The challenge
To harness the advantages of Information Technology (IT), big amounts of investment are not required: the benefits won’t be larger for the company that invested millions than for the one that optimized its available technology or seized the most affordable option. The key relies on IT’s capacity to support a business strategy, while enabling practices and relevant decisions.
With this purpose, in the last decades, organizations have invested greatly in two types of business systems:
- Transactional systems or ERP (Enterprise Resource Planning): They integrate and control information regarding different functions inside an organization (Finance/Accounting, Human Resources, Sales, Production, Buying, etc.) These systems have some extensions, like CRM (Customer Relationship Management) modules.
- Business Intelligence systems (BI): They consolidate the information form transactional systems and other sources (in the shape of reports, dashboards, scorecards, among others) to support management and decision-making.
Nevertheless, they both share a weakness: they focus on historical information, which explains what happened in the past. With the advances in mathematical models and technological capacity in recent years, a new group of tools dedicated to predicting the future and suggesting courses of action has risen. This new group of tools is called BAO (Business Analytics and Optimization).
BAO
Predicting the future of the business doesn’t just make room for a company to become market leader, it also gives its members the opportunity to develop skills that other players don’t have or that will take longer for them to master. Impossible to achieve? Not for BAO it isn’t. BAO tools focus on the future; predict all possible outcomes; and suggest courses of action, through data analysis and mathematical models.
BAO allows companies to manage higher complexity and data volume; it translates them into concrete, valuable and integral information for decision-making; and optimizes operative, commercial and organizational strategies. This way, IT stops being just another corporate department; it turns into an enabler that generates aggregated value for the company.
Benefits
The companies that implement BAO to their strategy will be able to benefit commercially, operationally and organizationally, based on its capacity to support the business strategy while enabling practices and decisions relevant to the strategy.
Some examples of benefits with BAO are:
- Commercial
- Increase in sales through basket analysis, price optimization and buying suggestions.
- Effective and profitable promotions.
- Client segmentation and strategies according to each segment.
- Retention of good clients.
- Operations
- Alignment of production, distribution and stocking activities in the supply chain.
- Production resources optimization.
- Logistic spending reduction through network analysis.
- Working capital reduction through optimization of finished products inventory, raw material and production in process.
- Organizational
- Employee recruitment and retention.
- Prediction of employee performance, based on their profile.
How do BAO tools work?
To ensure BAO tools’ accuracy, we use advanced mathematical models of predictive analysis, data mining, forecasts, simulation and optimization, among others.
BI (Business Intelligence) tools, the most popular method among companies nowadays, focus on an area known as “descriptive analysis”. The objective of such analysis is understanding what happened in the past, through answering questions like: What happened? And, why did it happen?
But just like business scenarios change, an evolution in the processes with which companies compete is required. Hence, BAO tool focus on two types of analysis: predictive and prescriptive.
Predictive analysis answer questions such as:
- What are the causal variables that impact a process?
- Which clients answer better to this type of promotion?
- What clients are more prone to stop paying their debt?
- Which employees could quit?
- Which people are more prone to get sick?
- Which criminals are more likely to relapse? Where?
Prescriptive analysis answers questions such as:
- What are the variable combinations to maximize success probability as an optimal result?
- What demand should I cover when facing a capacity restriction?
- How can transport resources be optimized, in order to reduce distribution spending without impacting on the service level?
- How can productive and human resources be better assigned to satisfy demand?
- In which location should a new plant or distribution center be opened? With what capacity?
BAO’s focus
Because of its nature, current BI tools are solely used on an executive level. Besides, they don’t allow immediate decision-making, since it requires information consolidation processes, executed on defined schedules.
With the objective of improving decision-making processes and permeating this competency throughout the organization, BAO tools enable decisions on all levels (strategic, tactical and operational), to then be incorporated in the daily operations of the company. For example:
- When searching/buying a movie on an Internet portal, the site recommends other films that people with a similar profile have bought.
- When soliciting a bank loan, the applicant must give personal information (age, marital status, occupation, income level, etc.). When registering this information in a system, it shows the solicitor’s breach chances.
- For organizations that have expensive machinery, any unplanned stopping represents a loss of income and profitability. When analyzing its information (temperature, pressure…) in real time, probability of failure can be predicted.
Big Data: complement, not component
Recently, one of the most used terms when talking about Information Technology is “Big Data”, and it is usually associated with BAO tools. Even though, it is not the same thing.
The term Big Data refers to the compilation, storage and handling of a massive amount of data that comply with the 3 Vs: volume, velocity and variety.
- High Volume: Every day 2.5 X 1018 bytes of data are generated, which means 90% of all data in the world has been created in the last 2 years.
- High Velocity: To serve some purposes, there is a real need to capture, analyze and take decisions in real time.
- High Variety: A company’s transactional data have a defined structure; but, there are many others without structuring: text messages, email, blogs, social media, multimedia, and data generated by machines and/or sensors.
It is impossible to run a Big Data analysis without BAO tools; but, BAO tools can be used without Big Data, meaning, they can be applied to analyze data in any repository: from an Excel sheet to a corporate data base.
First steps: few investment and low risk
The first few steps to incorporate BAO into an organization are rather simple, since they don’t involve big investments of time nor money, besides they implicate low risk.
There are 3 basic elements to get started:
- A hypothesis.
For example: Which current clients have a bigger probability of cancelling their subscription?
- Data to support the hypothesis.
For example: Subscriber profile (including subscribers that have cancelled their service in the past), payment deadlines, and recent activities, among others.
- BAO tools.
Even though there are expensive tools in the market, there are also free options (Open Source) or others at a very low cost.
In general, once the hypothesis has been defined and data has been obtained, a concept trail can be run in 3 to 6 weeks. If the hypothesis is valid, it can be tried with a higher volume and then take it into the operation.
Case Study
A consumer products company offers sales promotions to 1 million potential clients, and their response rate (clients that buy the product) is 1%, meaning 10,000 people.
If the buying margin is $200 and the cost for sending the promotion is $1 per client, the promotion’s profitability is:
Profitability = Margin – Cost
= ($200 X 1% X 10,000) – ($1 X 1,000,000)
= $2,000,000 – $1,000,000
= $1,000,000
As a result from a predictive analysis exercise, it is detected that one quarter of the potential clients are more prone to buy, with a response rate of 3%.
When sending the promotion just to the group of clients with the highest response rate, profitability is:
Profitability = Margin – Cost
= ($200 X 250,000 X 3%) – ($1 X 250,000)
= $1,500,000 – $250,000
= $1,250,000
With the new promotion, profitability rises 25%, with just a quarter of the previous promotional cost.
Companies that have applied this type of analysis observe better response rates per client segment, which has translated in promotional cost reduction and an increase of revenue and profitability.
Summary
The objective of Information Technology is enabling processes to support a business strategy. To achieve this, organizations have invested greatly in business systems. But, current systems are limited to explaining the past. Recent advances in mathematical models and technological capacity give room to a new group of analysis and optimization tools (BAO), which predicts the future and suggests courses of action.
BAO tools can be useful for different organizational areas (Finance, Sales, Operations, Human Resources, etc.) they also support decisions on different levels (strategic, tactical and operational). Organizations that have adopted this kind of tools obtain tangible benefits in the shape of an increase in sales, spending reductions, asset use, and client & personnel retention, among others. It just takes a hypothesis and the necessary data to prove it.