In our last instalment we talked about some of the organizational issues that impede the spread of analytics. But we would be remiss if we didn’t make clear that sometimes the failure can be laid at the feet of the analysts themselves. Over the past 20 years I’ve seen good analysts and bad ones.

Contrary to popular opinion, the difference between the two is very rarely about their analytic capabilities. Of course, analytical skills themselves are important, and some people are better at crafting creative solutions to tricky marketing challenges or diagnosing what is going on in the customer base. But good analytical skills are really only the price of entry. In this instalment we’ll outline some of the factors that help analysts do their jobs more effectively and contribute to the growth of analytics within their organization.

Understand the business

Analysts must understand the industry and the company they work for. That may sound obvious, but I’ve seen many analysts who have tunnel vision about their statistical expertise and have no idea how their organization actually operates in the marketplace. Some questions to explore include: How does the company make money (this is not always as obvious as it might appear on the surface)? What is the competitive environment? What are the major issues facing your industry and your customers? There are many sources of information on these issues, including annual reports, industry websites and business publications. Investing a little bit of free time will reap big dividends in your ability to make relevant and impactful analytical recommendations.

Treat internal groups as if they were customers

Whether analysts are attached to a specific marketing team or serve in a centralized support group, the reality is that Marketing pays the bills, so you would be wise to treat them as if they were your customers, because that’s what they are. Fundamentally, that means listening to what they are trying to accomplish, recommending an approach that will work, engaging them in the analytical process to the extent that they need or want to be engaged, and providing regular status reports as to milestones and delivery dates. In short, you want to be part of the solution, not part of the problem. As they come to look upon you and your colleagues as valuable allies they will seek out your help more often and be more open to your recommendations.

Unfortunately, analytical groups often take a different approach, acting more like they are doing Marketers a favour just by showing up. One organization we did work for had a centralized analytical team housed in another country. Periodically they would fly up to Canada and deliver a predictive model to the local Marketing team. Usually that model was not chosen by the local team, was not needed by that team, and thus was shelved as soon as the analysts got back on their plane. No one would think to engage them in helping to solve any of their problems. Good for us, as outside consultants, but arguably not good for the organization that had the talent but ended up wasting it.

Speak “business” not “statistics”

Analysts are not part of some divine priesthood whose secrets cannot be shared with mere mortals. If analysts can’t communicate their ideas in ways that are understandable to end users then they won’t be doing their jobs properly.

Nobody cares that you can derive some long-dead Russian statistician’s algorithm from first principles.

So drop the jargon and explain yourself in terms that your audience will understand. That doesn’t mean talking down to them – it means explaining objectives and methodologies in business terms, not statistical ones. I’ve long maintained that an analyst who can’t explain their approach to a non-technical audience (like, say, their mother or most CMO’s) doesn’t really understand what they are doing.

Take the example of proper sample size, an on-going area of debate. You can bring out all the formulas you want to justify a sufficiently large sample for a given test cell. But if you explain to the marketer in charge of the program that an insufficient sample size means that any learning they think they will gain may not be repeated in subsequent campaigns, and that will mean that they may not get their bonus, or could even get fired, then you are more likely to make your case for proper test design.

Help marketing prioritize analytical resources

A key role that analysts can play is helping the organization as a whole prioritize scarce analytical resources. Analysts often have a better idea of the areas where their efforts offer the greatest potential than others in the organization. A proactive approach at budget or planning time can make all the difference in ensuring that the right resources are applied to the right projects.

Sell yourself and your craft

As a species, analysts tend not be very good at selling themselves. To some extent this is the communication issue we talked about above. But more fundamentally I think it stems from the view that if they can see good results from their efforts then they assume that other rational people can see them too. Unfortunately, that is not always the case. Analysts would be advised to trumpet their success at all appropriate points to make sure that the powers-that-be understand where some of the value added came from.

Failing to do this can be bad for both the firm and the analyst. I once worked with a colleague whose prior position had been part of an analytical team in a financial institution. His entire group had been fired because, he believed, they had been unsuccessful in convincing their organization that they were adding value. When times got tough, they were among the first to go. I have no doubt that they were adding value, but the organization reverted to a “seat-of-the-pants” approach to marketing and eventually got taken over by a competitor.

If there’s a theme to these suggestions, it’s that good analysts understand that they are part of a team. They try to make it easy for the rest of the organization to work with them and in the process help enhance the role and impact of analytics. They are not the only force pushing in that direction, but if they remain locked in an ivory tower or fail to effectively engage those around them, it will be that much harder for analytics to deliver on its promise.

http://dmn.ca/Articles/Issues/DM_Mar2010.pdf