Time to go outside again, I guess. I’ll need a sun hat. Sunscreen. Maybe some new sandals? What else?

With the Recommendations AI service, I might be reminded to grab a reusable water bottle and a swimsuit. Or some after-sun aloe lotion. Good thing, cause I’ll need it.

Recommendations AI is a solution that uses machine learning to bring product recommendations to their shoppers across any catalog or client list. This service is part of our full suite of Retail solutions. When you integrate with the Retail API, you get the benefit of Google’s Product Discovery. Integrating once to reap the benefits over and over. Recommendations is the starting point, and you can easily extend into Retail Search and Vision Product Search in the upcoming future. 

The Recommendations solution is fully managed, global-scale and powered by deep learning, so you can focus on a great shopping experience and let someone else worry about the infrastructure.

Compared to baseline recommendation systems used by customers, Recommendations AI showed double digit uplift in conversion and clickthrough rates in A/B experiments controlled by the customers. You can optimize for click-through, conversion or session revenue, and fine tune the models to make sure you omit out-of-stock items or duplicates, for example.

So how does it work, and how do you get started? Read on, and we’ll walk you through the pipeline, starting with the data you already have to placement in your online store. 

Formula: Data -> Model -> Placement

You start with your catalog, the list of all the things (postcards, movies, pie recipes) that you want to show your customers. Then you ingest your PII-redacted user events -this is the historic event data like home page views, add to cart events and more along with real time user events. This user event is joined with the product catalog and items that allows us to construct the sequence of shoppers’ activity, thus being able to predict what the shopper has a high propensity to purchase next. The user events can come from both online activity across devices or offline store purchases

The recommendation model will return a list of products, which are the recommendations. The brains of the operation, if you will. This model is trained using all the data that you ingest, using the latest neural network models and techniques that Google has built expertise over the years in flagship products like Youtube and News, that allows us to uncover shopper intent,  so it can best predict the right recommendations to show to the right people.

Every model outputs a list of product identifiers, but where do they go? They go into placements, the spots, panels, carousels on your customer’s journey interacting with your brand that you’ve set aside to highlight recommendations. A model can send recommendations to one or more placements, but each placement only receives information from one recommendation model. Your pages will then need to render the products with the right images, text or other metadata, using the product ID that is returned by the model.

What do recommendations look like?

Let’s start by browsing our postcard-selling website, where I’ve been buying some vintage California postcards already. The recommendations algorithm has caught on to my interest, showing me other potential cards to purchase based on my history:

Put your data to work

To get started we need to bring your data into the recommendation model, so it can understand your customers, your inventory, and your sales patterns. 

The model takes in the product catalog you use, and metadata about those products to better understand nuances in assortment, pricing and variables like size and style. You might already have this data stored in BigQuery or Merchant Center, and hence we provide easy integrations that you can leverage to get started even faster.

As for the user events, don’t worry if you already have systems in place to capture web and mobile activity. We make it easy to bring in your real time event logs by providing seamless integrations with Google Tag Manager, Javascript pixel, or even historic events from Cloud Storage, BigQuery or using inline API or JSON, so you can immediately train the models on this imported  data. All this allows you to kickstart integrating with Recommendations AI in a matter of days.

The models then construct a sequence of activities that the user went through and joins with the products that the user engaged with. Once your data is ready to go, it takes a few days to train the model. Next onto making the data work for you.

Quickly customize your model

Setting up your own recommendations project in the console gives you the ability to choose what sort of model to train (based on what recommendations you want to generate) and your objective. Are you optimizing for click-through rate–more people click on the recommendation links or products–or for conversion rate–more people choose or buy what was suggested or revenue ?

Different models can be optimized for different optimization goals.; the GCP console explains what each one can do and how you can choose to optimize it.

Let’s unpack some of this terminology real quick.

We’ve got three model types:

• Recommended for you – Means we think these are items you’ll want to buy, based on your history; this is usually used on a home page to showcase items.
• Others you may like – Means if you’re browsing the page of a water bottle, we will recommend  alternative brands of water bottles that you may like as well as a backpack, based on your engagement  history.
• Frequently bought together – Means that when anyone buys sunscreen, we notice that they often also buy aloe lotion, so we will surface those items when someone adds any one of them to their cart.

