Embedded Analytics
Vendor Selection Guide
A holistic evaluation criteria for your
Introduction
Integrating a rich analytics offering into your software product can bring substantial
benefits, including enhanced competitive differentiation, increased customer
satis-faction, and the ability to tap new markets for growth. Further, when compared to
a ‘build in-house’ approach, embedding analytics often requires a much lower
investment of time and development resources, resulting in superior time-to-value
for the analytics initiative.
However, these advantages are only reachable with the right vendor selection and OEM
partnership – in terms of analytics feature set, compatibility with your environment, and
strength of success offerings, as well as other criteria. Too often, product teams turn
to visualization or reporting generalists for embedded analytics only to find that “good
enough” actually turns out to miss the mark, resulting in deployment delays, service
cost overruns, and end customer dissatisfaction.
As such, this document is intended to help product leaders cover all the bases during
their embedded analytics evaluation. We have categorized the evaluation guidelines
into 6 areas:
Data Access, Integration, and Management
End User Analytics Functionality
Embedding Analytics into the Application User Interface
Deployment and Security
Data Access, Integration, and Management
While visual analytics is likely what your customers will use, we see again and again that the major
challenge in business intelligence is making sure data is properly consolidated and prepared before
analysis. As such, you should find out if each vendor provides a full-fledged solution for data
inte-gration, including data blending, miinte-gration, enrichment, and cleansing. Some vendors may only
provide lightweight data ‘connectors’ that lack these capabilities. Stepping back, you may also realize
that there are a variety of potential databases and data sources you will want to access now and in
the future to enhance your analytics offering. For instance, you may want to supplement the insights
you provide with social media metrics or you might be thinking about implementing a new data
technology, like Hadoop or MongoDB.
Key Requirement Why It Matters
Can the solution perform ETL between one or more RDBMS (relational database management systems) and a data warehouse or data mart?
Being able to visually manage and schedule the process of integrating data streamlines the path to analytics
Can the solution pull raw source data (for example, from log files) and prepare it for analytics?
Ability to access diverse source data addresses a wider variety of use cases
Does the solution allow you to optimize data in a star schema for analytics with OLAP (online analyt- ical processing)?
This enhances analytics speed and performance for your users
Does the solution enable blending multi-source data in a controlled manner? (for example - bending data from two relational databases and a 3rd party web service)
Ability to blend accurate governed data from different sources expands the insights you can deliver – value of data is diminished if it is isolated
Does the vendor provide fully featured integration with Big Data sources like Hadoop distributions and NoSQL databases, including visual design for ingestion, processing, & other workflow?
If you aren’t investing in these architectures already, you are probably at least considering it. Accommodat-ing Big Data can help future-proof your analytics solution as your own software becomes more scalable and intelligent.
End User Analytics Functionality
Self-service data visualization for business users has become a core requirement in today’s analytics
landscape. However, other elements like production quality reporting, user-driven dashboards, and
custom ‘slice and dice’ pivot-style analysis are often equally important. Selecting a flexible toolset that
provides all of these different capabilities ensures that you will be able to accommodate a greater
array of user needs. It also allows you to cater to a wider variety of user roles and to ensure that you
won’t have to buy or build other analytics features in the future.
Key Requirement Why It Matters
Does the solution empower business users (not just IT staff) to create and edit reports, visualizations, and analysis by themselves?
If your customers need to call on their own IT staff to create new reports and analysis with the embedded solution, there will be significant barriers to analytics adoption.
Does the solution offer intuitive tools for production-quality row-based reporting, covering everything from standard operational reporting to account statements and invoices?
While visualization is important, many users across industries continue to demand high quality linear reporting of metrics.
Does the vendor offer data visualizations, including bar charts, bubble charts, heat grids, geographic maps, and more?
Offering a variety of visualizations enables you to provide a variety of different insights.
Can several visualizations and reports be combined into dashboards?
Dashboards are often a format of choice for keeping users apprised of several different trends at a high level.
Does the solution include a drag and drop ad hoc analysis tool, allowing users to bring a variety of data fields into a ‘pivot’ interface for custom ‘slice and dice’ analysis?
Your customers likely include business and data analysts who require the ability to create their own analysis and visualization.
Do visualizations and reports include the ability to drill through to underlying data and filter subsets of data in
Provides maximum flexibility for end users to identify the most important data points.
Embedding Analytics into the Application User Interface
As you might expect, it is absolutely crucial that the analytics solution you choose can integrate
seamlessly with your product’s front end to deliver a seamless user experience. This includes not
only the look and feel of analytics, but also the types of analytics content and functionality that can
be included as a part of your existing product. Moreover, it is also important to minimize barriers to
analytics adoption, such as additional prompts for user credentials.
Key Requirement Why It Matters
Does the vendor provide flexible embedding options – i.e. ability to embed only specific reports or a full analytics user interface inside the existing UI?
Keeping your embedding options open helps ensure that you can address changes in your user needs and analytics roadmap.
Does the vendor offer robust REST-based APIs that can call a full range of functionality from the analytics software, including but not limited to viewing and editing analytics content, report distribution, and platform administration.
The more functionality that is accessible via a modern API, the more questions your users will be able to answer from your existing product interfaces.
Can the look and feel of the analytics be customized according to your application’s branding? This includes colors, logos, buttons, menus, and more.
The correct branding and theming is necessary for a seamless user experience.
Is the look and feel customization compatible with the latest web standards, such as current versions of HTML and CSS?
This ensures maximum consistency with your product’s existing user experience.
Does the vendor have a recommended process for single sign-on (SSO) with your application? (For example, token-based authentication to serve the proper content)
Compatibility with single sign-on best practices pro-motes a secure ‘one product’ feel for your users.
