From report to insight: increasing analytics maturity
Today’s companies face a gigantic challenge: how to extract value from information and operate on a fast Data-Driven environment. Good insights have long played a key role on business and executives know it. But, never before has it been the source of companywide frustration. Simply put, there is a difference between wanting to be Data-Driven and get it done. The reason, a massive infrastructure and skill gap between data and analytics.
An organization business is not to write reports but to act based on insights
Today’s analytics solutions are often tested and presented on simple and hypothetical scenarios which fail to reflect day-to-day trade and the nuances of real organizations. It falsely prompts business leaders to believe that it is as easy as to plug a data set into a tool and it will generate immediate business impact. It isn’t, data requires a lot of refinement before any visible result.
There are three key elements to successfully implement a data-driven culture:
- Leaders need to make a leap towards a new view of data and analytics. To move past the old mindset of keeping siloed data, business applications and analytics applications grounded in data warehousing and business intelligence systems, primarily for enterprise reporting;
- Teams need to start treating data as the raw material of decisions and planning;
- And Companies need to embrace new methods and tools which will guide them on the path to analytics maturity.
Towards Insightful decisions
One of such tools is the Analytics Continuum. Introduced by Gartner, it represents the different stages towards an organization which actively acts on data. It is a helpful tool for organizations understand their current situation and plan for the road ahead.
Based on the experience of helping numerous companies, on a variety of industries we at Math Marketing present the Analytics Continuum as follows:
Descriptive Analytics: allows companies to determine what happened during previous periods. It helps managers measure performance and key metrics and is often translated into reports and dashboards. It focuses on past performance and triggers reaction when a deviation is detected.
Unfortunately there is a trend to either report the bare minimum required by regulations and compliance (e.g. Quarterly Financial Reports and Tax Reports) or generate an overwhelming amount of reports, which require significant effort, are seldom timely relevant and are prone to inaction. Nonetheless, a good Descriptive Analytics is an invaluable foundation to build upon.
Explanatory Analytics: by focusing on the why, explanatory analytics aims at understanding reasons behind results reported by Descriptive Analytics. It often requires the study of correlations and regressions among variables. However it still provides only a backward vision. It is about understanding relationships and why certain things happen and others don’t. Good Explanatory models might turn out to be good Predictive models but the starting goal is different.
Every stage of the continuum imply a significant increase of skills and difficulty. But, none is so dramatic as the shift from Descriptive into Explanatory Analytics. Regulations, off shelves ratios and business acumen can produce acceptable, even if inconsequential, reports. But explanations require a minimum statistical knowledge. Corporations graveyard is full of rushed conclusions and anecdotal correlations.
Exploratory Analytics: enables organizations to explore the causes behind the events and explanations detected. It implies a broaden look at patterns, trends, outliers, unexpected results and makes use of a mix of quantitative and visual methods to get the sense of what the story data wants to tell. Causation is a step further than correlation. It says any change in the value of one variable will cause a change in the value of other variable. Understanding the cause is a stepping stone into prediction.
Predictive Analytics: is about anticipating what could happen, estimating impacts and forecasting the future. It allows organizations to create and explore different scenarios and plan ahead. It requires a deep knowledge of the environment and a great level of control over data. Although the tool set plays a role here, the main part is on advance statistics/econometrics and business knowledge.
Prescriptive Analytics: given the deep knowledge already gathered, it provides the organization with a path towards reaching a specific goal. It basically answers the question, “what should be done to achieve result X”. At this stage, human intervention can be at its minimum and is possible to delegate decisions on machines to achieve business objectives.(e.g.: hedge funds trading algorithms)
In the context of a Marketing department of a financial services company, Descriptive Analytics provides information on last month’s engagement of a particular campaign. Explanatory Analytics looks into the relation between engagement and marketing expenditure. Exploratory Analytics explores audience cluster to discover possible trends. Predictive Analytics concludes about future campaign engagements and Predictive Analytics informs management about the best campaign to achieve the desired engagement goal.
How deep is your data
One aspect often overlooked is the role different data sources play. Value can be generated from expanding data sources even without increasing the analytics maturity. In some cases, a Descriptive Analytics with a vast array of data sources can be as instrumental as a predictive analysis based in internal information. We call this Analytics Depth and consider to be on one of the following categories:
Internal structured data: Information collected by companies during their normal transactions like sales value and amount, customer contact info or product logs. They are the result of intentional information gathering processes. It is the easiest and sometimes the only source of information available to companies.
External structured data: information from external database, either publicly available or from a data provider. It includes the likes of government statistics, customers reports or market researches. Information is often available on a tabular form or any other structured and is up to organizations to find correlations with Internal data. It increases the volume of data available for companies.
Internal unstructured data: not all internal information is readily available for consumption. Business knowledge is often disseminated among informal tools like e-mails, chat rooms, forums or wikis. Some might even not be available on any digital form like contracts or product manuals. Incorporate unstructured data sources is a complex endeavour. Increases the variety of data available for companies
External unstructured data: includes information from social media, community blogs and other online forums where the company interact with customers or stakeholders. Given the share volume and velocity of data being created nowadays, it is a challenge to incorporate into the decision process.
What Analytics Depth helps organizations to understand is that the road to analytics maturity is more on process evaluation, asking the right questions and on diversifying data sources and less on investing on new tools.
Insights require new set of skills
The core reason why organizations spend so much time and effort on analytics shouldn’t be to just have access to the latest reports. While knowing what just happened is important and useful, it only provides details into how well things performed and gives and overview of financial performance in the paste. However, it often fails to provide any insight into what should be done tomorrow.
A large number of companies spend too much time and money on descriptive analytics and tools. Unfortunately, and much to management frustration, this tools require more than they expected before they produce insights at their full potential. Skills remain the biggest challenge for organizations. These additional capabilities often require new roles with additional quantitative, content and analytical skills, perhaps statistical analysts or data scientists.
Most organizations fail to make the kind of cultural or business model adjustments to really leverage information. It’s great to come up with ideas for how to use information or analytics. But, if the organization is not prepared to actually act on it, then really it might turn out to be a wate of time. Lots of insights but not actually linking them to business
Final thoughts
Organisations make hundreds if not thousands of decisions every day at different levels, with different degrees of impact, complexity, frequency and use. The only thing that’s constant is change, and analytics represent a way of adapting to change. Identifying patterns, anticipating outcomes and responding proactively will be the basis of competition in the future.
Math Marketing has helped companies from a variety of industries to implement Data Driven processes. By mixing statistical proficiency, business knowledge and advance tools we help our customer answer though questions. Some of the largest brands share with us with their data challenges, and trust our data generated insights on everyday operational and strategic decisions.
If you want to learn about our projects and methodology reach us over the e-mail or contact any of our team managers directly on Linkedin. We are always available to share our insights with you and listen to your experience.
