.

2. Attribution caveats stated up front. Cross-screen attribution via Comscore plus IRIS_ID plus Automatic Content Recognition household graphs is the 2026 standard (Nielsen lost sole-source TV currency in 2024-25; the market is now multi-currency by default). Marketing Mix Modelling plus incrementality testing supplements last-touch attribution, which under-credits Connected TV by 40-60 percent typically.

3. The bets. Which Demand-Side Platform combination delivered (DV360, The Trade Desk, Amazon DSP, Yahoo), which Connected TV deal earned its place, what Made-for-Advertising exposure looked like, and how Supply Path Optimisation changed take-rates.

For brand-led programmatic (Connected TV reach, programmatic Out-of-Home, premium display): payback math runs through Marketing Mix Modelling, not last-click. Brand programmatic earns its place when modelling shows incremental revenue contribution net of fully-loaded cost (media + DSP fees + verification + measurement). Positive return on investment within 18-24 months on a brand-led Connected TV programme is healthy; faster ROI is suspicious (probably last-touch over-attributing direct-response that programmatic only assisted).

For performance-led programmatic (Dynamic Product Ads, contextual display, audience-targeted): payback runs in months, not years. Under 12 months on B2C, under 18 months on B2B. Fully-loaded customer-acquisition cost includes DSP take rates plus supply-side platform markups (which is why compressing supply paths to 5-8 named lanes materially improves payback).

Eight things 90 days before launch.

1. Demand-Side Platform setup health audit (DV360, Trade Desk, Amazon DSP, Yahoo) with brand-safety lists and exclusion lists imported.
2. Identity-graph integration (UID2, RampID, ID5, LiveRamp ATS) wired before launch. These identifiers were always needed for Safari, Firefox, in-app, and Connected TV regardless of the Privacy Sandbox reversal.
3. Made-for-Advertising filter stack (DoubleVerify + Integral Ad Science + Jounce Media) deployed pre-launch.
4. Supply Path Optimisation compressed to 5-8 named supply-side paths (down from 30+ in typical legacy setups).
5. Connected TV Private Marketplace deals secured (Netflix Ads, Disney+, Amazon Prime Video, Roku, Samsung Ads where available in your target market).
6. Brand Lift Study designed with country tier and question count locked.
7. Cross-screen measurement integration (Comscore + IRIS_ID + Automatic Content Recognition household graphs).
8. For Singapore financial-services clients: MAS FAA-N03 burnt-in disclosure pipeline for any Connected TV creative; creative-review lead-time extends by 3-7 days after FAA-N03.