Improving Programmatic Video Ad Performance
Context & Problem:
We're running programmatic video ads for my SaaS, aiming squarely for user acquisition. The goal is to drive sign-ups and trials, not just impressions or views. However, as mentioned, our overall programmatic video ad performance is underwhelming: we're struggling with high CPMs, low VCRs, inconsistent viewability across different placements, and a high cost per conversion which is really hurting our ROI.
What I've Tried:
I've tried quite a few things already, hoping to crack the code on better programmatic video optimization:
- Adjusting bid strategies: Experimented with both fixed and dynamic bidding, trying to find the sweet spot between reach and efficiency.
- A/B testing different video creatives: We've iterated on numerous video creatives, testing various hooks, calls-to-action, and lengths.
- Refining audience targeting: Deep dived into first-party data, lookalikes, and interest-based segments.
- Setting frequency caps: Implemented caps to avoid ad fatigue and ensure we're not overexposing users.
- Optimizing supply sources: Actively blacklisting low-performing domains and whitelisting those with better viewability and VCRs.
- Monitoring and optimizing for viewability scores: Constantly checking viewability metrics and trying to shift spend towards higher-quality inventory.
Specific Challenges:
Despite these efforts, I'm still facing some persistent challenges:
- Differentiating between low-quality inventory and actual ad fraud: It's tough to tell if poor performance is due to genuinely bad placements or sophisticated bot traffic.
- Balancing reach and scale with viewability and completion rates: Sometimes, when I optimize too heavily for viewability, our reach shrinks significantly, making it hard to scale.
- Optimizing for down-funnel KPIs (e.g., sign-ups) beyond just video engagement metrics: It feels like many DSPs are built for impression/view-based optimization, making it hard to tie video performance directly to actual sign-ups efficiently.
Main Question:
So, I'm really seeking advanced strategies, specific DSP features, or third-party tools that you've found effective for deeper programmatic video optimization to significantly boost results. I'm looking beyond the basic optimizations and into more sophisticated approaches.
Specifically, any practical tips on more robust fraud detection methods within programmatic video, and how to allocate budget more efficiently when the primary goal is down-funnel conversions rather than just video views?
Closing:
Anyone faced similar challenges and found effective solutions? Would love to hear your experiences and recommendations!
2 Answers
MD Alamgir Hossain Nahid
Answered 1 week ago- Robust Fraud Detection & Prevention:
- Pre-Bid & Post-Bid Verification: Beyond basic blacklisting, integrate with advanced third-party verification services like Integral Ad Science (IAS), DoubleVerify, or Moat. These platforms offer sophisticated algorithms and real-time filtering to detect bot traffic, domain spoofing, and other forms of ad fraud *before* your bid is even placed. Many DSPs have native integrations, but ensure you're leveraging their full capabilities.
- Custom Fraud Thresholds: Work with your DSP account manager to set custom, stricter fraud thresholds. While this might slightly reduce reach, it drastically improves the quality of impressions and completion rates, ultimately leading to better conversion efficiency.
- Contextual Targeting & Brand Safety: While not directly fraud detection, focusing heavily on contextual targeting (placing ads on pages highly relevant to your SaaS product) can naturally steer you away from low-quality, generic inventory often plagued by fraud. Combine this with strong brand safety measures.
- Balancing Reach, Viewability, & Completion Rates:
- Private Marketplaces (PMPs) & Deal IDs: This is crucial. Instead of relying solely on open exchange inventory, establish PMPs with premium publishers known for high viewability and engaged audiences relevant to your SaaS. These deals give you preferential access to quality inventory at negotiated rates, often with built-in viewability guarantees. It sacrifices some scale for quality, but the ROI often makes up for it.
- Advanced Viewability Metrics & Optimization: Go beyond basic viewability. Look at metrics like "audible and viewable on completion" (AVOC) or "time in view." Some advanced DSPs allow optimization towards these deeper viewability metrics.
- Creative Optimization for Programmatic Video: Your A/B testing is good, but consider optimizing creatives specifically for programmatic environments where users might skip or view silently. Think about:
- Quick Hooks: Capture attention in the first 3-5 seconds.
- Visual Storytelling: Convey your message effectively without sound.
- Clear, Early CTAs: Place calls-to-action strategically throughout the video, not just at the end.
- Optimizing for Down-Funnel KPIs (Sign-ups, Trials):
- Server-to-Server (S2S) Postbacks: This is a game-changer for accurate conversion tracking and optimization, especially in a privacy-centric world. Instead of relying solely on client-side pixels (which can be blocked or inaccurate), implement S2S postbacks. When a user converts on your site, your server sends a direct signal to the DSP's server. This provides a much more robust and reliable data stream for the DSP's algorithms to optimize against. Itโs foundational for effective conversion attribution modeling.
- CRM & DMP Integration: Integrate your customer relationship management (CRM) system and any Data Management Platform (DMP) with your DSP. This allows you to feed first-party data (e.g., existing customer segments, trial users, high-value leads) into your programmatic campaigns for more precise lookalike modeling and exclusion targeting. You can also push conversion data back to your CRM to understand the true value of users acquired through video.
- Value-Based Bidding & Custom Algorithms: If your DSP supports it (e.g., Google's Value-Based Bidding, The Trade Desk's Koa), leverage bidding strategies that optimize for the *value* of a conversion rather than just the conversion event itself. By assigning different values to sign-ups versus trials versus paid subscriptions, the DSP can bid more intelligently. For highly sophisticated teams, some DSPs allow for custom bidding algorithms or scripts that can be tailored to your specific business logic and KPIs.
- Attribution Modeling Beyond Last-Click: Programmatic video often plays a role higher up in the funnel. Implement multi-touch attribution models (e.g., linear, time decay, position-based) to give appropriate credit to video touchpoints, rather than just relying on last-click, which often undervalues video's contribution. This helps you understand video's true impact on your overall ad tech stack performance.
Salma Hassan
Answered 1 week agoWow MD Alamgir Hossain Nahid, didn't expect such a thorough reply, this is super helpful!