A great deal of market and competitive data is collected during the early phases of in the development of a new product. Some of the information is accurate and valuable, and some of it is little more than hearsay.

The analysis of this data,collected during research, customer interviews, and discussions with sales and other customer-facing departments, should provide a refined definition of the product in preparation for the development phase. However, if the raw data is flawed, no amount of analysis will provide successful results.

The goal is to use the data we’ve gathered to create the product specifications. Creating specs from the large set of customer data is a challenging task and it is helpful to divided the initialdata into broad categories.  

1. What data is important?
2. What should be ignored as either not applicable or unreliable?
3. Customer requests

Let’s explore each one of these in some detail to determine how to effectively apply the data in each step.

What data is important?

You will gather a large amount of raw data during the dozens of interviews and in the many of hours of research. Not all of that data will be important; some of it can be damaging.

Part of the job of creation is to determine which portions of these data are important enough to include as part of this project. In general, if you find the same data from more than two reliable sources, consider it valid.

Of course there is the risk that some of the data labeled as “unimportant” is in fact important. This is a small risk for which you can compensate for by retaining all of your raw data until the project is complete.

If any questions arise, just refer back to the original data set. Otherwise, trust your intellect and insight during this process of qualifying the data.

What should be ignored?

Some of our colleagues would say that all of the data we collect during the early portions of a new project is important. This is not only untrue but it can lead you down many dead end paths. Here are a few examples of data I consider unimportant:

• Information gathered during interviews that apply to different product categories
• I recommend logging these data and sending it to the product manager responsible for that product.
• The data is from second- and third-hand sources
• The data point is unique
• This is particularly important if you have multiple other sources that contradict this single point
• The data that is not applicable to this project
• All data from unreliable or unproven sources

Customer requests

The requests for features that your competitors have but you lack will be the most common set of data you receive. It will also be the largest set of data you gather because it is simplest to obtain.

Most of your customers and potential customers are aware of the many products in the market. They also know the details and operational characteristics of the products they frequently use. Lastly, it is a straightforward exercise to compare the products’ literature and by actual product testing.

The main problem you will find with data in this category is deciding which features you will implement. A large majority of these features will be classified as a “must have” by your sources simply because the competition provides the feature, even if unused.

While it is difficult to ignore or refute these requests, having the discipline to question the need and the use cases not only clarifies the need, but can help you win the business without implementing the feature.

Conclusion

During the early phase of product development, what I call the “Definition” phase, your will gather and be bombarded with a great deal of data.

Some of it will be valid–some will not be. The success of your product depends on the product manager’s ability to separate the wheat from the chaff and retain only the high-value information.