When establishing, expanding, or staffing an analytics and/or reporting function within your organization, there are a few key considerations that will strongly impact the types of services the group can provide, how the group provides value to the firm, what skillsets should be considered for each role, and what kinds of tools will be required in order to adequately support your Analytics function. Getting this right isn’t difficult, provided there’s an honest assessment of what this role’s major responsibilities will be. A small amount of effort in planning and staffing ensures that your organization gets the timely operational information it needs to make day-to-day tactical decisions while leaving you positioned to deliver strategic insights and shape the Analytics function into a revenue-producing powerhouse.
To set some background we go way back to 1992. Wayne’s World graced the big screen, the Maastricht Treaty was signed, founding the European Union, and Robert Reich presciently described in his book “[amazon-product region=”us” text=”The Work of Nations” type=”text”]0679736158[/amazon-product]” how two major business trends — a shift from “high volume to high value” production, and increasing globalization — would impact the nature and value of work over the coming decades. Reich identified three major classifications of jobs that would comprise the global work force. These classifications are highly relevant in examining the roles we all serve at work, and they’re indispensable in describing the roles we want analysts to play in the firm. (It’s worth a moment’s though reflecting on where you fall within this scheme, as well.) Reich’s job classifications are:
follow link 1. Routine Producers are responsible for “the kinds of repetitive tasks performed by the old foot soldiers of American capitalism in the high-volume enterprise.” These roles include data-entry clerks as well as “foremen, line managers, clerical supervisers, section chiefs — involving repetitive checks on subordinates’ work and the enforcement of standard operating procedures.” The primary value these workers provide is the ability to follow rules, to remain loyal to a company, and to work accurately and quickly. The general trend is that these jobs are being outsourced, displaced, and cost-reduced to the extent firms are able.
purchase Topiramate 2. In-Person Servicers deal directly with people. Like Routine Producers, in-person service workers may also complete standardized, repetitive tasks under supervision. However, work in this segment requires “a pleasant demeanor. They must smile and exude confidence and good cheer, even when they feel morose. They must be courteous and helpful, even to the most obnoxious of patrons. Above all, they must make others feel happy and at ease.” This description is relevant to an increasing number of job roles, and almost all of us will find that some percentage of our jobs (too much, I say!) falls within this category.
go here 3. Symbolic Analysts identify opportunities and “solve, identify, and broker problems by manipulating symbols.” Essentially, their job is to represent or model business realities in a way that’s consumable to management. They create insights, tools, and processes that allow the firm to capitalize on and monetize knowledge. Applied to general Analyst role in organizations, this is the role you need to cultivate if you want to “compete on Analytics.”
Many technology and knowledge workers have elements of all three classifications in our roles at work. However, while skills across these categories can compliment each other, it’s important to recognize that workers who specialize in one particular arena will perform best in their specialized role. Accordingly, once you know which of these roles you want your analyst to fill, you’ll be much more effective in planning for, staffing, and supporting the Analytics role — whatever function you want it to serve. To aid in identifying the kind of role that will best serve you in your organization (and at the risk of ticking more than a few people off), here are three prototypical examples of different reporting roles I’ve worked with in various organizations, along with some high-level costs and benefits:
Financial Reporting Analyst / Marketing Analyst (the “Reporting Manager”)
- Typical Responsibilities: Compiles data from known, consistent sources. Maintains existing reports and compiles new reports to specification *provided* those data are readily available. Highly dependent on existing processes or internal engineers to provide standardized and “cleansed” datasets. May spend a fair degree of time manually producing reports.
- Supervision Level: Typically high. Work may frequently involve financial or outside reporting that requires multiple levels of quality assurance and error-checking.
- Technology: Frequently dependent on an established internal reporting system, so this role is often seen in large to very large organizations that have a need to distribute large volumes of pre-summarized data to multiple internal audiences. Very frequently work in Excel, occasionally in Access. SQL skill level may vary from none to moderate, though application for SQL frequently comes down to pasting results into a spreadsheet.
- Responsiveness: High for updating existing reports, moderate for developing new reports based on standardized datasets.
- Organizational Support: Microsoft Office technology or similar. Dependent on engineering to provide data to known repositories in accessible formats, possibly in a centralized data warehouse, or on multiple parties to provide spreadsheet data for cut-and-paste operations.
- Ramp-up Time: Very fast. This position deals largely with established processes and matches output to known or expected outcomes.
- Cost: Low, possibly in the range of an associate Product Manager or beginning Marketer.
- Return on Investment: This role is often a purely supportive one, so return on the headcount is indirect.
- Outsourceability: Typically low; this work often involves secure data, proprietary reporting systems, and constant incremental changes to existing reports.