And then we have three business objectives that the models optimize for:

• Click-through rate – How frequently did somebody click on a recommended item?
• Conversion rate– How frequently did somebody add a recommended item to their cart?
• Revenue per session – How much money did the recommendations generate for you?

Deliver anywhere along the journey

Now that you’re all set up in the Retail AI console, you can test out the recommendations in the console, even before you deploy to production.

You can integrate Recommendations into your frontend by calling the Predict APIt. The placements of recommendations will report data back into the dashboard and you can analyze and measure success for future iterations. 

On top of that you can use the recommendations for other parts of your customer’s journey. Email promotions, storefront kiosks, display ads or follow-up notifications can include recommendations based on past activity and cart contents. The model gives you useful product recommendations for a wide variety of touchpoints and steps in the purchasing process.

More best practices, and guides, are available inside our documentation.

How to get started

Training your own models can be tedious, time-consuming, and expensive. On top of that it requires deeper data science expertise to set up. Let us do it instead!

You can see how IKEA Retail uses Recommendations AI in this recent talk and blog from the Google Cloud Retail Summit..

To get started today you’ll need to make a Cloud project and enable the Retail API, which then allows you to access all the recommendation tools in one menu. Bring in your catalog and purchasing data, define a placement or two, and you can start putting recommendations on your site in a matter of days.

Digitec Galaxus AG is the biggest online retailer in Switzerland, operating two online stores: Digitec, Switzerland’s online market leader for consumer electronics and media products, and Galaxus, the largest Swiss online shop with a steadily growing range of consistently low-priced products for almost all daily needs. 

Known for its efficient, personalized shopping experiences, it’s clear that Digitec Galaxus understands what it takes to deliver a platform that is interesting and relevant to customers every time they shop. 

The problem: Personalizing decisions for every situation

Digitec Galaxus already had established an engine to help them personalize experiences for shoppers when they reached out to Google Cloud. They had multiple recommendation systems in place and were also extensive early adopters of Recommendations AI, which already enabled them to offer personalized content in places like their homepages, product detail pages, and their newsletter. 

But those same systems sometimes made it difficult to understand how best to combine and optimize to create the most personalized experiences for their shoppers. Their requirements were threefold:

1. Personalization: They have over 12 recommenders they can display on the app, however they would like to contextualize this and choose different recommenders (which in turn select the items) for different users. Furthermore they would like to exploit existing trends as well as experiment with new ones.
2. Latency: They would like to ensure that the solution is architected so that the ranked list of recommenders can be retrieved with sub 50 ms latency.
3. End-to-end easy to maintain & generalizable/modular architecture: Digitec wanted the solution to be architected using an easy to maintain, open source stack, complete with all MLops capabilities required to train and use contextual bandits models. It was also important to them that it is built in a modular fashion such that it can be adapted easily to other use cases which have in mind such as recommendations on the homepage, Smartags and more . 

To improve, they asked us to help them implement a machine learning (ML) contextual bandit based recommender system on Google Cloud taking all the above factors into consideration to take their personalization to the next level. 

Contextual bandits algorithms are a simplified form of reinforcement learning and help aid real-world decision making by factoring in additional information about the visitor (context) to help learn what is most engaging for each individual. They also excel at exploiting trends which work well, as well as exploring new untested trends which can yield potentially even better results. For instance, imagine that you are personalizing a homepage image where you could show a comfy living room couch or pet supplies. 

Without a contextual bandit algorithm, one of these images would be shown to someone at random without considering information you may have observed about them during previous visits. Contextual bandits enable businesses to consider outside context, such as previously visited pages or other purchases, and then observe the final outcome (a click on the image) to help determine what works best. 

Creating a personalization system with contextual bandits

While Digitec Galaxus heavily personalizes their website homepages, they are very very sensitive and also require more cross-team collaboration to update and make changes. 

Together with the Digitec Galaxus team, we decided to narrow the scope and focus on building a contextual bandit personalization system for the newsletter first. The digitec Galaxus team has complete control over newsletter decisions and testing various ML experiments on a newsletter would have less chance of adverse revenue impact than a website homepage. 

