In predictive analysis, an analyst collects and organizes data, selects the most suitable mathematical model, and draws essential conclusions based on the results. In other words, predictive analysis uses mathematical modeling tools for decision-making about unknown facts, characteristics, or events.

Now, you might wonder, what advantages can this type of analytics have for my business? Below, I outline the benefits of predictive analytics to help you confidently embrace it.

Optimize decision-making capabilities

With predictive analysis, companies or organizations can discover patterns in customer purchasing or interaction behavior, assess which methods promote or hinder the benefits of these patterns, and decide which measures to choose to improve the organization’s success.

Strengthen risk management

Each sector faces a degree of risk in its daily operations. Proper risk management allows businesses to grow and do so in an appropriate manner. Predictive analysis can analyze large volumes of data to anticipate and prevent fraud, identify weak points, and avoid significant financial losses.

Companies that use predictive data analysis can learn from their failures and create a more prosperous future.

Boost customer loyalty

Before customers begin their journey, we must understand their desires and behaviors. Thus, we can offer them a unique alternative. With predictive analytics, you can create customer profiles that anticipate individual needs and address their “whys” by combining existing data.

Analytics has allowed businesses to provide a more personalized and trustworthy customer experience, opening the doors to new customer habits. It’s what’s been called a “predictive analysis customer retention strategy”.

Refine targeting and segmentation

A good segmentation strategy is essential for retaining existing customers and acquiring new potential customers. Both brands and customers benefit from segmentation, as it improves marketing strategy and audience experiences while increasing brand awareness and customer loyalty. Extensive use of predictive analysis techniques helps target markets based on accurate perceptions and indications about customer interests regarding what the company offers. Therefore, you can make data-driven decisions for every element of your organization using these predictive analysis practices, which will allow you to recognize entire markets that you were previously unaware of.

Reduce costs and maximize budget efficiency

Predictive analysis allows you to save money by optimizing your budget.

For example, companies can now make personnel selections based on a predictive algorithm. It also saves money by detecting risks and directing resources effectively to address them.

Predictive analytics also reduces costs through fraud detection and prevention. Historically used by banks and credit card companies, other businesses now benefit from it. Organizations can identify fraud patterns and prevent exploitation using predictive analysis.

Do you want to learn more about predictive analytics and how to properly activate your data? Our experts can help you! Contact us and tell us about your specific case. We look forward to hearing from you! 🚀

Contact us

Google’s BigQuery serverless data warehouse makes life so much easier and so much safer in terms of having all the data in one location however, the house-keeping and data governance are left for you to administer and manage.

At Making Science we work with numerous Google BigQuery customers, some small, some large and some very large corporate groups.  Most of our experience shows that customers are rolling out data governance tools but most of our customers are typically still offering the “SQL knowledgeable/connected” departments direct access to BigQuery. This all in the spirit of self-service analytics moving their business to be more data-driven.

Even with the advances in Google’s data catalog and data governance capabilities, the data administrators are left with challenges keeping data lakes and data warehouses operating efficiently.

The Making Science teams are always looking for ways to make themselves more effective at performing their work with the construction of methodologies and tools.  Making Science has tools specifically for Data Quality projects, Data Discovery and a most recent addition is BigQuery Intelligence.  The tool offers value to the roles presented in the data governance model below:[image_with_animation image_url=”48047″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]The BigQuery Intelligence tool was built to help the data warehouse/lake architects/administrators/custodians manage house-keeping activities regarding datasets, tables and views and optimize these same objects.  Many of our larger and very large customers have 100’s of datasets and 100’s of tables and 100s of views within each dataset making it very challenging to keep on-top of house-keeping activities.

