The past decade has seen a resurgence of interest in data. To name a few, there are poll trackers for elections, sports blogs that dissect the performance of your favorite team, and fitness tracker apps that show you stats on your health and wellbeing. Even products and services we use in our daily lives now have analytics on usage. For example, utility companies show you how much energy or water you consumed, and credit card companies show your spending patterns. Data is everywhere!
Companies are only scratching the surface when it comes to providing data visibility to their extended ecosystem of customers and partners. With data becoming so ubiquitous in the everyday, it’s no longer a question of whether or not to offer users insights, but how and to what extent. Today, our utility companies don’t just offer a single metric on usage, they show your usage patterns over time and how that compares to your neighbors’.
Constituents are hungry for data, and exposing data to them can create massive upsell and cross-sell opportunities. The embedded analytics market is projected to grow from $24 billion to $46 billion over the next five years.
Embedding Analytics Isn’t Easy
Building analytics into your application is an uphill battle. You are building a product that’s [likely] not your core offering. You will have to divert engineering time and resources to build it, and more resources to maintain it over time. You’ll have to keep it in mind when you upgrade your technology stack in the coming months and years. Building an analytics portal is expensive in time and resources.
Enter embedded analytics vendors. Over the past decade these vendors have enabled many businesses to bring analytics into their product with varying degrees of customization and difficulty. But these have primarily been either at smaller data or user volumes, and with limited ad-hoc analysis or data governance capabilities.
With the consumerization of the enterprise, there’s a huge opportunity in the market for next generation embedded analytics solutions. These should provide a consumer-like fast and easy user experience within the context of the native application. This is how the best companies are satisfying their consumers’ appetite for data.
How To Give Them What They Need
So what does it take to build an embedded analytics product that’s ready for the next decades of business needs?
1. Your Portal Must Be Flexible.<br>The process of getting insights from data needs to be flexible. Our personal lives have been transformed with search by companies such as Google, Amazon, Kayak and LinkedIn. We are used to asking a question and immediately getting an answer. A customer analytics portal needs to be more than just a few static charts. Customers should be able to ask additional questions and go deeper into their data. And this flexibility cannot come at the expense of hours of custom work by business analysts curating reports to answer those questions while consumers wait.
2. Response Times Must Be Fast.<br>When someone asks a question, response times should be sub-second—even at the scale of billions of rows of data, across a variety of data sources. Today, companies looking for embedded analytics solution providers are looking for a single solution that can connect to all of their data. These companies have product usage data sitting in Hadoop data lakes, MongoDB or Cassandra instances, billing data sitting in their ERP systems or data warehouses, and customer account information from Salesforce or other CRM solutions in the cloud. With the consumerization of the enterprise, customers want instant responses from traditionally “enterprise” web experiences. So, the analytics provider needs to return results from billions of rows, across multiple sources, in less than a second.
3. Data Governance Needs to be Granular.<br>Providing fine-grained data security and governance at scale is more important than ever. Embedded analytics solutions that can provide pretty charts are par for the course. But buyers of embedded analytics solutions want to easily provision granular column and row level access control to a single, shared data model for their users. For some companies, this can be hundreds of thousands of security groups. Giving data providers the ability to easily manage permissions on a granular level will help them minimize the time required to upkeep security and governance rules in the long run.
4. It Should ‘Just Work’ with Your Schema.<br>Real-life schemas are complex, and embedded analytics should “just work” with them. An embedded analytics solution provider should not require a months-long project to remodel your data to fit their modeling limitations. Even complex queries across multiple data sources and complex schemas should result in fast, accurate answers.
5. It Must Scale Without Falling Down.<br>Companies need the ability to start small, and then grow their footprint organically. The embedded analytics user experience has to stay consistently fast and easy, even when the system grows to accommodate millions of users and billions of rows of data. Companies also want pricing models that are adoption-friendly—a “per-user tax” inhibits usage. What is preferable is a model that scales as data consumption grows.
Next Generation Embedded Analytics
ThoughtSpot’s search-driven embedded analytics experience answers these needs of business leaders who are looking for a next generation embedded analytics solution provider. With ThoughtSpot, companies can offer their data consumers the familiar experience of search, backed by powerful relational search technology that calculates answers on the fly across billions of rows of data from any source. ThoughtSpot helps simplify the data pipeline, allowing IT and BI teams to blend data from any source. And this experience can be delivered to end users with row level security, keeping necessary data governance requirements intact.
We recently hosted a webinar with the Aberdeen Group, The Power of Embedded Search-Driven Analytics—Injecting Intelligence into Every Business App. We discussed leading strategies on how to embed analytics and empower customers, and the impact this can have on end users. We also discuss challenges in implementing an effective embedded analytics solution, and reviewed key use cases where companies have been successful. Watch it on-demand to learn how you can get started with your embedded analytics strategy.