Calculating ROI on AI projects is often critical to secure resources; however, these calculations can also be notoriously challenging due to the complexity of quantification, number of teams and people involved, and often long delay in outcome.Learn More
If there’s one sector that has huge potential when it comes to bringing automation and efficiency through AI-enabled services, it’s advertising. Advertisers have troves of data from online interactions — so much so that the only real way to leverage it today is through machine learning. But that doesn’t mean that building AI systems that leverage that data to bring tangible business value is easy.
This story examines how Solocal, a digital solutions provider, has been able to harness data scientists and the power of machine learning to start building AI systems at scale. In particular, it will take a look at how they built their pricer, which builds customized quotes for potential customers of one of their advertising products, as well as their budgetizer, which runs dynamic bidding on Google and Bing ads.
Solocal is a digital services company whose product — Booster Contact — offers customers higher visibility on search engines through optimized campaigns and bids on Bing and Google AdWords. Solocal offers these services for a fixed monthly fee, which is generated on a
prospect-by-prospect basis depending on their needs.
Tools and Tech Stack
The data team at Solocal recently focused on the deployment and operationalization of two models:
- A pricer that provides automated yet customized, per-customer quotes for BoosterContact products. The machine-learning based model takes into account both geo-demographics as well as characteristics of the potential client’s business.
- A budgetizer that generates dynamic bidding on Google and Bing ads and is automatically tuned each day to keep up with recent performance and campaign behavior. The system integrates historical business data, keeps all data in sync with ad platforms, and is built on Python-based statistical modeling.
Dataiku accelerates the operationalization of models as well as makes data pipelines robust, flexible, and accessible for all profiles — data scientists, analysts, or business.”
Teamwork and Solution
Solocal used Dataiku as the cornerstone of their operationalization strategy to both put — and maintain — the pricer and budgetizer in production. Here is an overview of each AI-enabled system and its components represented in the Dataiku flow:
Following the implementation of Dataiku, Solocal saw three clear areas of improvement in their data science processes:
- Data scientists are more independent: The team moved from an in-house app developed by Python developers to Dataiku on Google Cloud Platform, which removed the barrier between research/development of algorithms and productization. This ultimately gives data scientists more independence and more control over the data pipeline.
- Time to production decreased: Dataiku allowed Solocal’s data team to reduce time-to production by at least 3x. This change was driven by the nature of Dataiku as a robust solution, including allowing for altorighmic R&D to be facilitated by the
- Smoother overall processes: In going from a mass of different tools and attempting to cobble them together in-house (data connections, Python recipes, Jupyter Notebooks, libraries integration for development, wiki, scheduled scenarios, monitoring) to Dataiku as an all-in-one solution, the overall processes and efficiency of the team are improved.