The main goal was to architect a system that could be easily ported over to the homepage and other services offered by Digitec with minimal adaptations. It would also need to satisfy the functional and non-functional requirements of the homepage as well as other internal use cases.

Below is a diagram of how the newsletter’s personalization recommendation system works:

• The system is given some context features about the newsletter subscriber such as their purchase history and demographics. Features are sometimes referred to as variables or attributes, and can vary widely depending on what data is being analyzed. 
• The contextual bandit model trains recommendations using those context features and 12 available recommenders (potential actions). 
• The model then calculates which action is most likely to enhance the chance of reward (a user clicking in the newsletter) and also minimize the problem (an unsubscribe). 

Calculating whether a click was a newsletter or an unsubscribe enabled the system to optimize for increasing clicks and avoid showing non-relevant content to the user (click-bait). This enabled Digitec Galaxus to exploit popular trends while also exploring potentially better-performing trends. 

How Google Cloud helps

The newsletter context-driven personalization system was built on Google Cloud architecture using the ML recommendation training and prediction solutions available within our ecosystem. 

Below is a diagram of the high-level architecture used:

The architecture covers three phases of generating context-driven ML predictions, including: 

ML Development: Designing and building the ML models and pipeline 
Vertex Notebooks are used as data science environments for experimentation and prototyping. Notebooks are also used to implement model training, scoring components, and pipelines. The source code is version controlled in Github. A continuous integration (CI) pipeline is set up to automatically run unit tests, build pipeline components, and store the container images to Cloud Container Registry. 

ML Training: Large-scale training and storing of ML models 
The training pipeline is executed on Vertex Pipelines. In essence, the pipeline trains the model using new training data extracted from BigQuery and produces a trained, validated contextual bandit model stored in the model registry. In our system, the model registry is a curated Cloud Storage. 

The training pipeline uses Dataflow for large scale data extraction, validation, processing, and model evaluation, and Vertex Training for large-scale distributed training of the model. AI Platform Pipelines also stores artifacts, the output of training models, produced by the various pipeline steps to Cloud Storage. Information about these artifacts are then stored in an ML metadata database in Cloud SQL. To learn more about how to build a Continuous Training Pipeline, read the documentation guide.

ML Serving: Deploying new algorithms and experiments in production 
The training pipeline uses batch prediction to generate many predictions at once using AI Platform Pipelines, allowing Digitec Galaxus to score large data sets. Once the predictions are produced, they are stored in Cloud Datastore for consumption. The pipeline uses the most recent contextual bandit model in the model registry to evaluate the inference dataset in BigQuery and give a ranked list of the best newsletters for each user, and persist it in Datastore. A Cloud Function is provided as a REST/HTTP endpoint to retrieve the precomputed predictions from Datastore.

All components of the code and architecture are modular and easy to use, which means they can be adapted and tweaked to several other use cases within the company as well.

Better newsletter predictions for millions

The newsletter prediction system was first deployed in production in February, and Digitec Galaxus has been using it to personalize over 2 million newsletters a week for subscribers. The results have been impressive, 50% higher than our baseline. However, the collaboration is still ongoing to improve the results even more. 

“Working at this level in direct exchange with Google’s machine learning experts is a unique opportunity for us. The use of contextual bandits in the targeting of our recommendations enables us to pursue completely new approaches in personalization by also personalizing the delivery of the respective recommender to the user. We have already achieved good results in our newsletter in initial experiments and are now working on extending the approach to the entire newsletter by including more contextual data about the bandits arms. Furthermore, as a next step, we intend to apply the system to our online store as well, in order to provide our users with an even more personalized experience. To build this scalable solution, we are using Google’s open source tools such as TFX and TF Agents, as well as Google Cloud Services such as Compute Engine, Cloud Machine Learning Engine, Kubernetes Engine and Cloud Dataflow.”—Christian Sager, Product Owner, Personalization ( Digitec Galaxus)

Since the existing architecture and system is also dynamic, it will automatically adapt to new behaviours, trends, and users. As a result, Digitec Galaxus plans to re-use the same components and extend the existing system to help them improve the personalization of their homepage and other current use cases they have within the company. Beyond clicks and user engagement, the system’s flexibility also allows for future optimization of other criteria. It’s a very exciting time and we can’t wait to see what they build next!

