Subscribe by Email

Wednesday, February 15, 2012

What are different aspects of bucket testing?


1. Bucket testing is often known by many other names like split testing and A/ B testing.
2. Bucket testing is often considered to be a market testing methodology.
3. The bucket testing methodology is used to compare a variety of baseline control samples with the single variables test samples.
4. This is done for improving the responding rates of the whole mechanisms.
5. Bucket testing is a direct mail strategy which is classic in nature.
6. Bucket testing has been employed in the interactive space to make use of features like landing pages, emails and banner ads etc.
7. Improvements can be made by testing the elements such as following:
- Layouts
- Colors
- Images and
- Text
8. One cannot expect the other elements to show the same improvement.
9. By observing the outcomes of the tests one can analyze that which elements are effective at making great improvements.
10.They are easy to identify.
11.The tester has to provide different samples of the test and controls in order to check out that which single variable is responsible for the improvement.
12.In order to make the test more effective, it has to be distributed to a sufficient audience such that any meaningful difference can be pointed out between the features and the controls.
13.Bucket testing is completely different from the multi- variate testing.
14.The multi- variate testing employs the statistical modeling using which many variables can be tested in the samples.

Today there are several companies that make use of this bucket testing like:
- Zynga
- Netflix
- Google
- Microsoft
- Amazon. Com and
- eBay
and many other companies are using this bucket testing to make effective marketing decisions and this method is gaining popularity.

It has proven to be a good methodology for maximizing the profits. Other than this, the following companies are also making use of the bucket testing tools:
- Omniture test
- Omniture target
- Google web site optimizer
- Personyze


With improving technology many types of bucket testing have been developed and are mentioned below:
1. Multi- variate Testing
this testing is a designed experiment and has been developed whenever isolation of effects is required from two or more than two factors is required.

2. Multi- variant Testing
This type of testing is similar to the multi- variate testing.

3. A/ B/ N Testing
This type of bucket testing involves more than two options or cells denoted by ‘N’.

4. A/ B/.. Z Testing
This type of testing is same as the A/ B/ N testing as described above.

5. A/ B/ A Testing
- This type of bucket testing involves only two variables but in repetition.
- One of the options or cells is repeated.


- Bucket testing is being largely used by the web sites which cater to more traffic.
- It is employed for evaluating any feature of a web site usually exposing it to a small part of the total audience.
- Then the effects on the audience are measured.
- Earlier bucket testing was used as the uniform independent sampling of the whole audience.
- This sampling was considered to be enough for producing an audience that can be used to generate a statistical data corresponding to the whole audience.
- Bucket testing holds good for social networking sites also like for example, testing a new feature that will affect only the user and his/ her friends.
- This type of testing involves much of the complex processes.
- In such a case uniform sampling method is not going to work.
- The sample should be constructed as per the structure of the user’s account.
- This is in a way helps in overcoming the challenges of the combination problems.

No comments:

Facebook activity