With the explosion of business analytics (the business analytics software market is poised to reach over $59 billion by 2018) and a widely publicized lack of analytics talent (a McKinsey report estimates that by 2018, about 140,000 to 190,000 big data jobs will be unfilled due to a lack of applicants with expertise and experience), many enterprise organizations, from IT to marketing, are turning to analytics service providers for help.
Leaders seeking to build a competitive advantage, as well as drive their personal career development, are now asking how to evaluate and select the right analytics partner. This decision can’t be made lightly, as the consequences can easily make or break a critical initiative, or a career.
Analytics service providers range from tiny niche boutiques to midsized established consulting firms to massive organizations embedded inside larger consulting or platform providers. Offerings range from single-point solutions or projects to specialty industry expertise to retained teams with broad and deep expertise.
Which provider you select depends on the following factors:
- The complexity of the business problem(s)
- How you are currently addressing the issue
- The nature of the desired solution
- The depth and level of service desired.
These are selected criteria you might want to use to evaluate analytics providers.
Complexity of the Problem
Is your problem well defined or ambiguous? Your answer will determine the solution approach and depth of problem-solving experience you will need when working with your analytics provider.
Considerations when evaluating your analytics provider include the following:
- Degree of complexity – simple or massively complex
- Degree of experience – historical issue or new
- Organizational scope – single department or enterprise-wide
- Timing – transient or recurring issue
A sample question or proof you should ask of candidate providers: “What approach do you use to collaborate on problem definition?”
Current Approach
How extensive are your current analytics and big data capabilities? Your answer will determine the gaps that your analytics partner needs to fill to augment your current approach.
Considerations when evaluating your analytics provider include the following:
- Current analytics sophistication – internal and other provider skills, infrastructure, tools
- Types of data – structured, unstructured, streaming
- Data sources – internal, public, third party
- Industry and problem expertise – broad, deep, relevant
A sample question or proof you should ask of candidate providers: “Can you demonstrate expertise in the range and depth of my internal and external data?”
Solution Requirements
The solution you need will depend of the type of analytics you need to build, how they will be operationalized, and whether you need ongoing support.
Considerations when evaluating your analytics provider include the following:
- Desired analytics sophistication – machine learning, artificial intelligence, automation
- Innovation – proven approach , push the envelope, bleeding edge
- Scale tradeoffs – speed, scope, accuracy, consistency
- Customization – out of the box, tailored, completely customized
- Data security – regulatory compliance, on-site versus off-site protocols, certifications
- Implementation – integration across platforms, data connections, and decision processes
- Operationalizing – business rules, visual storytelling, optimization flows
A sample question or proof you should ask of candidate providers: “Can you demonstrate expertise in building and operationalizing machine learning?”
Provider Requirements
Once you have defined the problems you want solved and identified gaps in your current approach and the parameters of your desired solution, you are prepared to define the type of partnership that will best fit your needs. Are you seeking a short-term project solution or a long-term partnership that will scale and position you and your enterprise-wide initiatives for highly visible success?
Considerations when evaluating your analytics provider include the following:
- Problem-solving capabilities – data management, insight development, predictive modeling, machine learning, automation, visual storytelling, systems integration
- Industry experience – same or relevant, cross-pollinated ideas
- Team structure – analysts, consultants, data scientists, big data engineers
- Engagement approach – strategic partnership, project, ongoing program
- Process – solution requirements, milestone management, problem resolution, satisfaction measurement
- Skills and training – technical skills, advanced analytics, consulting skills
- Culture – flexible, reliable, supportive, client service orientation
- Results – impact, insight, and innovation delivered
A sample question or proof you should ask of candidate providers: “What can you tell me about your employee engagement and how you attract and retain the best talent?”
And of course, you will want references and a check for ethical business practices. Some companies seek business at any cost, and the sooner you know you can trust your provider, the better. Test your provider with hard questions that challenge its integrity and run for the hills if the provider outright lies or promises the moon. This wrong choice can cost you and your company dearly.
Careen Foster, SVP and Chief Marketing Officer, Fractal Analytics
Careen Foster is the Chief Marketing Officer at Fractal Analytics. Throughout her career, she has been on the forefront of Big Data analytics space as the head of product management and partner marketing at FICO’s Scores division (Fair Isaac), and product management and operations for risk management solutions at TRW (now Experian).
Careen holds an MBA from the University of California, Irvine, Paul Merage School of Business and a Bachelor’s in Psychology, from the University of California, Irvine.