13
WHITEPAPER How to Upgrade Analytics by Building a Single Source of Truth

WHITEPAPER How to Upgrade Analytics by Building …...visualizations and modeling in BI tools, such as Looker. A good BI tool that’s well supported by a strong warehouse can accomplish

  • Upload
    others

  • View
    7

  • Download
    0

Embed Size (px)

Citation preview

1HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

WHITEPAPER

How to Upgrade Analytics by Building a Single Source of Truth

HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Contents

1 Evolving Your Understanding of Customers Means Breaking Down Data Silos

3 Why Creating a Single Source of Truth Is So Challenging

5 What Should Your Source of Truth Look Like?

7 Customer Data as a Service Enables Your Source of Truth

9 Heap-Supported Warehouses

10 Conclusion

1HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Evolving Your Understanding of Customers Means Breaking Down Data SilosUnderstanding your customers is a journey. Most businesses take their first steps by measuring traffic and optimizing individual touchpoints and channels. However, customers interact with your business across many different channels and touchpoints. They may see an ad for your business on Facebook, decide to look you up on Google, sign up for your newsletter, interact with your emails, buy something and contact support. From a customers’ perspective, they expect these interactions to be personalized and consistent. However, each of these interactions lives in different tools and the data is siloed. Looking at individual channels or traffic doesn’t deliver enough insight anymore.

The next steps that businesses often take are to implement tools that allow them to become user-centric, and find ways to improve customer experience and optimize conversions on their websites and apps. This step in the maturity journey opens up the possibility of grouping users into cohorts and performing optimization efforts like A/B testing. At this stage, businesses start to understand their customers as segments as opposed to traffic numbers.

However, optimizing conversion rates is not the complete picture. The final step in the journey to understand customers is tying your complete customer behavioral data together with all of your relevant 3rd party data about customers to create one singular source of customer truth that you control fully.

2HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

The best approach to building a single source of truth today is centered around deploying a cloud data warehouse, where you join and unify all your customer data across touchpoints. The plummeting costs of compute and storage in the form of cloud data warehouses have made previously prohibitive efforts to build such a source of truth feasible for everyone. To create this source of truth requires a solution tailored to your specific needs and realities, as opposed to a “one size fits all” approach. There are many excellent guides for choosing the right modern data warehouse for you, and it’s a fundamental foundation for taking the next step in the journey to understand your customers.

If you’re just getting started on this journey, it can seem like a lot. Fortunately, the path has been carved out and there are lessons to be learned from others’ attempts. The idea of creating a single source of truth isn’t new. Businesses of all sizes have attempted to create a single source of truth and unify their disparate sources of customer data. So why have their initiatives failed time and time again?

Some false starts can be attributed to missteps like getting locked-in to a vendor and losing ownership of their data, as well as capacity and cost constraints on the data warehouses front for earlier efforts. However, even for the largest enterprises, the central reason for failed initiatives here is incomplete, messy, and untrustworthy data. You can’t have a single source of truth if data is getting left behind or lost in translation. In order to break down data silos and unify data in your warehouse in an accurate and thorough manner, as well as keep ownership over your data, you need to build your own customer data technology stack with tools and technologies that best support your particular ecosystem and needs.

3HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Why Creating a Single Source of Truth Is So Challenging

These days, there are excellent data warehouses and best-of-breed tools across the layers of a customer data technology stack. Choosing the right tools on their own can help you move along the analytics journey from measuring high level traffic metrics to building user-centric views of your customers. However, taking the next leap and creating a single source of truth that unifies all relevant customer data and enables highly actionable analyses is where many organizations stall out today.

What sinks initiatives around creating a single source of truth is not the quality of individual tools -- it’s getting data out of them and into the data warehouse in a well-structured, reliable, and trustworthy manner. There are major issues with joining the data together in a consistent and complete manner in one unified location. There are several crucial reasons why this is the case:

The number of data sources is constantly increasing

We live in a world of expanding data and customer interactions. The proliferation of specialized tools brings great functionality to each touchpoint, but also brings a new data source that must be brought into the fold to create a source of truth.

There are no standards for what happens outside the boundaries of a tool

Each individual tool is focused on the world unto itself. At best, a tool will provide an API of some flavor to assist with data out, but once data is leaving a tool, there are no standards or even best practices for the form, schema, or completeness of that data egress.

4HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Pipelines are fragile

Since many tools don’t focus on getting data out of them, organizations have to build their own pipelines or look to a 3rd party ETL tool in order to get the data from these tools into their data warehouse. For some sources of data, this works just fine. For other sources of data, in particular customer behavioral data, outside pipelines can cause a huge headache. When the source of data is constantly changing and needs to send new types and definitions of data, outside pipelines often break down.

The data flow into your warehouse is always changing

Despite the language commonly used around data pipelines and data flows, data is not water. It comes in different shapes, forms, and schemas, and you have to ensure that you can handle and unify all of these different data shapes and sizes together in your warehouse. It’s not something that happens automatically, and it’s something that requires a lot of attention. If you’re not able to send all of the data that’s relevant to your customers to your warehouse, then your source of truth falls apart.

5HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

What Should Your Source of Truth Look Like?So far, we’ve talked a lot about the challenges around building your single source of truth. So what are the solutions? What should your single source of truth look like in reality?

The first step is to deploy the right cloud data warehouse for your needs. Building out a stack that lets you retain ownership over your data while accomplishing the goal of unifying data in a sustainable way in your warehouse requires conscientious decision making when choosing best-of-breed tools. With a well-built data warehouse, you can unlock high-value visualizations and modeling in BI tools, such as Looker. A good BI tool that’s well supported by a strong warehouse can accomplish organizational goals of democratizing data and data access as well.

