What is Testing
Technically, the word testing refers to experimentation, therefore to try and verify the effectiveness and performance of an object.
The purpose of testing is to obtain, once changes have been made to the object in question, a better version. Thus, from this it can be deduced that the main purpose of testing is to improve.
Improve to create objects closer to perfection.
Taking the phrase "the best marketing is the one you can't see" as a starting point, we can say that even in marketing itself, we always try to achieve perfection. We always try to find the best strategy to adopt, we try to understand which is the most effective web page, what is the best time to get in touch with your audience and so on.
To do this it is necessary, even in marketing, to test and experiment.
In marketing it is possible to test anything, however, in traditional marketing it could lead the company to take expensive paths. What digital marketing instead offers is precisely the possibility of being able to experiment at low cost, with a risk of failure truly minimized.
Types of Testing
But how to do tests? What kind of specific tests to use?
A/B Test
This technique, which is also the most known, consists in testing two or more versions of a product (for example two different landing pages). There is usually always a control product, which is usually the original one. It is used as a benchmark. Once the various nuances of the product have been launched and made public, together with the original one, thanks to data analysis we’ll be able to find out the best one.
By applying this to digital marketing, as mentioned before, it is possible to verify the effectiveness of a specific internet page, a specific ad or an email written in one way or another. In short, there are no limits to the A / B test and it is possible to test endless digital marketing activities.
Pros:
- Immediate results
- Zero cost
- Simple to create
- Easy to communicate results
Multivariate Test (MVT)
Multivariate tests are more complicated. They consist, as in A / B tests, in testing several changes to several elements but all at the same time, for example on a web page. In case you want to optimize a web page and understand how it could perform at its best, it is necessary to understand a bit of IT.
In fact, in the multivariate test it is necessary to implement the JavaScript tags of the test tool on the page. In the mentioned case of optimization of a web page, the tags relating to the identified elements to be modified would have to be inserted. Based on the chosen settings, the various visitors will receive different pages created based on the chosen tags.
The page will start producing results immediately.
Pros:
- increasingly easier to perform
- you can stop it, pause it and start it again
- immediate results
Random Testing
In random testing the samples that are compared are selected totally at random. To select them, we rely on a computer in order to avoid any type of human error, or otherwise preference on the part of the person in charge of selecting.
In this way, with a completely randomized selection it will be possible to see as many and different results as possible and it will be possible to conduct a good work of data analysis.
Peer Group Testing
In peer group testing it happens that we tend to pair similar products and therefore not randomized. This type of test can also offer interesting results.
Analysis
As we have seen so far, the result that is produced by the tests is usually described by data (data driven). This results in a great and very close relationship with data analysis.
In fact, once the various test results have been obtained, it is also necessary to know how to read and interpret them. To this end, it is necessary to understand what you want to analyse and what type of analysis to conduct on your data.
In this case, we can distinguish 3 basic types of analysis.
Descriptive - Predictive - Prescriptive
Descriptive: it answers the question "what happened?"
In other words, descriptive analysis is a preliminary phase of data processing that creates a summary of historical data to provide knowledge and almost always prepares the data for further analysis. Descriptive analysis is useful because it allows us to learn from past behaviours and understand how they could influence future results.
Predictive: it answers the question "what could happen in the future?"
The main goal of this type of analysis is to transform raw data into useful information that could be used not only to understand past patterns and trends, but also to accurately predict future outcomes.
Such a model does not tell you what will happen in the future. Nobody exactly knows what will happen tomorrow, or a week from now. Rather it states that a certain event has a certain probability of happening. And this depends on the variables that influence the problem analyzed.
The greater the accuracy of the model used by predictive analysis, the greater the probability that a certain event will happen in the immediate future. For this reason, predictive analysis implies the search for significant relationships between variables and the representation of these relationships in models.
Prescriptive: it answers the question "how should we respond to those potential future events?"
The prescriptive analysis aims to help the company decision-making process in order to be able to make decisions that put the company in the best possible position to be successful. This type of analysis is effective only if organizations know what questions to ask and how to react to the answers.
How to combine these 3 with your testing methodology
As you can see, a vicious circle can be created between testing and analysis.
It can be said that the decision on what type of test to do will depend on the type of analysis you want to do, and so vice versa.
So basically, before you start testing you will need to understand what you want to find out. Based on this you will be able to choose in the most appropriate way the type of analysis and the type of test most suited to your situation and your purposes.
In order to create more clarity and timing for your experimentation activities, try to create a testing plan for yourself. Organize it according to the goals and improvements you want to make to your product. In addition, define the time: decide on time intervals and schedule your tests, it will help you to have clarity and keep the situation under control.
