实例探究 > Pixability Enhances Ad Performance with Snorkel Flow's NLP Capabilities

Pixability Enhances Ad Performance with Snorkel Flow's NLP Capabilities

公司规模
200-1,000
地区
  • America
国家
  • United States
产品
  • Snorkel Flow
技术栈
  • NLP
  • Foundation Models
  • Classification Models
实施规模
  • Enterprise-wide Deployment
影响指标
  • Customer Satisfaction
  • Innovation Output
  • Productivity Improvements
技术
  • 分析与建模 - 机器学习
  • 分析与建模 - 自然语言处理 (NLP)
  • 分析与建模 - 预测分析
适用行业
  • Software
适用功能
  • 商业运营
  • 销售与市场营销
服务
  • 软件设计与工程服务
  • 系统集成
关于客户
Pixability is a leading YouTube & Connected TV Ad Platform that focuses on helping customers maximize their reach and optimize their video ad spend. The company is known for its advanced advertising solutions that leverage data and technology to deliver targeted and effective ad campaigns. With a strong presence in the digital advertising industry, Pixability serves a diverse range of clients, including major brands and agencies, by providing tools and insights to enhance their advertising strategies. The company's commitment to innovation and customer satisfaction has positioned it as a key player in the ad tech space.
挑战
The time-consuming process of manually labeling high-cardinality training data blocked Pixability from expanding their NLP capabilities.
解决方案
With Snorkel Flow, Pixability distilled knowledge from foundation models to build smaller, deployable classification models with more than 90% accuracy in just days. This approach significantly improved ad performance and brand-suitable targeting. By leveraging Snorkel Flow, Pixability was able to automate the labeling process, which previously required manual effort and was a major bottleneck. The solution involved creating a 600+ class multi-label NLP model that provided greater granularity and support for custom content categories. This allowed Pixability to offer more precise targeting options to their clients, enhancing the overall effectiveness of their ad campaigns.
运营影响
  • 500K programmatic labels sourced from FM responses and keyword analysis with zero ground truth.
  • 600+ class multi-label NLP model that provides greater granularity and support for custom content categories.
  • Achieved 90% accuracy on a model with 26x more classes, demonstrating the scalability and effectiveness of the solution.
  • Improved ad performance and brand-suitable targeting, leading to better customer satisfaction and campaign outcomes.
数量效益
  • 500k programmatic labels
  • 600+ class multi-label NLP model
  • 90% accuracy

Case Study missing?

Start adding your own!

Register with your work email and create a new case study profile for your business.

Add New Record

相关案例.

联系我们

欢迎与我们交流!
* Required
* Required
* Required
* Invalid email address
提交此表单,即表示您同意 IoT ONE 可以与您联系并分享洞察和营销信息。
不,谢谢,我不想收到来自 IoT ONE 的任何营销电子邮件。
提交

感谢您的信息!
我们会很快与你取得联系。