BigQuery provides data lake administrators with a simple visual interface for listing datasets, tables, and views.  The BigQuery explorer panel interface allows you to collapse tables and views under their given dataset simplifying the visualization.[image_with_animation image_url=”48053″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]But the BigQuery Explorer does not offer data lake administrators a simple visual interface to understand usage statistics regarding datasets, tables and views; this is where BigQuery intelligence collects underlying data previously collected by the GCP + BigQuery environment and offers visualizations to understand the usage patterns across the nodes in the data warehouse.

The Data Scientists and Data Engineers at Making Science had a look and found that GCP offers a tremendous amount of data lake usage statistics. But, there was no easy way to retrieve the data and visualize the usage patterns in the data lake. The team chose to use Sankey visualizations, built with Python and D3, to better understand the usage patterns between the nodes (datasets, tables, views and jobs/queries) in the data warehouse. [image_with_animation image_url=”48059″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]As the team worked with some very large data lakes they quickly found it necessary to allow data custodians to apply filters to limit views of the data lake to make the Sankey diagram more readable.  Most typically they would select logical groupings of datasets to give a complete view but without making the visualization undecipherable.  BigQuery Intelligence actually allows filtering both on datasets and tables.[image_with_animation image_url=”48104″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]The team also found that by including the possibility to define color based on criteria; by default, the color criteria is “dataset” so that each dataset is distinguished by a specific color in the Sankey visualization.[image_with_animation image_url=”48073″ animation=”Fade In” hover_animation=”none” alignment=”” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]The team decided to include a “change color criteria” filter, where we can choose between additional criteria including bytes, jobs, inherited jobs, recency, and inherited recency. With this kind of plot, we can check if there exist inefficient nodes in the data lake.[image_with_animation image_url=”48079″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]

HotSpots

A couple of use cases helped by changing the color criteria is the location of “hot spots”, i.e. those nodes which process too many bytes or have too many jobs, and “cold spots” or those nodes that have not been used in a long time and hence could be removed/deleted from the data lake.  In the following screenshot, we can see that the color indicates the volume of data each node has processed. Red indicates too much, and green indicates a reasonable volume. Grey indicates that we don’t have information about this particular node. It is interesting to see nodes that have a deep level and are very red. This usually means that those nodes could be optimized by creating aggregated/intermediate tables.[image_with_animation image_url=”48085″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]

Unused nodes

If we change the color criteria to “inherited recency”, we can search for nodes that have not received any query/job in a long time. The meaning of “inherited” is that we take the smallest recency among one node and its children, given that if one child node receives a job, in some way, the parent node also receives that job. Here, red indicates very high inherited recency and green indicates very small inherited recency. Grey, indicates that we have no info about the particular node. We could analyze the plot to search for unused nodes (red and grey) and clean the data lake.[image_with_animation image_url=”48091″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]

Metrics

The team included a mouse-over capability to display relevant metrics of a node, as you can see in the following screenshot:[image_with_animation image_url=”48097″ animation=”Fade In” hover_animation=”none” alignment=”center” border_radius=”none” box_shadow=”none” image_loading=”default” max_width=”100%” max_width_mobile=”default”]

The Conclusion

Making Science teams, acting as Data Custodians for some of our customers, have found that conversations with Data Business Owners/Stewards regarding optimizations and housekeeping are greatly facilitated by the use of the BigQuery Intelligence tool.  Sitting down with the Data Steward with data and visualizations in hand enables a transparent, data-driven conversation.

The value to the business from a tool such as BigQuery Intelligence comes in the form of improved performance for systems accessing/consuming the data and ensures a reduced GCP operating costs.

The Making Science team members have feature additions already in the works and we will share some of the use cases and the corresponding benefits as they are available.

Currently, a major issue faced in e-commerce is correctly measuring and attributing the conversions of our digital assets. Who or what channel has made our users decide to buy our product? How can we assign or attribute that sale and to which channel?

Today we bring you some solutions to the attribution problem. But for this, we must first ask ourselves what factors are causing this problem. Let’s dive in!

