Analytics is an overused term. This webinar shows how BI, web analytics, data mining and predictive analytics all have a role but all need a focus on decisions - especially operational decisions - to maximize their value. Webinar recording available here: http://decisionmanagement.omnovia.com/archives/64147
3. The world of analytics Challenges in Business Intelligence Limitations of Web Analytics Problems with Advanced Analytics The Decision Management Solution Wrap Up
Analytics means putting all your data, not just your web data, to work improving the way your web site works. It means using your historical data not for reporting but for predicting and driving new and more effective behavior across the company. It means putting all your data to work improving decision-making.This webinar will show you the range of meanings of analytics, contrast some of the common meanings and show how Decision Management and a focus on operational decisions can focus your analytic efforts for maximum value.
Fixed reports assemble a repeatable set of data to present a set of information.Fixed reports are typically generated and delivered on a regular schedule – daily, weekly or monthly for instance.Although a report, even a fixed report, could inform the reader, many are poorly designed and simply deliver large volumes of data with little or no information. For instance, a report that simply lists all the purchases made at a store on a given day would be full of data but of little or no value to anyone.Most fixed reports are either required for compliance purposes or they are a poor solution to a business problem. If, for instance, our problem is deciding which products are selling well enough to justify an additional supply and which are selling poorly enough to be discounted for clearance then a report of sales will enable us to answer the question. Trawling through all the data in a sales report, however, is a very poor solution as it requires a store manager to spend a long time reviewing the data and information in the report to derive the relevant (and useful) insight buried within it.
Dashboards are a collection of reports, visualizations and other elements designed to give rapid access to critical information.They are often compared to the dashboard of an automobile or airplane.In general a dashboard is better compared to the instrument cluster of an automobile as it only reports on status, it does not allow changes to be made.Dashboards generally use a graphical metaphor that will be familiar to a user – such as a process flow or map – to organize the information displayed.Key Performance Indicators or KPIs and other measures are common topics of dashboards and Enterprise Performance Management is a term often used to describe the use of dashboards in this way.
OLAP (Online Analytical Processing)is designed to rapidly answer complex queries against dataBefore OLAP tools can be used to analyze data a multi-dimensional data model is specified by a designer. These dimensions – date, location, value for instance – can be used in the analysisThis essentially “de-normalizes” the data in a way that allows certain kinds of analysis to be conducted very rapidly. For instance, the way sales vary by date and by location can rapidly be seen without having to run slow database queries.In addition many summary results and possible aggregations are pre-calculated and stored for faster access. For instance, date dimensions are often pre-aggregated into weeks, months, years, fiscal reporting periods etc. so that these roll-ups can be examined easily.OLAP tools allow managers and knowledge workers to navigate through data and pick out information
Visualizations help you see information content that would otherwise stay hidden in data. Unlike a report, visualizations do not rely on rows, columns and numbers to display information.Visualizations use graphic design and visual clues to bring out the meaning – the information – within the data.Visualizations rely on the human ability to rapidly see and understand patterns to help a user understand the information content of data.More and more new visualization techniques are being developed and productized all the time and what follows are just some examples.For more information about this topic, see the section on Visualization in the Bibliography provided in the Attachments tab.
Descriptive analytics can be used to categorize customers into different categories – to find the relationships between customers - which can be useful in setting strategies and targeting treatment. But this analysis must be delivered not just to your analysts, also to your systems. Analysis is generally done offline, but the results can be used in automated decisions – such as offering a given product to a specific customer – often by developing rules that embody the analytics.For instance a decision tree can be created where each branch, each end node, identifies the segment for a particular member.Data mining can also create rules with less effort and with a quicker time to market in certain circumstances
Predictive analytics often rank-order individuals. For example, rank-order members by their likelihood of renewing – the higher the score, the more “completers” for every “non-completer”. The risk or opportunity is assessed in the context of a single customer or transaction and these models are not an overall pattern, even if they are predictive. Models are called by a business rules engine to “score” an individual or transaction, often in real time, though the analysis is done offline.These models are often represented by a scorecard where each characteristic of a member adds to the score and where the total score can then be returned.