Do you know how much money you could save by adjusting workload requests to better represent their actual usage? If you’re not rightsizing your workloads, you might be overpaying for resources that your workloads aren’t even using or worse, putting your workloads at risk for reliability issues due to under provisioning.

As we’ve previously discussed, setting the resources is the most important thing you can do to increase the reliability of your Kubernetes workloads. In this blog we will help you with the second key finding from the State of Kubernetes Cost Optimization report!

According to our research findings, workload rightsizing is the most important golden signal. Workload rightsizing measures the capacity of developers to properly use the CPU and memory they have requested for their applications. 

Rightsizing is challenging

It can be quite difficult to predict the resource needs of your applications, which historically has not been a concern for developers in traditional data center environments.In traditional data center environments, resources were typically over-provisioned upfront to ensure capacity for peak demand and future growth, so developers didn’t need to focus on accurately predicting resource needs as they were covered by the excess capacity, whereas in cloud environments, resources are consumed on-demand. Finding a balance between efficiency and reliability can often feel like a delicate balancing act.

Tools for workload rightsizing

There are native tools in Cloud Monitoring and the GKE UI you can use to rightsize your workloads running on GKE. 

Rightsizing in the console

The Workload Cost Optimization tab helps you identify workloads that can be optimized by displaying the resources used versus what’s requested.

To take advantage of potential cost savings, you can drill into clusters to see workload level resource recommendations.

To view workload resource recommendations for Deployment objects only:

1. In the GKE Cost Optimization.
2. Select a cluster.
3. Click Workloads > Cost Optimization.
4. Select one Deployment workloads
5. In the workload’s detail page, select Actions > Scale > Edit Resource Requests

Rightsizing with Cloud Monitoring

Cloud Monitoring provides built-in VPA scale recommendations metrics that you can use to monitor the performance of your workloads and to identify opportunities to rightsize them without the need to create VPA objects.

To view these metrics:

1. Go to the Cloud Monitoring > Metric Explore console.

2. In the Metric dropdown, select the metrics:

• Memory recommendations: 
  Kubernetes Scale > autoscaler > Recommended per replica request bytes
• CPU recommendations: 
  Kubernetes Scale > autoscaler > Recommended per replica request cores

Rightsizing at scale

If you’re interested in viewing recommendations across clusters and projects, We’ve created a guide that you can use today to help you right-size your GKE workloads at scale. This solution leverages your actual cluster’s metric data and built-in workload recommendations provided by Cloud Monitoring. You can determine the resource requirements for all your workloads without having to create additional VPA autoscaler objects in each of your clusters. The guide walks you through deploying the solution.

In conclusion

In conclusion, rightsizing your workloads is essential for both cost savings and reliability. By following the tips in this blog, you can ensure that your workloads are using the right amount of resources, which will save you money and increase your workload’s reliability.

Links to the solution presented in this blog and other useful tools to help you optimize your cluster are listed below:

• The Right-sizing workloads at scale solution guide
• Setting resource requests: the key to Kubernetes cost optimization
• The simple kube-requests-checker tool
• An interactive tutorial to get set up in GKE with a set of sample workloads

Download the State of Kubernetes Optimization report, review the key findings, and stay tuned for our next blog post.

What do CIOs and CTOs deliver for the company? If you said “technology,” that’s just the beginning. According to their research, McKinsey found that 85% of CIOs and CTOs interviewed in the spring of 2019 said they were essential for at least two of the three most common CEO priorities—revenue acceleration, improved agility and time to market, and cost reduction.

IT modernization – including migrating to the cloud – is key to business growth and agility. Yet, according to a recent McKinsey study, 80% of CIOs report that regardless of their level of cloud migration, they still haven’t reached their projected agility and business benefits. Sometimes, this is because of issues like training and skills gaps in the IT workforce. Surprisingly often though, the barrier to reaching the goals is based on trade-offs that CIOs themselves feel they must make to strike a balance between the perfect and the possible.