Deployment and Security
Your product, like every software product, has its own unique data environment and architecture.
It should be no surprise that a successful embedded analytics solution will have to work well with
these existing systems. In order to ensure success, you need to take a hard look at the vendor’s
compatibility with your deployment types, multi-tenancy configuration, as well as your security and
access frameworks. Finding a vendor that properly aligns with your architecture will reduce risks and
costs during implementation and beyond.
Key Requirement Why It Matters
Will the vendor’s analytics work in your products that are both deployed in the cloud and on-premise?
Analytics should be available to all of your users, regardless of how your software is delivered to them.
Does the solution work with your multi-tenant archi-tecture? Can the solution isolate or share reports, data sources, databases, and/or theming according to customer organization?
It is mission critical that the right customers and users see the right analytics and only those analytics. In some industries, it is required by law that customer data be separated.
Can the solution be configured to work with common services for security and access, such as LDAP or Active Directory?
Ensuring that your product’s access roles, groups, and security can be leveraged seamlessly for analytics.
Can the solution be configured to work with common access control frameworks, like Spring for security?
Ensure access to the software is controlled in a secure fashion.
Can the analytics software be extended to work with custom security hierarchies?
Not all access & security frameworks are the same – so flexibility can become quite crucial.
Extensibility, Customization, and Transparency
For anyone that has developed a product from a first concept to a finished customer-facing
solution, it is clear that software requirements and roadmaps change over time. In order to meet
evolving requirements and reduce risk, you not only need an ability to develop extensions to the
software but you also need to have a strong understanding of how it works – in terms of technology
standards as well as the vendor’s future plans for the platform. Good technology partnerships are
based on transparency.
Key Requirement Why It Matters
Does the vendor provide an open architecture that your team can extend and customize without having to wait for the vendor’s future releases? For example – to develop your own plugins for new data visualizations, analysis, or data integration workflows.
Your embedded requirements are not only unique but may change over time as your offering evolves. Open analytics architecture helps ‘future proof’ your solutions.
Is the technology based on open standards that the team can easily understand and take advantage of? Examples include Java, J2EE compliance, authentication, authorization and REST APIs.
If you want to extend the solution, you will be able to work more quickly and effectively if your team is already familiar with the frameworks involved.
Does the solution give you visibility into their roadmap and source code? Is there an active developer community?
Providing for these requirements boosts technology alignment between you and the analytics vendor, which tends to maximize project success.
Is a visualization API available if you want to create your own visualization or connect to third party charts?
Maximizes the visualization possibilities for your analytics.
Can you embed analytics libraries and/or engines into your own server-side source code? For instance, if you wanted to use a back-end reporting engine with your own separate front-end interface.
Allows you to meet a wider variety of business needs over time.
Services, Support, and Vendor Experience
A partnership is more than just software – you need a vendor with the staff and experience to
optimize your embedded analytics deployment and avoid the mistakes that others may have made
in the past. As illustrated thus far, embedded analytics involves unique proficiencies in integration,
architectural alignment, and customization. These abilities do not necessarily fall under the core
competencies of BI providers that sell stand-alone 100% proprietary products. Even if you are fairly
confident in your team’s ability to execute on your goals, it is still a good idea to select a vendor that
can ‘hold your hand’ and walk you through any challenges that may arise.
Key Requirement Why It Matters
Does the vendor have a stated success method- ology and deployment timeframe for embedded analytics projects?
Having an up-front plan, even at a high level, will begin to reduce risk. This also indicates vendor experience.
Is there a dedicated services group to support embedded analytics or OEM deployments?
The embedded use cases are usually different than direct BI projects for internal use – you need domain-specific knowledge.
Does the vendor offer collaborative architecture ses-sions with your team, including deployment best prac-tices and a personally tailored implementation plan?
Though your technical resources may be strong, it is still important to provide them with a low-risk plan to help them avoid potential technology pitfalls and deploy the software for ongoing customer success. Does the vendor provide training for your staff, to help
them become both proficient users of the toolset and provide adequate support to your own end customers?
Your people must know how to use the analytics in order to maximize results and support for your customers.
Does the vendor provide developer enablement – i.e. support to answer questions that your technical staff may have as they deliver the embedded solution?
This streamlines both implementation and ongoing solution delivery.
commit-Conclusion
We hope this was a useful set of guidelines for you as you evaluate possible embedded
analytics vendors for your product initiative. While you may have already had a fairly
complete vision of what your end user analytics will look like, you should now be better
prepared to address key buying criteria related to data preparation, architectural
alignment, platform flexibility, and vendor experience. Your analytics initiative is
mission-critical not only for your team but for your customers and your software’s overall value
proposition as well. As such, a holistic approach to vendor selection can lead the way to
reduced risk, rapid go-to-market, and more successful outcomes for your product and
your business.
Learn more about Pentaho Business Analytics
pentaho.com/contact
+1 (866) 660-7555.
Global Headquarters
Citadel International - Suite 340 5950 Hazeltine National Drive Orlando, FL 32822, USA
tel +1 407 812 6736 fax +1 407 517 4575
US & Worldwide Sales Office
353 Sacramento Street, Suite 1500 San Francisco, CA 94111, USA
tel +1 415 525 5540 toll free +1 866 660 7555
United Kingdom, Rest of
Europe, Middle East, Africa
London, United Kingdom tel +44 (0) 20 3574 4790 toll free (UK) 0 800 680 0693
FRANCE
Offices - Paris, France tel +33 97 51 82 296 toll free (France) 0800 915343
GERMANY, AUSTRIA, SWITZERLAND
Offices - Munich, Germany tel +49 (0) 322 2109 4279 toll free (Germany) 0800 186 0332
BELGIUM, NETHERLANDS, LUXEMBOURG