The Questions Our Clients Ask Us
What kind of services does Data Stats provide?Data Stats' goal is to help you enhance your product/productivity so that you can channel more proceeds from your business into increasing your profits or cash flow, lower your production costs or improve your quality control.
Your goals are met by working closely with your staff in designing experiments to statistically analyze the variance in factors that are unique to your company. After evaluating the results we try to present you with a variety of options to help you meet your objectives.
What is an Analysis of Variance (AOV)?
An analysis of variance also known as multivariate testing is a mathematical process for systematically evaluating experimental results. It is used for simultaneously identifying relationships and their significances between many factors and observed results.
A process is a collection of various factors (ingredients, concentrations, rates of addition, temperature, etc.) that affects the total outcome. Multivariate testing is a unique approach which helps indentify the factors that have the most beneficial or the most negative effects on your product.
After consulting with your staff, an AOV is designed with a focus on your objectives and takes into consideration factors unique to your operation. The results allow us to develop "What If" scenarios that gives you options and flexibility in choosing how to meet your goals.
How is an AOV different from A/B testing?Your research & development or marketing team is already probably conducting A/B testing. It is the "classical" form of experimental design.
You divide your experiment into two groups. One group represents your original version and becomes your control. The second group is your alternative in which one and only one change is made. Once the results are measured and compared, the direction of the next A/B split is determined. The procedure is repeated until you are satisfied that all the factors have been examined.
A/B testing is the simplest and easiest method of comparing various factors. It is not complicated. However there are serious drawbacks to this method.
First, it is very time consuming. You can only accurately test one element at a time. To do otherwise eliminates your ability to track the results of a change.
Secondly, A/B testing does not measure potential interactions between factors. You cannot ignore that most situations are the result of many variables working together. Sometimes these variables do not affect each other. Sometimes they do and in unexpected ways. At times the interactions of two variables will have a larger effect than the sum of the individual variations. You have no way of knowing this unless you measure for it.
An analysis of variance also measures one factor at a time but in a group environment that increases the pace of results while measuring any interactions that may exist between factors.
What kind of results can I expect?The certainty of results can never be guaranteed. To do so would only be deceitful. At the same time it can be said with almost complete confidence that there never has been a product or process that couldn't be improved. The system that you design today can be further improved upon tomorrow.
It is very likely that your productivity, product or system can be optimized further.The chances are that you will get real improvements in your profits. The rewards can be very significant for your company and continue to payoff in the future. For example, say you manufacture 1.5 million pounds per year of a specialty chemical that you sell for US$0.50/lb. That works out to a revenue source of US$750,000.
Now say that you improve your productivity by 10%. You now manufacture 1.65 million pounds per year and do so without increasing your raw materials costs. Your product immediately generates revenue of US$825,000 or US$75,000 more than you did previously. Since this is the result of productivity improvement, the additional US$75,000 is pure profit. You can use this additional cash to reduce your borrowing, purchase new equipment, hire new employees, pay your company's increased benefit costs or you can retain these earnings as increased profits.
What kind of competitive edge would I get?Some of our clients consider Data Stats a competitive edge.
Chances are that there is more potential in every product or service you have. What you have to do is to seek out and unlock that potential.
Your products and processes are optimized for your work environment. Your benefits can be a lowering of the cost of producing these products or services leading to industry beating profitability, earning more from maturing products with declining sales, improving your quality levels or your product's performance, enlarging production output or you may be able choose a combination of the above.
Even if a competitor reversed-engineered your product, he may not be able to reproduce the exact factors that gives you a boost over him. There will always be certain characteristics that will give your company a unique selling position. Data Stats tries to locate these factors so that you may use them to your advantage and protection from the competition.
Why should we hire someone to analyze our processes?Information analysis is part science and part art. There is the basic component of knowledge and skills plus the craft. Both have a steep learning curve. Few companies have the resources to statistically evaluate their data and even less have the expertise to know what to look for and how to interpret their observations.
It is not unusual for a company that has been operating successfully for many years to find out that their production can be improved by 20% with only minor adjustments to their procedures
Also there are situations where your staff becomes so close to their work that they cannot see alternatives. In such cases a fresh examination by someone not directly associated with your operation can make a big difference to your return on investment.
Who is responsible for data integrity?Any real benefits you gain will only result from real work. You as the client will have to make a serious commitment to the project. This must be a partnership between your company and ours in order to succeed. But, the bottom line is that it is your company.
It is very easy to be misled by poorly conceived and carelessly handled data collection and analysis. If you make an assumption based on inaccurate data it may be a time bomb just waiting to explode.
Consultants you hire will generally have broad industry knowledge but can rarely match that of your own staff when it comes to how things work in your organization. We depend upon you and your staff to supply the critical information needed to meet the goals you summarize.
Both Data Stats and the client have their respective responsibilities. The consultants assigned to your project will outline and guide you on data collection procedures, but it is critically important to have your staff collect the data. The validity of data is your responsibility. The accuracy of analysis along with its interpretation is Data Stats' responsibility.
How many factors can an analysis of variance (AOV) analyze simultaneously?The number of factors that an AOV can handle is probably only limited by your computing power and how much in preparations you wish to make. However, it is not recommended to go beyond three.
By increasing the number of analysis factors you also increase the complexity of the AOV exponentially. Increasing the complexity of the mathematical models, you will also increase the potential for errors occurring.
We recommend the first effort only be devoted to three factors. If your situation calls for the study of more than three factors and many do, it is suggested that you optimize for the three most controllable aspects, run your optimized processes for several months to confirm your improvements and then use this new system as a control in a new analysis of variance.
How do we proceed to the next step?Contact us by sending an email to service@datastats.com. A confirmation response will be sent to you within 24 hours. If you do not receive a response, please resend your email.
Your correspondence should include a contact name, title, company name, mailing address, telephone number and a brief description of your objectives and requirements.
As a follow up, you will receive a discovery questionnaire and Confidentiality Agreement by certified mail. If further clarification is needed, a telephone conference will be arranged. Outside the U.S. and Canada alternate communication methods may be required.
After evaluating and accepting your request, you will be sent an outline of proposed experimental procedures, estimates of the time plus costs of completing your project along with a contract for your review and approval.
|
