The emerging role of the analyst…

6 10 2011

A few days ago, I wrote about the analyst function being dead, which spurred conversations about the emergence of a new breed of analysts. Organizations, with all of their investment in data capture, data generation, and business intelligence, still struggle to use data effectively to make decisions. With the explosion of data over the last couple of decades, the analyst moved away from business and morphed into an IT role.  The role became more about writing business requirements and providing reports than understanding data and helping the organization digest meaning from it.

Now the analyst needs to move back to a business role, but with more of a mathematics and statistics background.  They have to be curious about improving the business and have the acumen to do it.  They need to know how to blend traditional data with today’s non-traditional data feeds from blogs, social media, video, etc.  The value of the analyst is back in creating business value through relevance, context, and timeliness.

To achieve this, the emerging role of the analyst requires a new skill set and must:

  • Understand how to derive information out of data and present it in business terms – this is perhaps the most important of all of the new skills.  The analyst must be able to take a tremendous amount of information and coalesce that information into business terms leading to action.
  • Integrate various types of information – as data is coming from all different places and in new forms, it is increasingly important to understand how and when to leverage potentially rich data, and decipher what is irrelevant quickly.
  • Design problems with various concepts – the analyst needs a consultative style in which different models are applied to solve ever more complex issues.
  • Use technology – with Business Intelligence, Planning, and Predictive Analytic style software becoming easier to use, the analyst needs to know not only how to use these tools, but when to use them.
  • Delegate – traditionally the analyst needed to do everything.  Now as technology, data sources, and businesses have become more diverse, the analyst needs to know how to farm out some of the analytics to specific expertise at the right time, guide the project, and integrate the results.

The analyst also needs to transform informational projects into a process where requests for information are appropriately managed. This includes breaking down information into four areas: persistent information or basic reporting of facts on a regular timeframe; performance measures that have higher level KPIs; problem analysis; and data exploration.

Gone are the days where the analyst was a report writer, spending too much time on data acquisition.  They must now know how to enhance data to get more out of it in a timely and fashion and present that back to the business in a manner that drives value creation.

Advertisements




The analyst function is dead

8 09 2011

The role of the operational analyst has moved from the business into both Finance and into IT.  The Finance team typically focuses only upon the financial outcomes of the business and has left the operational side of the business to the IT team.

Here is a conversation a client of mine recently had with their analyst…

ANALYST: ” Here is the report on units sold this year.”

BUSINESS:  “What happened here?”

ANALYST:  “That is a spike in the data.”

BUSINESS:  “Right.  But what happened?”

ANALYST:  “That is what the data is showing.”

Sadly, this is not uncommon in the business world today.  Billions of dollars are spent every year on Business Intelligence software to help us visualize what is happening within the business, yet we are really no better off in terms of insight.

WHY is this happening?

  1. The biggest reason why this is happening is we have changed the role of the analyst.  It used to be a marketing person looking at marketing data, or operations looking at manufacturing information.  We have now moved that role to IT, or IT has promised that that can do it better with their understanding of data structures.
  2. We have wrongly assumed that a picture is worth a thousand words.  In BI terms, a chart is worth a handful of questions. IT can not predict that next series of questions and is then left to prioritize what questions to tackle next.
  3. The pace of business, or at least the pace and variety of business questions (like the data we collect) has risen exponentially and scaled faster than our ability to respond.
  4. IT is over burdened and lacks the political power and will to say “no.”  They are in complete reaction mode and lack the resources to cover the demand.

WHAT can we do to fix this?

  • First off, we need to understand the analytical gap within the organization.  IT can manage the data and needs to partner with the business, but the business needs to own the intelligence.  It is easier to teach the business a little about technology, than teach the IT resources about the business.  The business side needs to find that type of person who understands a little about technology, but has a solid mathematical or statistical mind with a curiosity about improving the business.
  • The organization needs to find a better way to integrate better analysis back into the management process.  We need to give the analysts a frame of reference in which to explore ideas and present results.  Some of this will follow reporting upon weekly/monthly operational outcomes, while most will likely by ad-hoc hypothesis or what-if scenarios about some aspect of the business.
  • The culture has to reward critical thinking.  This is not true in most corporate cultures.  All to often, the analyst is criticized for not “going along” with the current belief.  If the culture does not reward new thinking, then the analysis will quickly fall in line with visualizations that support the status quo.
  • Invest in tools and training beyond just the core cubes and reports of the BI market.  While a good portion of analysis can be done with Microsoft Excel and a data dump, the more we want out of our analysts, the more we need to give them.  We need them to look at market baskets, threshold containment, frequency curves, optimization models, assumption testing, correlations, and many other types of analytical tools.