But what if you could have it all without the trade-offs? As Will Grannis, Managing Director of the CTO Office at Google, and Arul Elumalai, Partner at McKinsey & Company discussed in our recent digital conference, many of the compromises CIOs make can be avoided with new technology, modern architectures and by encouraging a transformation mindset across the business. In interviews, CIOs explained how they’ve leveraged the best of the cloud without compromising on security, agility, and flexibility. Here’s how these leaders avoid three of the top perceived trade-offs—both with technology and by transforming their operating model.

Trade-off #1: Developer agility vs. control and governance

Moving to the cloud offers new opportunities for speed, but 69% of organizations indicate that stringent security guidelines and code review processes can slow developers significantly. One CISO of a multinational company mentioned that cloud development was so fast that they had to institute manual checks on their developers’ code. So much for agility. 

To overcome this trade-off and maintain both speed and security, some respondents found success in DevOps, hiring security-experienced talent and introducing automation for security and quality. Building in security into the CI/CD pipeline and increasing automation don’t just eliminate the tradeoff, they result in higher quality and faster innovation.

At Google Cloud, we’ve also observed that customers with strong DevOps practices have increased speed-to-market and product/service quality. From our own journey, we’ve learned seven critical lessons essential to adopting a DevOps model, ranging from taking up small projects and embracing open source to building an overall DevOps culture.

Trade-off #2: Single-vendor benefits vs. freedom from lock-in

CIOs perceive benefits to using the fewest number of clouds, specifically avoiding introducing multiple systems that require their teams to develop and maintain multiple skillsets. Unfortunately, 83% of the CIOs interviewed said that while they would prefer fewer clouds, the potential financial and technical lock-in drives them to multiple providers. 

Successful CIOs said that they can avoid lock-in pitfalls not just with contractual guardrails and executive and board education, but with evolving hybrid cloud technologies that provide additional choices. Hybrid cloud platforms based on containers can further mitigate the risk of using a single cloud vendor. The key to successful hybrid architectures is the infrastructure abstraction and portability that containers create for them, enabling disparate environments to work together. 

This notion has been at the heart of our strategy at Google Cloud with Anthos, which provides an abstraction layer and an application modernization platform for hybrid and multi-cloud environments. Enterprises can use Anthos to modernize how they develop, secure, and operate hybrid-cloud environments and enable consistency across cloud environments.

Trade-off #3: Best-of-breed tools vs. standardization and familiarity

Optimizing tool chains for different environments can improve productivity, but many CIOs believe that this means reduced functionality and tools. While 77% of CIOs said they had to standardize to the lowest common denominator, some have found a better solution. Rather than giving up the languages, libraries, and frameworks that their teams prefer, effective leaders said that they found success by investing in training programs to upscale talent and adopting new open and vendor-agnostic solutions. Architectures that are based on open-source components have been the keys that helped remove this tradeoff, and eliminate the notion of a lowest common denominator. 

This is why we have built Anthos on open-source components like Kubernetes, Istio and Knative. Anthos gives your business the choice you need. With the ability to create code that works in most environments using the tools, languages, and systems you prefer, you can do more without major changes to how you work.

Regardless of your current cloud adoption level, check out “Unlock business acceleration in a hybrid cloud world” to discover more about McKinsey’s findings, including how CIOs drive agility, methods to make trade-offs unnecessary, and how to prepare your team for the cloud. Then, stay tuned for subsequent posts that  take a closer look at how hybrid solutions and strategies can help CIOs drive a transformation mindset across the business—without compromising on security, agility, and flexibility.

The financial services industry is changing—an estimated $68 trillion in wealth transferring from baby boomers to millennials.¹ This means financial service providers will have to deliver the speed, ease-of-use, technological sophistication, and tailored services that millennials have come to expect. In fact, half of all millennials are willing to switch to a competing institution if it offers a better digital experience.² This and many other trends are driving unprecedented growth for mobile Fintech experiences in banking, digital payments, financial management and insurance.³

Google Maps Platform financial services solutions

To help you respond to customer’s changing demands, we’re launching financial services solutions that can help you improve your customer experience, security and operations. We’ve outlined the technical guidance and APIs you need to build out three financial services solutions: Enriched Transactions, Quick and Verified Sign-up, and Branch and ATM Locator Plus. We’ve also highlighted two use cases that customers are using our APIs to solve: Contextual Experiences and Fraud Detection. 

