Choosing the best big data partners: Eight questions to ask
Image: iStock/MangoStar_Studio
In May 2017, Interana, which provides interactive behavioral analytics, surveyed 200 people from companies that ranged from startups to Fortune 500 firms. Although 88 percent of respondents said that their companies recognized the importance of being "data-informed," and 91 percent said that having an analytics solution in place was important, most acknowledged that their companies were struggling when it came to "delivering analytics solutions that are accessible, fast, and flexible enough to give them the insights they need."
Big data plays a major role in analytics. It enters the analytics process via social media and web-based information, graphics, Internet of Things (IoT) and sensor-based input, and many other sources of unstructured data that companies take in every day.
SEE: IT leader's guide to big data security (Tech Pro Research)
Because many companies lack the internal expertise to turn big data into actionable analytics for the business, and because most businesses no longer see big data initiatives as 'new' and are expecting results, obtaining a skilled and trustworthy big data partner is more important now than ever.
Here are eight questions that will help you identify the best big data/analytics partner for your organization.
1: Do they know your business?
Big data projects, like other IT projects, are funded and supported only because the business expects value from them. To deliver that value, big data projects must have specific business cases and pain points they can solve so that C-level executives see tangible results.
Generic big data vendors that provide IT platforms and applications for processing and storing data are not as well positioned to help you define and execute your business cases as vendors that really understand the nuts, bolts, and strategic issues of your business.
You want to find a vendor-partner that has deep big data and analytics skills, as well as deep knowledge of your business and the industry sector you're in. These vendors can even help you identify relevant business cases for your big data projects because they already know the needs of your industry.
2: Do they have scalable solutions and costs?
Big data can mean big money. It's just as important to find a vendor you can work with that can scale your dollar investment based upon the amount of resources you are consuming as it is to find a vendor that has processing and storage platforms that can scale out as your big data projects grow.
When you sit down with a potential vendor-partner, you want to discuss cost scale-out as well as resource scale-out. If the vendor tries to lock you into a fixed amount that you must spend even if your big data projects don't materialize as fast as you thought they would, it's best to look for another vendor that will require you to pay only for what you actually use.
3: Are they willing to pilot their technology with you?
No matter how convincingly a vendor presents its solution, you need to know that the solution will really work in your business before you invest in it.
Vendors also understand this need to 'prove a concept', but not every vendor offers a try-and-buy policy or is willing to pilot a project and deliver some business results before you enter into a formal contract with them.
Look for a vendor-partner that will enter into a pilot proof-of-concept of its solution for your business before you buy. That will help you formulate a more long-term strategic roadmap for your company where the vendor's solutions continue to deliver value.
SEE: In-memory computing: Where fast data meets big data (ZDNet)
4: How strong is the vendor's commitment to you?
Once you sign on the dotted line, the vendor-partner should demonstrate its ongoing commitment to you through strong support services and a robust set of service level agreements (SLAs).
In the SLA area, speed of performance and 365/24/7 uptime are important metrics. There should also be SLAs in place for time to response to questions, time to resolution for open production issues, and time to restoration for disaster recovery.
In the support area, the vendor should have training and certification programs and online resources available so that your staff can get up to speed with the solution.
You, not the vendor, should ensure in your contract that vendor support and SLA commitments are documented in an appendix to the contract and that there's a way for you to break or renegotiate the contract if the vendor fails to stand by its support and SLA commitments.
5: How solid are the vendor's references?
Every vendor will handpick the best references from its client base and give you those people to contact. Instead, you should ask the vendor for a complete (or near complete) list of clients you can randomly call. This helps you obtain more objective reviews of the vendor and its performance. It's also advantageous to contact references that are in the same industry as you are.
6: is the vendor compliant?
If a prospective vendor-partner can't meet the same governance levels for security, data safekeeping, and so on, that you expect of yourself, you should keep looking for another vendor.
Your vendor-partner should be able to furnish you with current IT and financial audits. You want the former to ensure that the vendor has the level of security and data protection you expect. You want the latter to ensure that the vendor is solvent and will be in business for the foreseeable future -- because you don't want to have to change partners.
SEE: Why enterprises are finally paying up for big data security (TechRepublic)
7: Does the vendor have skills and knowledge you don't have?
In August 2017, Harvard Business School reported that "Big data is not used well. Companies are better at collecting data -- about their customers, about their products, about competitors -- than analyzing that data and designing strategy around it."
Yours might be one of the many organizations having trouble getting the most from their big data. This makes it imperative to find a vendor-partner that can bring knowledge and "on the ground" skills to your efforts, as well as a technology solution.
If the vendor can't provide deep big data knowledge and practical skills, you might be better off looking elsewhere until you find a partner that can.
8: Is the vendor open to skills transfer to your organization?
Most companies want to develop big data skills within their own organizations. What they find is that they must retrain staff in these new skills.
When you negotiate with a vendor-partner, and before you sign on with it, you should discuss your staff training goals and ideally come up with a knowledge transfer policy that everyone participates in.
This strategy goes beyond just training your staff to operate the vendor's tool set or using the vendor to provide expert consultation when you need it. It could also involve the vendor's working side by side with your staff, with knowledge transfer and project execution occurring simultaneously. The end goal of such an effort is to accomplish the project, as well as equipping your staff with the skills they need to do a similar project on their own the next time around.
Final remarks
The process of choosing a vendor-partner should be approached carefully and thoughtfully. You want fair pricing and an excellent solution set from the vendor, but you also want a partner that will be there when projects encounter difficulty and you need vendor support. The real test of a strong partner is a company that's in there working with you for as long as it takes to resolve an issue and bring your project to a successful close.
Also read...
- ASPI urges government to see beyond the 'hype' of big data (ZDNet)
- Why streaming data is the future of big data, and Apache Kafka is leading the charge (TechRepublic)
- Big Data Power Tools Bundle (ZDNet Academy)
- How to succeed with big data on a tight budget (TechRepublic)