In order to most successfully democratize data and perform high-value analyses, you’ll need to focus on how the different tools you implement as you build your customer data technology stack are going to connect to your warehouse. The quality of the connections that different tools have with the main data warehouses on the market varies wildly. The first thing to keep in mind when approaching this problem is what data warehouse you’re moving forward with, as that will impact which tools are the best fits for you.

Creating a complete, single view of your customers means bringing data into your warehouse from your 3rd party data tools -- like marketing automation, email service providers, and CRMs -- and your 1st party customer behavioral data tool. In your data warehouse, you must be able to standardize that data into a consistent form and ensure that it is always up-to-date and accurate.

Some of this 3rd party data is best suited to be brought into your warehouse via an ETL tool. ETL tools are well equipped to replicate table-based data in consistent batches into your warehouse. If the format, schema, and type of data you want to send to your warehouse is generally the same, then ETL tools will do a great job.

6HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Customer behavioral data, however, requires special attention. The end-to-end process for 1st party customer behavioral data entails the ability to capture all of the data, organize it and make it sensible, and then make it available in your warehouse for analysis and joining with other datasets -- all in a fully complete and standardized manner to ensure that you can always trust what you see. ETL tools often run into problems with the ever-changing and evolving nature of customer behavioral data, making this a sticking point where single source of truth initiatives often break down.

Not every customer behavioral data tool has direct integrations with the major data warehouse providers, but this is one source of data that’s worth paying extra attention to. A fragile pipeline for your customer behavioral tool will often lead to missing and inaccurate data and require a full-time team of people dedicated to updating and maintaining it.

7HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Customer Data as a Service Enables Your Source of Truth

At Heap, we believe that you should always have all the data you need to answer any question you can think of -- complete, structured, and ready for use. Heap’s platform automatically captures every customer interaction from web, mobile, and other touchpoints and makes that data available for you to analyze retroactively -- in Heap’s analysis UI or in your data warehouse. With Heap, you always have the data that matters, and the ability to analyze what’s relevant and important, even as “what’s important” evolves.

With the introduction of Customer Data as a Service, Heap delivers the best-in-class solution for capturing 1st party customer event data and sending it to your data warehouse. As an end-to-end solution for capturing your event data, virtualizing it, and sending it to your data warehouse, Customer Data as a Service gives you the most control and smoothest process available for bringing this crucial data into your single source of truth for analysis in your BI tool. Customer Data as a Service offers:

1. User-focused schema optimized for behavioral analytics

a. Heap’s automatic data capture makes gathering customer behavioral data an automated process, so you never have to worry about missing crucial data.

b. Heap uses a standardized schema, making integrations with other data sources and modeling in your data warehouse straightforward.

c. Heap is the only tool to offer retroactive data access and syncing in your warehouse. This means historical data is always available in your warehouse, and when you update an event definition in Heap’s UI, it retroactively updates your warehouse as well.

8HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

2. Trustworthy data in your warehouse

a. Heap offers complete data in your warehouse at all times -- no caveats, no uncertain versions of the truth.

b. High data quality is guaranteed. The data that you see in Heap’s UI is always the data you see in your warehouse. It’s that simple.

3. Large savings in engineering time and resources

a. The ETL process is finicky and painful for customer behavior data. Heap’s integrations with data warehouses and fully managed ETL makes getting data out of Heap and into your warehouse fast and easy.

b. Heap reduces the amount of transform work your engineers need to do in the warehouse. Heap has a standardized schema, so there’s no need to manually derive sessions and user tables. Heap also has automatic, retroactive identity resolution, so you don’t have to manually solve identity issues during the ETL process.

In Heap Connect, Heap has built fully managed, robust data out solutions to your data warehouse of choice so you can spend time analyzing data instead of maintaining brittle ETL pipelines for your customer behavioral data.

9HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Heap-Supported Warehouses

Heap Connect brings direct integrations with Amazon Redshift, Google BigQuery, and Amazon S3. S3, in turn, integrates well with Snowflake and systems like Hadoop for organizations that deploy that sort of setup.

Heap’s integrations offer:

• Standardized user-focused schemas

• Retroactive events

• Complete datasets

• User migration mapping

• Automatic identity resolution (not available for S3)

10HOW TO UPGRADE ANALYTICS BY BUILDING A SINGLE SOURCE OF TRUTH

Conclusion

Understanding your users is a journey. In order to move up the customer data maturity curve, you must build a resilient source of truth. Your source of truth will live in your cloud data warehouse. It’s the place where you can break down customer data silos and create a 360 degree view of your customers. With a unified dataset, you can then visualize and model customer data in BI tools like Looker to unlock high value, actionable insights.

Creating that single source of truth is not a simple matter. There are many challenges and time-consuming obstacles in the way. Building your own best-of-breed stack that breaks down silos and gives you full ownership of your data doesn’t happen by accident. It requires a careful analysis of which tools can most effectively capture important customer data as well as deeply integrate with your warehouse of choice. However, thanks to advances in data warehouses, ETL tools, and now Customer Data as a Service, it is possible to realize that long sought after goal.

Heap automatically captures every customer touchpoint and automates away the pain of data. Other analytics tools require you to tag events upfront and manually instrument tracking code. Instead, Heap automatically captures everything: clicks, taps, swipes, form changes, and more. Get answers in seconds and make decisions faster.

To learn more about Customer Data as a Service and Heap’s other solutions, contact [email protected] or visit heapanalytics.com/signup for a free trial.

WWW.H EAPANA LY T I C S . COM