Optimize ROAS like industry leaders with AI-powered bidding
The digital advertising landscape is undergoing its most significant transformation since the advent of programmatic buying. With Chrome’s evolving approach to third-party cookies and the meteoric rise of AI-powered bidding, marketers face both unprecedented challenges and opportunities. Recent case studies from functional cosmetics brand Miamo and African travel platform Travelstart (+27% revenue lift) demonstrate how combining-focused bidding with AI optimization creates resilient performance in this new era.
These successes aren’t accidental—they reflect fundamental shifts in advertising philosophy. Where Travelstart once relied on volume-based TCPA bidding (target cost per acquisition), their 2025 strategy embraced value-based ROAS algorithms that prioritize booking profitability over mere conversion counts. Similarly, Miamo’s protocol-based skincare business achieved 94% search impression share by letting AI dynamically adjust bids between minimum profitability thresholds and peak-season volume targets. As privacy regulations reshape measurement capabilities, these advertisers prove that first-party data integration and machine learning aren’t just advantageous—they’re existential necessities for modern performance marketing.
Ⅰ. Transitioning from Volume-Based to Value-Based Bidding
For years, Travelstart’s marketing team measured success through a singular lens: booking volume. Their TCPA bidding strategy optimized for cost-efficient acquisitions but ignored critical revenue variables—a limitation that became glaring during Africa’s 2024 travel resurgence. When analyzing campaigns for domestic routes like Johannesburg-to-Cape-Town, they discovered that 22% of bookings generated 63% of total revenue, yet their bidding treated all conversions equally. This volume-centric approach forced unsustainable bid increases to maintain share in South Africa’s carrier-constrained market, eroding profitability.
Miamo’s strategy presented a stark contrast. By implementing ROAS target ranges—where the floor ensured baseline profitability and the ceiling allowed for demand surges—their AI could bid aggressively during Black Friday while maintaining margins during slower periods. Google’s Smart Bidding algorithms analyzed each search query’s likelihood to deliver not just a sale, but a high-value sale. The results were transformative: a 70% revenue surge versus prior campaigns, with conversion rates jumping 40% year-over-year. Travelstart’s parallel shift to value-based bidding through Search Ads 360 yielded a 27% revenue boost and 50% higher revenue per booking, proving that even in transaction-heavy industries, not all conversions hold equal worth.
Ⅱ. AI-Powered Campaign Management Techniques
At the core of these successes lies AI’s ability to process multivariate signals in real time. Travelstart’s adoption of dynamic templates allowed automatic campaign adjustments based on fluctuating variables—flight availability, route popularity, and seasonal search trends. When queries for “Madagascar holiday packages” spiked 70% among South African users, their system instantly generated route-specific Google Ads without manual intervention. This responsiveness was particularly crucial during peak booking windows, where hour-by-hour price adjustments (like JNB-CPT fares dropping from ZAR909 to ZAR599) captured demand surges while protecting margins.
Miamo leveraged similar automation but applied it to product-level profitability. Their AI continuously recalibrated bids for items like the Global Eye Defense Sunscreen Concealer based on inventory levels, margin structures, and cross-sell potential within their “protocol” skincare system. During their YouTube Masthead campaign, machine learning allocated spend toward audiences most likely to complete high-value multi-product purchases, achieving a 6M user reach while maintaining it targets. Both cases underscore AI’s role as a force multiplier—transforming static campaigns into adaptive systems that respond to micro-moments.
Ⅲ. Privacy-Centric Advertising in the AI Era
The deprecation of third-party cookies has forced advertisers to reinvent measurement frameworks, driving demand for privacy-compliant solutions. These tools enable granular conversion tracking by associating multiple advertising accounts with precise event tagging, capturing actual transaction values rather than binary conversions. By processing first-party datasets through privacy-preserving techniques such as differential privacy, advertisers can maintain attribution accuracy despite signal loss.
