“Quality information is increasingly recognised as the most valuable asset of the firm. Firms are grappling with how to capitalise on information and knowledge. Companies are striving, more often silently, to remedy business impacts rooted in poor quality information and knowledge.” - Kuan-Tsai Huang, Yang W. Lee and Richard Y. Wang, Quality Information and Knowledge
The concept of Intelligent Quality (IQ) is a new idea that reflects the end state of the well-known Data, Information, Knowledge, Wisdom hierarchy. It goes beyond how well the data fits the system specifications - it concentrates on how well the data provides for the business information needs of the organisation.
By looking at data from a business point of view, applying business rules and acknowledging the impact of business processes on the quality of data, information and knowledge, data IQ can be achieved. Intelligent Quality will form the information foundation which can then be analysed and interpreted and on which you can add business experience and external intelligence to form wisdom.
It must be stated that IQ is not an IT activity. It may start in IT, but it must include the business in order to be successful. All business areas that create, utilise or report on business information need to be involved. Subject matter experts need to be found and engaged in the programme. This is very important.
Even more important is finding the right sponsorship. You may be able to get funding for an IQ initiative without a sponsor. However, you will struggle to get everything implemented and to build a long-term environment without one.
So, who would be the right sponsor? It must be a senior manager. Improvement processes will require changes to business processes and systems. They will also require financial, system and human resources. In fact, most initiatives will have an impact on job roles and may even require changes to entire departments. Senior managers alone can enable these to happen.
To find the right sponsor, you need to find a senior manager who is in pain! Unless you have at least one person, at a very high level in the company, terribly worried due to a problem associated with data, you won’t get the right backing to ensure success. The worry may be about missed sales, runaway costs or may be due to an awkward mistake that has come to light in a very embarrassing way.
Even if your senior sponsor is really worried, you will need to highlight the connection between his or her pain and poor quality information. If you can relate the pain back to a data quality problem, you will have the keys to getting an Intelligent Quality initiative underway and valuable input into that all-important business case.
Three things you’ll need to gain buy-in
To achieve long-term buy-in from the myriad people you’ll need to attain success, you’ll need three things – a vision, the visualisation of that vision and the roadmap that will get you there.
Vision - This should focus on the benefits that will be gained by realising wisdom in the organisation and should also highlight the relief of whatever pain your sponsor and the business are experiencing. This is why it is so important to find and measure the pain.
Visualisation - you need to help people visualise the vision and to do that, you need to understand it fully. Review the situation and walk your way through to your solution. Draw yourself a picture of this solution - literally. Based on the problems you’ve discovered and the pain they are experiencing, ask yourself the following questions:
•What systems have problems?
•What business areas are experiencing issues?
•How is it impacting the customer?
•After the improvements you will undertake, what will be different?
•What will the business area be able to do to capitalise on the improvements?
Graphically depict the answers to these questions. Keep your visualisation simple so people can hold onto it and take it away with them. Once this is complete, review it with several people in the business to ensure your ideas are as clear to them as they are to you. This vision will form part of a “roadshow” pack that will form the basis of your campaign to get your Intelligent Quality initiative off the ground once you get the go-ahead.
Roadmap - If you can’t show people that you can get them to that vision, you will not get the backing you need. A roadmap, including milestones with deliverables, is required to engender confidence in you ability to accomplish your improvement goals.
It is here that you must make it crystal clear that this is an improvement programme, not a quick-fix solution. It takes time to get everything under control and will need a continuing effort to ensure the quality remains for the long term. However, good planning will reap rewards in short order. Prioritise the milestones so that the biggest problems get relief first. Include easy wins along with big problems so you can prove return on investment quickly.
“Quality is Free”
“It’s not a gift, but it is free. What costs money are the unquality things – all the actions that involve not doing jobs right the first time. Quality is not only free, it is an honest-to-everything profit maker. Every penny you don’t spend on doing things wrong, over, or instead, becomes half a penny right on the bottom line.” - Philip Crosby, Quality is Free
Companies have created whole departments dedicated to correcting poorly-captured data. In fact, these departments have become part of the very fabric of most organisations and require and receive substantial funds to make this happen. If quality principles were employed at all points where data is created, altered or utilised for business purposes, the need for the scrap and re-work undertaken by these departments would be greatly reduced, saving large amounts of money.
