Delivering dashboards based on new data sources is 43% faster using a full-stack analytics platform.
How well is your data analytics tech stack meeting your current business needs? If you said not very well, you’re in good company.
Nine in 10 respondents in a February 2021 Forrester Consulting survey of analytics and business intelligence professionals said their organizations’ current analytics solutions cannot meet all of their business objectives.This is one of the key findings reported in the June 2021 study, “Accelerate Business with a Modern Full Stack Analytics Solution,” a Forrester Thought Leadership Paper commissioned by Incorta.
The shortcomings of current solutions are wide ranging. Fewer than 20 percent of respondents say their analytics tech stack enables better business agility or ability to meet customer needs. Fewer than 15 percent report seeing more innovation, business performance or faster time to market. Fewer than 10 percent say their current analytics systems and approaches help increase revenue.
In short, only a minority of organizations today appear to have analytics solutions in place that are actually helping them make significant headway on their business objectives. Prime reasons include the inability to architect for real time analytics, data security challenges, and solution integration challenges.
But what about tomorrow?
Data sprawl and velocity are already pushing existing solutions to the breaking point. It takes too long to add new data sources, make changes to existing ones, and spin up reports and dashboards. But the data deluge is not going to end any time soon. According to the survey, the average number of data sources feeding analytics solutions is set to almost double in the next 12 months, from 320 to 708.
If existing solutions aren’t helping organizations meet their business objectives now, how can they possibly rise to the occasion as even more data sources are added?
To handle the volumes of data, data complexity, and speed of analysis required today, organizations are moving away from a mix of proprietary and homespun solutions, turning instead to full-stack data analytics platforms. Besides reducing time spent integrating components and building dashboards and reports, these end-to-end solutions offer the ability to analyze data as is, without complex transformation efforts. This can really speed up the time to insights – according to the Forrester study, creating dashboards based on new data sources is 43% faster using a full-stack analytics platform.
So how can organizations implement a full-stack platform?
Following our recent webinar, we asked our guest speaker Boris Evelson, Vice President and Principal Analyst at Forrester and a leading expert in insights-driven business (IDB) capabilities, for some practical tips on how to go about making the shift to a full-stack analytics solution. Here’s his advice:
Q: How do we build a business case for a full-stack platform?
A: It depends on your current state and the level of detail you are looking for. For example, if you are still at the beginner stage of your insights-driven business (IDB) capabilities, have not yet invested in building or buying and deploying any of the key data warehouse/analytics technologies, and need an estimate for budgeting and planning purposes, please consider leveraging the 2x-6x ROI that Forrester detailed in our recent research report, Build A Maturity-Based Business Case For Insights Driven Investments.
If on the other hand you are at an intermediate or advanced level of IDB maturity and have already built, integrated, and deployed data warehouse/analytics solutions based on multiple individual components, then build your business case based on “% faster”, “% more/less likely” and other full-stack platform benefits that we uncovered in this thought leadership study.
Q: We already have a component-based solution, but are considering rearchitecting based on a full-stack platform. How do we migrate from the current environment to the new one?
A: In most situations I am not a fan of big bang approaches to migrations; rather I prefer and recommend a steady step-by-step approach. For example, consider developing all new analytical applications on the new full-stack platform, while continuing to maintain current applications on the component-based platform.
When you are ready to start migrating current applications, do it in the following four step sequence:
- Replicate data pipelines and data models in the new full-stack platform and leave the new and the current data pipelines running in parallel for now.
- Next migrate the front end (dashboards, reports) by simply pointing existing ones to the new full-stack platform based data warehouse (DW). You can use applications built on 3rd party business intelligence (BI) to query your new DW. These first two steps should be almost invisible and least disruptive to the business users.
- At this point you can turn off your current/old data pipelines and decommission the current/old DW. Now build a new and optimized data model and pipelines based on the new full-stack platform’s capabilities to analyze data as is.
- Once it is ready, migrate the old reports and dashboards to the new platform.
Q: What about vendor lock-in associated with full-stack platforms?
A: This is a good question and indeed something you need to consider and have a Plan B (but this is true for all other enterprise commercial software and applications).
To minimize risk, evaluate the platform’s extensibility via APIs – can you extend the platform to meet your requirements in case the vendor can’t/won’t? Additionally, full-stack platforms are still based on three major components: data pipeline; DW and/or DBMS, and dashboards and reports. If a need arises you can always substitute the platform’s native capabilities with 3rd party data pipeline and/or BI platforms. You will lose some of the benefits of a full-stack platform, so that should only be your Plan B to mitigate potential future vendor risk.
To learn more about how organizations are using full-stack analytics platforms to overcome the limitations of proprietary and homespun analytics solutions, explore the full Forrester study here.
Forrester Consulting Study Methodology
In this study, Forrester conducted an online survey of 215 financial services and insurance, manufacturing and materials, retail, and supply chain respondents in Canada and the US to evaluate current data and analytics solutions impact on the data to insights cycle. Survey participants included decision-makers responsible for analytics and business intelligence strategy decisions across IT and the business. The study was completed in February 2021.