• 0

best tools to implement this


Question

I want to create a questionnaire about a specific set of products for example: Laptops or say Video gaming consoles.

The user on the website answers a specific set of questions, based on the answers (ranging from least important to most important) it will calculate the results and get the result. That result is then tallied and then the best product is recommended.

I know a lot of websites do this. What is the solution for me to go through with this

Link to comment
Share on other sites

4 answers to this question

Recommended Posts

  • 0

I know a lot of websites do this. What is the solution for me to go through with this

The one you know.

You've basically asked "what's the best kind of paper to take a phone message on." While there are many kinds of paper, and some may be more suitable for writing a resume or death-certificate on, the task you're trying to accomplish is so simple that even a tissue or corner of a newspaper will do the job.

The standard three are PHP, ASP, and one of the popular script languages via CGI (Ruby, Python, Perl). The task is simple enough that all will do the job equally well with minimal effort. Pick one and run with it. Every answer you get recommending a specific language will likely be somebody recommending their favourite language. I'd use Ruby (without rails), somebody else would recommend PHP, and a third will say ASP.Net with C#. It really doesn't matter what you go with because they'll all work.

Link to comment
Share on other sites

  • 0

The one you know.

You've basically asked "what's the best kind of paper to take a phone message on." While there are many kinds of paper, and some may be more suitable for writing a resume or death-certificate on, the task you're trying to accomplish is so simple that even a tissue or corner of a newspaper will do the job.

The standard three are PHP, ASP, and one of the popular script languages via CGI (Ruby, Python, Perl). The task is simple enough that all will do the job equally well with minimal effort. Pick one and run with it. Every answer you get recommending a specific language will likely be somebody recommending their favourite language. I'd use Ruby (without rails), somebody else would recommend PHP, and a third will say ASP.Net with C#. It really doesn't matter what you go with because they'll all work.

any suggestions other than tools then. like which methodology is the best? I was leaning towards (least important to most important) and then tallying the results but there has to be a better, more unique and more accurate method

Link to comment
Share on other sites

  • 0

The process of asking questions and getting answers is easy enough:

  • build a list of questions and a list of ansewers: you'll probably want a 1 to many relationship (one question has one or more answers).
  • build a list of surveys. Surveys have a one to many, or many to many relationship with questions depending on if one question might apply to more than one survey.
  • Build a list of products - link them as possible outcomes for a survey. A product has a many to many relationship with surveys.

Building the list of responses is as simple as looking up the questions that belong to a survey and displaying the answers in a form.

You'll probably need a few more lists:

  • Criteria (ie: cost, american, itallian, japanese, action, rpg, mpg)
  • products (ie: ferrari 250, ford fiesta, xbox, playstation, space shuttle)
  • A table linking the answer to a question to a critiera + score
  • A table linking products to their criteria + score

Once you have your list of answers you need to weight them (ie: "I can't afford a $250,000 car" is going to carry a lot more weight than "I like cup holders" when trying to recommend a vehicle).

When you have weighted responses to questions you'll want to group them in to criteria (ie: cost, style, country, etc.) and sum them up. A person who answers "I only want American products" and "I only have $5000 to spend" is going to have different scores for the criteria of cost and country than somebody who answers "I don't care where stuff comes from" and "I'm a billionaire".

Your recommendations will be lists of products with various criteria and values assigned to them. A 1991 Ford Fiesta will have a "cost" score of 1 and an "American" score of 8 and an itallian score of 0. A 1960s Ferrari 250 GTO will have a cost score of 100, an "American" score of 0, and an "Italian" score of 10.

You take the weighted scores and match them up to products' criteria scores: the closer the match, the more recommended the product.

Your "secret sauce" is choosing good questions, assigning good weights, and accurately assigning criteria to products. The programming part is pretty simple, it's the product & market research that is difficult.

Link to comment
Share on other sites

  • 0

The process of asking questions and getting answers is easy enough:

  • build a list of questions and a list of ansewers: you'll probably want a 1 to many relationship (one question has one or more answers).
  • build a list of surveys. Surveys have a one to many, or many to many relationship with questions depending on if one question might apply to more than one survey.
  • Build a list of products - link them as possible outcomes for a survey. A product has a many to many relationship with surveys.

Building the list of responses is as simple as looking up the questions that belong to a survey and displaying the answers in a form.

You'll probably need a few more lists:

  • Criteria (ie: cost, american, itallian, japanese, action, rpg, mpg)
  • products (ie: ferrari 250, ford fiesta, xbox, playstation, space shuttle)
  • A table linking the answer to a question to a critiera + score
  • A table linking products to their criteria + score

Once you have your list of answers you need to weight them (ie: "I can't afford a $250,000 car" is going to carry a lot more weight than "I like cup holders" when trying to recommend a vehicle).

When you have weighted responses to questions you'll want to group them in to criteria (ie: cost, style, country, etc.) and sum them up. A person who answers "I only want American products" and "I only have $5000 to spend" is going to have different scores for the criteria of cost and country than somebody who answers "I don't care where stuff comes from" and "I'm a billionaire".

Your recommendations will be lists of products with various criteria and values assigned to them. A 1991 Ford Fiesta will have a "cost" score of 1 and an "American" score of 8 and an itallian score of 0. A 1960s Ferrari 250 GTO will have a cost score of 100, an "American" score of 0, and an "Italian" score of 10.

You take the weighted scores and match them up to products' criteria scores: the closer the match, the more recommended the product.

Your "secret sauce" is choosing good questions, assigning good weights, and accurately assigning criteria to products. The programming part is pretty simple, it's the product & market research that is difficult.

yeah now that i come to think of it, the product/marketing will be the killer

Link to comment
Share on other sites

This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.