Of course, it costs money to set up a quality improvement programme. However, the returns are far greater than any costs. I have yet to find an organisation that didn’t get more than a 200 per cent return on investment from Intelligent Quality that is focused on enabling the business with high-quality information that will help it to become wise.
Where it is relatively easy to quantify scrap and re-work costs, it is sometimes difficult to get a company to admit that back office processes are not necessary parts of their business. These operations have become ingrained in the fabric of companies and do not disappear easily. However, improving data capture and usage processes will have the desired effect, and the people working in those back offices can be re-deployed elsewhere in the organisation doing work that earns money, rather than wasting it.
Implementing an Intelligent Quality programme
Once the business case has been approved, you are ready to undertake your first improvement project. A clearly defined methodology will go a long way to creating confidence in your abilities. The key starting point of any data IQ programme is what may be called a Forensic Diagnostic. This is a multi-faceted approach (People, Process, Technology) to finding and analysing data quality problems and then identifying the improvements required.
Investigate - Start by investigating the extent of the data and data quality problems. You would have started this process when putting together the business case, but this now goes further and deeper. First, profile the data again. This time look further and check relationships across tables and duplicates within tables. This is where profiling tools can be a big help.
Review data definitions - Take a review of all the documented data definitions and compare them against the profiling results. It is possible that the definitions have not been implemented according to specification. Is this because the specification was wrong and the data never could measure up? Where definitions do not align with the actual data, discuss this disparity with the subject matter experts. It could be that the definitions do not reflect business needs. In this case, the definition and associated metadata needs to be changed to reflect this.
Assess data architecture - It is important to have a good analysis of the data architecture involved with the key data you are improving. Check the various relationships between entities and attributes to make sure the relationships are identified correctly and look closely at the lifecycle of the data to make sure it meets business needs. Look to the results of the data profile as this may provide some clues to data architecture problems, as may the data definition review.
Questionnaires and interviews - Before conducting structured workshops, develop questionnaires and distribute these to workshop attendees and other individuals around the business who may be able to provide valuable insight into data and data quality problems. Questions should centre on data usage, data and information issues, business information needs, costs of poor quality, information customers/suppliers, and business goals. In some cases, you may want to conduct structured interviews with individuals who may not be able to attend workshops or who appear to be so key to the investigation that a one-to-one discuss is warranted.
Structured workshops - Workshops with subject matter experts and knowledge workers should be themed around specific areas of key data and should include all the individuals who have a stake in the quality of the data being investigated.
Three distinct types of workshops should be employed, each with a specific purpose:
The first is an investigation workshop, aimed at reviewing data issues identified when gathering evidence for the business case and also data concerns discovered during the data profiling and data definition reviews. Walk through the data profile results and discuss the data definitions with the business people involved to identify any further issues and to start to find the reasons for any problems that may exist.
The second is an information chain workshop. The purpose of this workshop is to map the information flows of key data through the business. Help the attendees to document the customer/supplier relationships within their organisation with regard to their key data.
The third is a root cause workshop. Here you review the problems that have been found and work with the business to identify the root causes of the problems and start to find ways of making improvements that make sense to the business. Depending on the extent of the investigation you have underway, you may want to combine the investigation and information chain workshops together.
Develop recommendations - Improvements need to be based on their practicality and effectiveness with a close eye on cost and effort required to implement. Be sure to consider all possible remedies. These should include business process improvements as well as system and architecture changes. Data cleansing could be needed to clean the existing data but you should combine this with a more permanent repair to the underlying issues. This will prevent a recurrence of the quality issues. Make sure the recommendations set appropriate goals and include quick wins along with more meaty work streams. Continually stress that this is a programme dedicated to continual improvements, not big bang solutions.
Making improvements - The final step in the process is implementing the recommendations to start the continuing improvement cycle. This will involve making improvements to business processes, system changes, changes to the data definitions and architecture, data cleansing, measures for success, communications strategies and more.