- When to staff this role: When you have very low data volumes, a larger organization with consistent and accessible datasets, a need to manually compile information from different sources into a single source for distribution, or a need to push out a large volume of standardized reports to multiple audiences.
- Qualifications to Look For: Look for applied experience working in broadly available tools; experience solely in another firm’s proprietary reporting system will not be immediately generalizable to your firm’s systems.
Business Intelligence Manager / Web Analytics Manager (the “Reporting Analyst)
- Typical Responsibilities: Compiles data from known, mostly consistent sources. Maintains some existing reports and compiles new reports to specification. Sets up, conducts, and analyzes custom A/B tests inside of existing frameworks (Google Analytics, etc.) Works to identify, source, and integrate data from within known data frameworks (internal databases, Google Analytics or SiteCatalyst, MailChimp, etc.) Dependent on the integrity of existing data sources if the organization does not provide a reporting “sandbox” database where the analyst can manipulate data themselves. Dependent on internal engineers to capture basic data, but able to standardize and “cleanse” those datasets themselves. Focuses on compiling reports into automated presentation frameworks such as Google Analytics or SiteCatalyst dashboards, SQL Server Reporting Services, or Crystal Reports. Aside from specialized applications, producing and maintaining reports in something like Excel may fall somewhat below this pay grade.
- Supervision Level: This is a largely autonomous role. May require moderate supervision when learning to use new datasets; after becoming proficient, this role will be the SME for those same datasets. Work output will typically see little to no quality assurance or error-checking aside from that which occurs when the reported information is consumed.
- Technology: Often dependent on one or several third-party frameworks such as Google Analytics or SiteCatalyst, so this role *may* encounter some difficulty in bridging datasets from across multiple platforms (depending on complimentary skillsets.) This role is seen in organizations of all sizes, particularly with the rise of tools such as Google Analytics (and, frequently, the separation of financial reporting into a separate function.) May do a small to moderate amount of pasting to Excel or similar to make report distribution easier. SQL skill level may vary from none to moderate; application for SQL will vary depending on the degree to which the organization requires internally-stored data in reports.
- Responsiveness: Typically this role sets up production reports to maintain themselves. High to very high responsiveness for developing new reports based within known frameworks (SiteCatalyst, etc.) depending on skill level. Developing (non-manual) reports across multiple sources may significantly challenge this role’s skillset.
- Organizational Support: Highly dependent on reliable site tagging if working in GA, SiteCatalyst, or similar. Mild to moderate ability in manipulating multiple large datasets with different structures will likely still require some support from development to import and/or standardize data for reporting and analysis.
- Ramp-up Time: Very fast. This position deals largely with established processes and matches output to known or expected outcomes. Very often will bring in best-practices that can have an immediate impact on business practices in the analyst’s skilled areas.
- Return on Investment: This role is largely supportive, though often at a higher level in the organization. Analysts will contribute new insights as experience accumulates, which can be packaged into business cases typically by another role. When a complimentary skill is present, the analyst will often use these skills together to contribute substantial value at the intersection (e.g., SEO.)
- Outsourceability: Typically high; expertise in third-party tools is often very transferable. The other side of the coin is that there is substantial job demand for this class of analyst.
- When to staff this role: When you’re heavily reliant on third-party tools such as Google Analytics or SiteCatalyst, want to reduce the load / increase the consistency for business managers pulling their own reports, and/or have a need to set up and manage third-party reporting dashboards or centralize these datasets into something like SQL Server Reporting Services or Excel.
- Qualifications to Look For: Look for significant applied experience in one or more third-party tools such as Google Analytics — preferably in the tool your firm is using, though these skills are moderately generalizable across tools. This role is largely business-trained, so look for a degree or a complimentary applied skill to add value to the role.
Modeling and Analysis Manager / Decision Sciences Manager (the “Symbolic Analyst”)
- Typical Responsibilities: Consults with internal clients to help identify key business drivers and to match reporting/analysis requirements to those business drivers. Maintains some fully-automated existing reports and compiles new reports to specification. Sets up, conducts, and analyzes custom A/B tests and provides statistical analysis both inside and outside of existing data frameworks. Works to identify, source, and integrate data from familiar and unfamiliar data sources of all kinds (internal databases, Google Analytics or SiteCatalyst, third-party demographic and economic forecast data, etc. etc.) in an effort to develop a fuller understanding of the business. Dependent on rich internal data sources to fully leverage the skillset; focuses in generating and storing as much data internally as possible. Often works to independently spec and compile a working data mart, provided adequate organizational support is available. Fully able to acquire, manipulate, and standardize datasets from multiple sources. Focuses on generating business insights and modeling tools resulting from key drivers analysis, predictive modeling, and segmentation with correspondingly lesser focus on reporting. Depending on complimentary skillset, may also conduct pricing optimization studies.
