In the earliest days of data analytics, our new clients would typically say, “Solve this problem for us.” As they saw the enduring power of analytics, their request then became, “We want to launch our own data analytics capability. Will you help us set it up?” However, growing an analytics capability from scratch is a huge challenge, and today more companies appreciate its difficulty.
Instead of expecting an analytics consultant to come in, tell them what to do, and then leave, knowledgeable organizations say, “Come in, and help us solve this problem. If you do a good job, we’ll ask you to help us with other problems. Meanwhile, we’d like your support over the long run as we set up our own analytics capability.”
Launching into data analytics without professional assistance usually leads to trouble. Most organizations need quite a bit of help over a substantial period of time before they can “take off the training wheels.” Even companies with considerable experience in the field usually require ongoing
professional support with the hard science around recent advances. You can waste a great deal of time and money trying to go it alone. And the costs really soar if a weak model leads to erroneous decisions!
Hiring a Consultant or Building a Team
If your organization intends to rely on data science on a consistent and growing basis, you will likely want to build your own in-house analytics capability. Do it gradually. Don’t rush out and start hiring data scientists. Engage a consultant to help you define your needs and develop a plan to meet them. The knowledge you gain by working with a consultant will benefit you over the long run, and in the short run you’ll do a better job of planning and staffing your analytics initiative. Also, early successes will serve to protect the initiative against competing budget priorities.
A large government agency recently asked us to help them decide whether they should increase their analytics capability by hiring more people or by expanding their use of consultants. After identifying the problems they wanted to solve and assessing their staff, tools, and data, we made a recommendation that allowed them to meet their needs in the most cost effective manner. If they had launched into their expansion without this assessment, they could have wasted money on buying the wrong kind of software or hiring the wrong kind of people.
Discerning Fact from Hype
There are many types of consultants in the analytics world, and unfortunately some may not be what they seem. A few intentionally try to inflate their capabilities, but the main problem stems from the complexity and fluidity of the field of data science, where the definitions of key roles can be confusing, and the demarcation lines between disciplines are blurred.
When you’re hiring consultants, you may have to wade through hype. Many consulting companies that offer business intelligence (BI) services accept higher-level analytics engagements in hopes that they can figure out what’s needed as they go along. Before hiring an analytics consultant learn enough about data analytics to be able to separate fact from hype. My book Mining Your Own Business, co-authored with our CEO Gerhard Pilcher, will give you a good start.
Next, do a thorough job of interviewing prospective consultants. Ask them what types of problems they have solved, and verify the results by checking references. Find out how they manage projects and what tools they use.
Evaluating Industry Experience
In most cases, it’s not necessary or even particularly helpful for the consultant to have experience in your specific industry. In most cases, “data is data.” An analytics technique that works in one domain (say Healthcare) will work fine with the same type of application in another domain, such as Financial Services. However, the consultants must be willing to listen to you and employ your domain expertise to best effect. Beware if they assume their magic software will take the place of domain knowledge.
Certain firms specialize in providing data analytics services in narrowly defined areas, such as credit card scoring, insurance risk scoring, or web analytics. Because of specialization, they are sometimes able to develop automated systems and other techniques to increase their cost-effectiveness. If you have such an “analytically mature” application area, you might want to check out vendors that specialize in your industry. Still, don’t assume that such a firm is automatically the best choice. A strong data analytics consultant, unfamiliar with your specific problem area, can often take a fresh look at your need and come up with a new and better solution.
Evaluating Analytics Experience
Although experience in your organization’s industry generally isn’t necessary, experience in analytics definitely is. Evaluating a consultant’s sophistication in analytics can be tricky, however. When hiring consultants, it’s especially important to understand the difference between data analytics
and business intelligence.
Data analytics relates to Levels 5 through 10 on the hierarchy of analytics shown in our Ten Levels of Analytics (Figure 1). It involves the use of very sophisticated analytical techniques, such as clustering, predictive modeling, machine learning, text mining, or link analysis.
Business intelligence (BI), on the other hand, relates more to traditional descriptive analysis and advanced spreadsheets. It is useful for transforming raw data into meaningful information for business analysis and decision making, but it is far less powerful than data analytics. Make sure the consultant you hire matches the need you have.
Launching into data analytics without professional assistance often leads to trouble. Even companies with considerable experience in the field usually require ongoing professional support with the hard science. When you’re hiring consultants, it can be difficult to wade through the hype. Many consulting companies, especially those that offer business intelligence (BI) services, may accept higher-level analytics engagements in hopes that they can muddle their way through. In most cases, it’s not necessary, or even particularly helpful, for analytics consultants to have experience in your specific industry, as long as they are proven experts in solving data analysis problems. Subject-matter experts (SMEs) need to be on call for the analytics team, but their commitment level is usually modest and affordable.