We significantly simplify data analytics. You can unlock the potential hidden in your data with a cloud-native, serverless approach that decouples storage from compute and lets you analyze gigabytes to petabytes of data in minutes.
This allows you to remove the traditional constraints of scale, performance, and cost to ask any question of data and solve business problems. As a result, it becomes easier to operationalize insights across the enterprise with a single, trusted data fabric.
To achieve the goals of customer centricity, customer growth, and long-term focus on value, the company mentioned in this Private eCommerce case opted to leverage the functionality of Google Marketing Platform together with Google Cloud Platform to test and implement strategies that would allow for improved targeting, ease of activation, and insights on marketing performance. Resulting in better customer experiences and lifting Return on Investment significantly. The company’s new Marketing Cloud in Google Cloud, also called Customer Data Platform, has enabled them to seamlessly develop intelligent tooling for optimizing and deploying audiences in campaigns, with the ability to actively monitor and measure within the same environment.
In this same environment, you can, for example:
- Aggregate revenue per day in real-time to see how your business is doing
- Understand through which affiliates those new customers landed on your website
- Discover what articles will be low on stock soon
- What effect will showing particular content/articles to a specific customer have on the expected revenue
- Identify, define, visualize, and continuous update customer journeys
- Identify high-value customers before the acquisition, i.e., the ability to model look-alikes
- Apply multi-touch attribution measurement
Looking at the Private eCommerce Case, the ML model automation include:
- Re-training models weekly on the latest session data
- Comparing the model quality vs. previous models
- Deploying if acceptable and sending an alert otherwise
- Monitoring model scoring performance and statistical controls
- Monitoring scoring distributions (high/medium/low or 5% increments)
- Compare predictions with actuals (after 15 days)
Supporting models types
We support many model types, including deep neural networks, time series analysis, K-means clustering, and linear and logistic regression. Read here how it works and here how we used Big Query ML for predicting conversion intent.
For custom AI, we use Vertex AI components and Notebooks.
We also assist you in developing AI out of the box, such as Recommendations AI, Vision AI, and propensity to buy modeling of Tinyclues.
For analytical use cases, we are skilled in working with:
- Google Cloud Dataprep
- Google’s Vertex AI, Jupyter Notebook
- Crystalloids proprietary InsightOS with a graphical interface for analyzing, selecting, and visualizing
- Looker data platform
- Many other non-Google products for visualizing, reporting such as Tableau, PowerBI
Google Cloud BigLake allows organizations to unify data lakes and warehouses on a single data set. This simplifies data governance and lifecycle management and reduces duplicate copies of data while simplifying and accelerating ingest and data transformation.
BigQuery Omni allows BigQuery to access data in other public clouds and BigSearch provides querying capability for JSON-based log analytics.
With the ease of configuration, these processes run effectively with continuous improvement, allowing you, and in the Private eCommerce Case, the ability to test and learn quickly, scale to new markets efficiently, and deploy your strategies for customer growth and profitability effectively.