Clarify financial statements with Enriched Transactions solution

Transaction statements are often hard for customers to understand, using abbreviations like “ACMEHCORP” instead of customer-facing names like “Acme Houseware”. Our Enriched Transactions solution clarifies these transactions and makes them instantly recognizable by adding the merchant name and business category, a photo of the storefront, its location on a map, and full contact info. Making transactions easier to recognize not only boosts consumer confidence, with reported increases in NPS of 15% or higher, but decreases costly support calls by approximately 67%.⁴

In addition, you can help customers easily visualize a series of transactions by adding the merchant name to the transaction amount and date, and displaying their transactions on a Google map. This enables you to give customers insights about where and how they spend money. See the guide to implement Enriched Transactions today.

Enable faster sign-up with Quick and Verified Sign-up solution

Manually entered addresses can lead to lowered conversions, erroneous customer data, and costly delivery mistakes. Our Quick and Verified Sign-up solution makes sign-up faster, suggesting nearby addresses with just a few thumb taps—cutting sign-up time by up to 64% and increasing conversion rates by up to 15%.⁵  

The solution also provides one additional level of address verification that helps reduce the risk of fraudulent account sign-ups—and companies have decreased fraudulent account setups by approximately 30% through using geospatial data to verify customer identities.⁴  See the Quick and Verified Sign-up solution guide to get started today.

• Faster sign-ups 1An application form requires an address
• Faster sign-ups 2Autocomplete quickly suggests addresses
• Faster sign-ups 3Select the address with visual confirmation
• Faster sign-ups 4Address verification options are presented
• Faster sign-ups 5Location permission is granted by the user
• Faster sign-ups 6The address is verified

Help customers visit you with Branch and ATM Locator Plus solution

74% of customers now search for specific details prior to their visit, which makes detailed, accurate profiles for each location a must.⁵ Our Branch and ATM Locator Plus solution enhances your own websites and apps with the same information shown about your branches and ATMs on Google Maps. Include hours of operation, available services, user reviews, photos of the location, driving directions and more.

Financial services companies using geospatial data to provide additional information (e.g. opening hours, available services, etc.) on branch and ATM services have seen a 14% increase in Net Promoter Score (NPS), and a 7% decrease in customer support calls.⁴  Implement Branch and ATM Locator Plus today using the guide or build it in minutes with Quick Builder.

• ATM locator 1Customers can enable location permissions, or enter their address
• ATM locator 2Quickly enter the address with Autocomplete
• ATM locator 3Nearby location listings, ranked by distance and ETA
• ATM locator 4Map view and directions

Enable offers and rewards with Contextual Experiences                    

Real-time, geo-targeted offers can power deals, rewards, and cash-back programs—all visualized with rich Google Maps. By combining the insights of purchase histories with customer opt-in to location-based features, companies can implement the Contextual Experiences use case to enable personalized offers and rewards programs that drive engagement with brands while putting money in customers’ pockets at the same time.This is a win for banks and their customers, validated by encouraging metrics like NPS rating boosts of 8% or higher, and an increase of 8% or more time spent in-app.⁴  Learn how Current uses Google Maps Platform to create innovative customer rewards programs with location intelligence.

Detect suspicious transactions with Fraud Detection

With the Fraud Detection use case, companies can use customer opted-in mobile device location to flag suspicious activity based on geographic distance, such as an ATM withdrawal that is far from the customer’s phone. Our APIs can also help companies recognize suspicious transaction patterns such as a purchase made at a location that is physically distant from a recent transaction. 

Financial services companies that use geospatial data to verify customers’ identities have reduced fraudulent transactions by approximately 70%, and false positives in fraud detection by 45%, on average.⁴  Learn how Starling Bank uses Google Maps Platform to enable real-time notification of transactions and their locations, and enhance data-driven decision-making.

Start elevating customer experiences, reducing risk and increasing efficiency today with our financial services offerings. Visit our financial services solutions page to learn more about how to start implementing these solutions.

For more information on Google Maps Platform, visit our website.