Miamo exemplifies this shift through its ROAS-driven budget protocol, which leverages consented first-party data to refine AI bidding strategies without compromising anonymity. Their 102% Q1 2025 conversion lift underscores how privacy-centric frameworks—can optimize performance while building trust. The integration of such solutions with Google Ads’s Privacy Sandbox APIs allows brands to balance precision and compliance, particularly in regulated industries like pharmaceuticals, where offline sales data must align with online engagement metrics.
Ⅳ. Industry-Specific Application Frameworks
The divergence between Miamo’s e-commerce playbook and Travelstart’s travel industry adaptations reveals how value-based bidding requires tailored execution. For Miamo’s direct-to-consumer model, ROAS targets were calculated against lifetime value (LTV) metrics, allowing temporary profitability sacrifices during customer acquisition bursts. Their UK/Switzerland expansion used these LTV models to bid aggressively on high-intent searches like “clinical skincare regimen,” driving 51% annual revenue growth.
Topkee’s Google Ads optimization framework prioritizes ROI-driven campaign performance through data-centric methodologies. Their comprehensive approach begins with website assessments using advanced scoring tools to identify SEO gaps, ensuring alignment between ad targeting and landing page relevance. The TTO CDP platform enables centralized management of conversion tracking across multiple ad accounts, while TM settings provide granular campaign analytics through customizable URL parameters. For heritage travel campaigns targeting culturally curious audiences, Topkee’s keyword research combines competitor analysis with AI-powered creative production, dynamically adjusting bids based on attribution models that show remarketing ads achieve 70% higher purchase intent. This systematic integration of conversion tracking, smart bidding, and cross-channel attribution mirrors the original case study’s focus on value drivers, substituting route profitability with digital marketing KPIs as the optimization compass.
Ⅴ. Strategic Budget Allocation and Flexibility
Miamo’s transition from rigid daily budgets to “open” -driven spending unlocked hidden opportunities. Previously, their ads disappeared for high-intent “Miamo” searches once budgets depleted—akin to a pharmacy closing during peak hours. By instead allowing AI to flex spending based on real-time profitability signals, they achieved 94% impression share while maintaining margins. This flexibility proved vital during promotional periods; lowering ROAS targets briefly for Black Friday generated revenue spikes without sacrificing annual profitability goals.
Travelstart adopted similar elasticity but with geographical nuance. Campaigns for high-demand domestic routes operated at lower thresholds than international flights, reflecting differing competitive pressures. Their AI automatically reallocated budgets toward emerging opportunities like Zimbabwean “shoulder season” searches, which delivered 27% higher revenue per impression than fixed-budget approaches could capture.
Ⅵ. Future-Proofing Google Ads Strategies
As Miamo tests Demand Gen campaigns for mid-funnel engagement, they’re pioneering a hybrid model where upper-funnel awareness efforts are directly tied to bottom-funnel metrics. Early experiments with Switzerland’s German-speaking market show that YouTube skippable ads, when optimized for eventual purchase value rather than just views, reduce CPAs by 19%. Travelstart’s parallel work with Privacy Sandbox’s Protected Audience API suggests that even cookie-less remarketing can drive value—their heritage travel retargeting pools now achieve 80% match rates using first-party data alone.
Both brands emphasize continuous testing frameworks. Miamo’s weekly AI “audits” identify underperforming product campaigns for rapid it adjustments, while Travelstart’s route templates undergo monthly profitability recalibrations. This iterative approach—combining AI’s real-time optimizations with human strategic oversight—creates a flywheel where each campaign’s performance data enhances future decisions.
Conclusion
The Miamo and Travelstart case studies illuminate advertising’s new paradigm: AI + Privacy as interdependent pillars. As Travelstart’s CFO notes, their value-based approach made Search Ads 360 the “most profitable marketing channel” despite Africa’s airline constraints. Miamo’s 102% conversion leap proves that even niche skincare brands can leverage these tools for global scalability.
Emerging trends—from Privacy Sandbox’s audience solutions to Demand Gen’s mid-funnel potential—will further blur lines between brand and performance. Google Ads who delay adopting these frameworks risk not just inefficiency, but strategic obsolescence. For teams ready to begin, the time is now: Start small with a pilot campaign, instrument first-party data rigorously, and let AI reveal the hidden profit pathways in your account.
