Client Cases

Challenge

Static banner performance could not be predicted before launch; most of the many creatives produced fell short of expectations.

Maneuva's Solution

Collected thousands of the client's and competitors' banners via Meta Ads Library API and decomposed them with Vision AI. Linked each element to past delivery results and built a machine-learning CPA model that reports expected CPA and improvement points before production.

Challenge

For products where sales and ad performance swing widely with weather (temperature, rainfall), daily bid tuning relied on the operator's gut feel.

Maneuva's Solution

Joined weather data with Meta / Google ad delivery data and visualized it. Auto-generated reports on weather-vs-CPA/ROAS correlation now inform next-day bid adjustments and pauses on a data-driven basis.

Challenge

Hundreds of Meta ad campaigns / ad sets / ads were uploaded by hand each week. Naming conventions and UTMs varied between operators, mistakes occurred, and one upload cycle took several days.

Maneuva's Solution

Built an in-house automated upload system on the Meta Marketing API. It reads the delivery plan from a spreadsheet, auto-creates campaigns/ad sets/ads, and applies naming rules and UTMs automatically — reducing upload errors to near zero.

Challenge

Prize campaigns were outsourced each time (millions to ten million yen, more than a month), and lead time and cost from planning to launch were pain points.

Maneuva's Solution

Built a composable CDP and a RAG over past campaign data on BigQuery. A multi-agent system now runs planning through page generation via conversation, completing the whole flow in tens of minutes.

Challenge

Video-ad creative iteration depended on the operator's gut feel; there was no way to project results before launch.

Maneuva's Solution

AI analyzes video elements from past ad performance. On upload of a new video, the AI agent returns predicted CPA and improvement suggestions.

Challenge

Existing social-listening tools charged high fees for post retrieval, making ongoing brand analysis expensive.

Maneuva's Solution

Built an in-house analytics app that pulls brand-related posts via the new X (Twitter) API and lets AI turn them into reports. Monthly tooling cost was cut while analysis stayed fully in-house.

Challenge

The team wanted an outbound chat system to lift LP CVR, but a SaaS solution would cost hundreds of thousands of yen per month plus meaningful integration work with the in-house CRM.

Maneuva's Solution

Built a chat system tailored to the site in one day. New chats trigger a Slack notification and can be handled straight from Slack. Inquiries flow into the in-house CRM and are auto-added to the customer list; responses can also be handled from a dedicated admin panel.