Models make predictions but predictions alone will not help much – you must ACT based on those predictions.When you are thinking about smarter systems, taking action means having the system take action in a way that uses the predictions you made. You need to make a decision based on those predictions and this means combining the models with rules about how and when to act.Let’s take our retention example from earlier. Knowing that a customer is a retention risk is interesting, acting appropriately and in time to prevent them leaving is usefulGrovel index story
Time is accuracy, accuracy is moneyThe delays caused by manual creation of models, calculated attributes, mapping to production systems, testing in production and all the other tasks involved in getting a model into production degrade the effectiveness, the accuracy of the model. Less accuracy means less money.It’s important to think about how long a model will take to get into production and consider that when building it – if changing the model makes it a little more accurate but much harder to implement, is that going to be worthwhile?Again, let’s consider our retention example. An analyst may find that including up to the minute call data makes the prediction of retention risk much more precise. But if it takes a week to implement a model that includes that data and only a day to implement one that does not the operational effect may be small or even negative.
Remember – decisions are where the business, analytics and IT all come together
Once deployed analytics cannot be a “black box”, we must understand analytic performanceObviously you need a 'hold out sample' or business as usual random group to compare to.You need to understand what's working and what's the next challenge – which segments are being retained, for instanceYou must understand operational negation.You need to track input variables, scores, decisions or actions taken (classic example is in collections where a strategy may dictate a 'do nothing' strategy, but the collections manager overrides the decision and puts the accounts into a calling queue) and operational data that fed the decisionBoth analysts and business users must think about what they can do to improve decision making, which is the foundation of adaptive controlIn our retention example I need to have some customers I don’t attempt to retain or that I don’t spend any money retaining. I have to capture what the call center representative ACTUALLY offered and what was actually accepted (if anything), not just what SHOULD have been offered and I have to be able to show the results to my business users in terms they understand.
At its heart a decision is a choice, a selection of a course of action. A decision is arrived at after consideration and it ends uncertainty or dispute about something.Decisions are made only after considering various facts or pieces of information about the situation and participants.Decisions select from alternatives, typically to find the one most profitable or appropriate for an organization.Decisions result in an action being taken, not just knowledge being added to what’s knownThe basic decision making process is simple. Data is gathered on which to base the decision. Some analysis of this data is performed and rules derived from company policy, regulations, best practices and experience is applied. A course of action, a selection from the possible options, is then made so that it can be acted on. When considering decisions in operational business processes, the way the decision is made is often constrained such that it can be described and automated effectively in many, even most, cases.
How many decisions are involved in sending a letter to a some customers?One view says a couple of decisionsWhat to put in the letter and who receives itA more complete view says that you have also made a decision for each customer to receive or not receive the letter. If you sent a letter to 10,000 customers, you just made 10,000 micro decisionsAdding a new option to your IVR system means deciding that everyone who calls will hear the option. changing your website means deciding that every visitor will see something new…Many strategic decisions can only be implemented if many supporting micro decisions are also made.
All these pieces contribute to ever-more sophisticated decision services that support your business processes.Decision Services externalize and manage the decisions production processes and systems needBusiness rules allow business users to collaborate in the declarative definition of decisionsAnalytics can create better more data-driven business rulesAnd ultimately additional predictive analyticsAdaptive control allows test and learn to become part of a continuous improvement loop
It is often helpful to walk through one example here. Let’s take some interaction with a customer – say making a retention offer – and see how it might work.Initially we have different channels and our approach to retention is probably different in each. The first step, then, is to take control of the decision so we can make it consistently across channels. We should also use rules to describe it so that the decision can be automated correctly and managed by business staff, not IT. However not all customers are the same so we should analyze them and segment them so we can retain them differently depending on what is going to work. Segmenting based only on the data we have is interesting but it would be more useful if we could also use predictions as to their risk of leaving, lifetime value of them etc as part of our decision. Back to the data, then, to build predictive insights. Applying adaptive control to continually improve the outcomes and we end up with an optimized decision.As we work our way through the class we will revisit this and discuss.
Begin!Identify your decisionsHidden decisions, transactional decisions, customer decisionsDecisions buried in complex processesDecisions that are the difference between two processesConsiderWho takes them nowWhat drives changes in themAssess Change ReadinessConsider Organizational changeAdopt decisioning technologyAdopt business rules approach and technologyInvestigate data mining and predictive analyticsThink about adaptive control
Decision Management Solutions can help youFind the right decisions to apply business rules, analyticsImplement a decision management blueprintDefine a strategy for business rule or analytic adoptionYou are welcome to email me directly, james at decision management solutions.com or you can go to decision management solutions.com / learn more. There you’ll find links to contact me, check out the blog and find more resources for learning about Decision Management.