Cookies and AdBlockers

Cookies are currently in the spotlight because many browsers are stacked against them. Safari, for example, blocks third-party cookies and limits first-party cookies. Other platforms, such as Google Chrome, also are fighting to eliminate them. Added to this are the adblockers that are installed by default on many of our computers. These can limit any script that we insert into a page.

How can we solve this problem?

Although certain digital analytics tools work without cookies, they provide us with very limited information. For this reason, we recommend working with the usual measurement tools such as Adobe or Google Analytics 4.

If you are using Google Analytics 4, we suggest integrating it with Google Consent Mode. This solution will allow us to work without cookies. In other words, if a user does not accept cookies, Google Analytics will continue to receive pings without the user’s personal information. However, it can still model this data to identify users and sessions. In this way, we can continue measuring even those users who did not accept cookies.

Another way to overcome the limitations associated with adblockers is to think about solutions such as server-side tagging, for example, Google Tag Manager server containers. They will prevent the blocking of the scripts or hits that we send since they will be sent to our own domain.

Transaction transfer

On many occasions we’ve seen that the transactions (or purchases) made are not being sent. With tools like Google Analytics, we must insert a tracking code on the thank you page. However, in some scenarios, these transactions are not reported.

The reasons are diverse. Many times the user is redirected to a point-of-sale (POS) to make the payment, thus leaving our website. The problem is that many times the POS never returns the user to the thank you page or the user directly closes the browser before reaching it. As a consequence, the transaction is never reported and Google Analytics does not show it in its reports.

How can we solve this problem?

Our recommendation would be to send the transactions from server to server. To do this, Google Analytics 4 offers us the measurement protocol, through which we can send the transaction directly from our e-commerce platform. In this way, even if the user does not reach the thank you page, at the time they report to the e-commerce platform, we can automatically send it to Google Analytics 4 and view it in our reports.

Attribution models

The third problem we commonly encounter is related to the attribution models we use. We typically work with a last-click attribution model that attributes all credit for the conversion to the last channel the user interacts with. This model prevents us from being able to attribute said conversion to the channels that presented the brand to the user or that participated in the different steps of the conversion funnel.

How can we solve this problem?

In order to go one step beyond last click, we suggest a data-driven attribution model. This model is already available in Google Analytics 4, although previously we also found it in the Universal Analytics 360 version.

With this model, we can analyze and understand the collaborative nature of our channels. That is, it allows us to assign part of the credit for the conversion to all the channels that had a relevant participation in said conversion.

We also suggest using your own attribution model, for example with BigQuery. With it, we can create our own attribution model in order to add the value of the contribution to final conversions.

After having analyzed the different factors that cause attribution issues and its different solutions, we can say that by following these tips you will improve the way you measure and attribute your conversions while always respecting current regulations. Do you need help with the implementation? Our team of experts can help you. Do not hesitate to contact us!

Get in touch

Within the new, increasingly private digital ecosystem, advertisers and agencies must adapt their strategies by adopting the most appropriate technology. This will allow them to continue carrying out their advertising campaigns. And, it will allow them to continue to measure, analyze, and activate their data similar to how they have in the past. Today we’re talking about Google Enhanced Conversions. This is a Google new feature that makes it easier to manage ad campaigns and improves the accuracy of conversion tracking while respecting user privacy. This is accomplished through the use of conversion modeling.

Enhanced Conversions lets you see conversions across Google Ads campaigns, even if the interactions occur across devices or via email. It allows companies to properly track the results of their campaigns while respecting the privacy of consumers.

With Enhanced Conversions, you can maintain high visibility of your data, making it easier to personalize campaigns in the future. But how does it work?

How Enhanced Conversions works

When a customer completes a conversion on your website, you receive their data (such as email address, name, home address, and/or phone number). This data can be captured via conversion tracking tags, sent to Google in its encrypted form, and then used to improve your conversion measurement.