- Supervision Level: Low, but requires interaction with database engineers and application developers to gain proficiency. Some work output is advanced enough that only other symbolic analysts can validate it, which may not fit well in all organizations. Due to the broad range of tasks typically available, this role needs special organizational support in managing priorities.
- Technology: Highly dependent on specialized software: typically SQL Server, Oracle, and/or SAS or SPSS (recently relabeled PASW), to enable this role to compile and manipulate data from multiple disparate sources. (Unfortunately, at this time MySQL offers very limited data warehousing capabilities without the aid of an additional dedicated resource to write and manage copy jobs, though no doubt there are analysts out there with this range… just be aware.) This role can be seen in smaller (30+) organizations seeking to drive greater insight from a significant volume of data or in mid-size companies seeking to establish a competitive position through analytics. Large organizations (such as CitiBank) may have dozens of these analysts. This role will possess a significant skill in at least one data-manipulation language, typically SAS; less often SQL or a development language, and, increasingly, R.
- Responsiveness: This role is likely to be *less* proficient in third-party tools such as Google Analytics because the specialized education typical of candidates often comes from study in the social sciences rather than from hands-on business training. However, once proficient, this role can fully support this demand as well as add substantial value by tying it to other data sources. This role will typically focus on delivering the information required to make a decision rather than delivering reports per se.
- Organizational Support: Dependent on basic data availability. Require development support to set up data retrieval from some third-party sources such as Google Analytics, or, depending on complimentary skillsets, may be capable of fully integrating these datasets without aid provided the software tools are made available (SAS, SQL, and/or development tools.) Also dependent on organizational priority to keep a good percentage of work requests at the strategic level so that the firm doesn’t employ this role in a purely reporting capacity.
- Ramp-up Time: Slow. Able to produce basic reports and insights from unstructured database data within weeks; proficient across the business spectrum within one year; able to produce very significant insights, predictive models, and proposals for business rule changes within six months to a year. This position deals with both well-known and unstructured datasets, so proficiency in some areas comes faster than in others.
- Cost: High to very high, depending on skill range and relevant industry experience. Though often working as individual contributors, this role often demands Group Manager or Director-level salary in exchange for highly specialized skills. Additionally, in order to get the most value out of this role, the organization will need to commit to moderate hardware and software investments.
- Return on Investment: Theoretically, high. This role is experienced in driving revenue through improving conversion, identifying key drivers, proposing business rule changes, and implementing predictive models (see an example here). This role is less likely to bring in a complimentary business skill (such as SEO) that can be immediately applied to the business.
- Outsourceability: Some skills such as predictive modeling and expertise in experimental design, are easily outsourceable; however, outsourcing one or two such projects can easily exceed the cost of this role. Other skills, such as knowledgeable consulting for sourcing data and combining disparate datasets for ad-hoc analysis, are not reliably outsourceable.
- When to staff this role: This is a good role to hire if you’re ready to build out a competitive position in Analytics; if you find that the firm needs or is outsourcing significant amounts of segmentation, predictive modeling, or statistical testing work; if you have multiple disparate datasets that the organization is having difficulty consolidating. Alternatively, this role can add significant value if you’re growing an existing team of analysts of other stripes.
- Qualifications to Look For: Often this skillset is acquired via Masters or PhD work in psychology, sociology, econometrics, or statistics. Occasionally you’ll find a business-trained statistical modeler with a different pedigree. This role should require knowledge of experimental design and applied statistical analysis, as well as proficiency in a statistical tool such as SAS, SPSS, or R, and depending on your environment, may require knowledge of SQL.
Overall, each of these three “example” Analyst roles can bring significant value to the firm — *provided* you hire the right skillset for the objectives you’ve set out for the role. Applying Robert Reich’s observations about the types of general job roles was helpful for me in understanding the kinds of general Analyst roles that are out there as well. On that note, I’d like to share a closing observation about the type of work symbolic analysts do and whether this role will be a good fit in your organization:
If you’re ready to “compete on analytics” — ready to dive in and hire a symbolic analyst — three things will determine the success of your endeavor: 1) this role needs a lot of detailed data to crunch on, arguably in a high-volume business; 2) this role requires specific tools and you need to be ready to commit to them; and 3) your organization must be strongly data-driven. At the end of the day, the primary job of a “reporting manager” is (usually) to help you get the information you communicate that you want; the primary job of a “symbolic analyst” is to conduct research into your business environment and to develop insights and tools to drive revenue — derived from objective, empirical analysis of the firm’s business data. To the extent the firm’s business objectives aren’t measured, or to the extent management/business objectives are opaque, unstated, or hidden, the symbolic analyst’s work may appear as out-of-touch or in conflict with the objectives of other groups. That’s no recipe for success. I hope you’ll keep it in mind.