Depending on the type of Enhanced Conversions we use, the hashed data will be used in different ways to improve measurement. The main types of Enhanced Conversions are for the web (online sales), and for leads (for offline sales). In the table below, we describe the differences between the two:

Enhanced Conversions for web

With Enhanced Conversions for your website, customers’ own data, such as email address, name, home address, or phone number, is captured in your conversion tracking tags, encoded, and sent to Google in their encrypted form. Then, they are used to link your customers with their Google accounts, which they logged into when they interacted with one of your ads.

Its main benefits include

• Recovering conversions that otherwise would not have been measured
• Improving the optimization of offers through better data
• Protecting privacy by hashing your own customer data

Enhanced Conversions for leads

With this functionality, you can upload or import conversion data into Google Ads using customer data from your website’s lead forms.

Unlike standard offline conversion imports, Enhanced Conversions for Leads doesn’t require you to modify your lead forms or customer relationship management (CRM) systems to receive a Google Click ID (GCLID). Instead, it uses information that’s already been captured about your leads, like email addresses, to measure conversions.

Its main benefits include

• Ease of configuration: you can configure the measurement directly from your Google Ads account.
• Better performance: you will be able to optimize your campaigns for sales and transactions that occur outside of your website.
• Flexibility: you can implement it using Google tags or with Google Tag Manager.

Enhanced Conversions can help you better monitor the performance of your campaigns. It also facilitates the administration of Google Ads campaigns, all while respecting the privacy of the users. This represents a great advance in the use and activation of data correctly and respecting current regulations. Do you want to implement Enhanced Conversions? 🚀

Get in touch

BigQuery is a Google Cloud database that allows you to process huge amounts of data in just a few seconds. When used in conjunction with Google Analytics 360, you can import Analytics data automatically, without needing to process it. This is where BigQuery really shines. You can utilize the same data that is in Google Analytics 360 without any limitations, making BigQuery the perfect analysis tool.

What are the advantages of BigQuery?

No Sampling

One of the limitations that our Google analysts have found inside the Analytics interface is  “sampling.” This most commonly occurs in personalized reports or a wide range of dates. With sampling, Google extrapolates the data it receives – counting only some of its data points and modeling the other information. As a result, the data that appears is approximate. With BigQuery, we can avoid sampling. We can include all unprocessed data, allowing us to consult, divide, or modulate it using 100% real data. It’s truly powerful!

Users as protagonists 

In Google Analytics, objectives are based on sessions. If a user completes an objective after multiple visits, it will not be observable in the Google Analytics funnel. On the other hand, BigQuery has the data as well as the capacity to show user-level information. This can have a great impact for ecommerce focused companies. If you’ve used the improved e-commerce reports inside Google Analytics, you’ll know that these are based on sessions. However, in the real world, a person may add an article to his cart during one visit and wait to complete the purchase on another visit. BigQuery allows you to monitor the purchasing behavior of users that need more than one session to convert.

Combined dimensions 

The data model in Google Analytics is structured so that session based dimensions (like campaigns) don’t work when combined with user or hit level dimension and metrics. There’s a limit to the number of dimensions we can see: usually only two within Google Analytics, or five within personalized reports. Using an API, we can achieve up to seven dimensions. But with BigQuery, there are no limits. For example, ecommerce companies may have trouble getting exact statistics about the quantity of social media users that saw a product page and then bought that product. In BigQuery, we can get that data with just one query.

Apply changes retroactively

Google Analytics has a specific way of processing data, for example with objectives or filters. When we gather data in Google Analytics, whether by web or app, Google will apply our configurations and store this data for future reports. As such, it’s not possible to go back and change data in Google Analytics. But with BigQuery this all changes. If you made a mistake when configuring your reports and you wish to filter or modify data (settings of objectives, channel groupings, etc.) you can adjust your consults in BigQuery to correct those issues retroactively. 

Now what??

If you’re not sure how to get started with BigQuery, get in touch with our team by shooting us an email at info@makingscience.com.