Jump to content

Cre8asiteforums Internet Marketing
and Conversion Web Design


Photo

Survey Application Beta Test - Justaskthem.com


  • Please log in to reply
56 replies to this topic

#1 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 06 September 2006 - 03:51 AM

New update, 9/25/06 - The survey app is now available in an iframe version. There is no need for JAVA, if you wish to not use that way. See this post for where to start

[Kim's note:

We have been awaiting the arrival of this survey application, and it has finally been offered to Cre8asiteforums members for BETA testing. This is a first for us!

JustASKthem.com is a strong supporter of these forums. For a few years, they (Alan Runfeldt) managed the server on which these forums sat, at no cost to us. We have been waiting to return that favor. What we propose to do is show you their application, which is still in development. It's ready for "people to bang on it". For now, we're not proceeding in any formal testing environment. JAT developers need user feedback, from all operating systems, browsers, software, and countries. Some code may still be in the rough, and that can only add to the fun :P

There's a lot to the survey app, and as you get to know it, more of the whole story will be told. For example, it has a back-end that rocks. For now, they're (Steve, Al and their developers) looking for basic feedback on what works and what doesn't. The more questions you ask the JAT people, the more we may get to know this app. We're trying something new...I'll let Steve take it from here.]

..........................................................................................................................

We finally have ASK3 ready for prime time.

Some Background - Why another survey application?

Most Internet market research survey companies still offer popup surveys as one of the primary tools to collect comprehensive feedback on a website. Other alternatives include having a blind link ('Click here to take our survey") or having a short form where all you basically get to ask are two or three questions.

And, of course, popus are so un-pop-ular that there there is a cottage industry in popup blockers and many people simply do not see them any more.

Almost no one clicks on the "click here to take our survey", mainly because they don't know how long it is going to take, or maybe they do and that's the problem. And, if they do click on the link, then they aren't on your page anymore and they aren't reading, shopping or buying.

Most importantly, the two or three question approach gets you feedback but doesn't answer all of your questions.

Ask3

Enter ASK3. We have a new application that we are now offering for beta testing. ASK3 takes a normal length survey and asks each web site visitor only three questions at a time. ASK3 has No Popups, No Blind Links, No Cookies and No Links that take your visitors off your page.

How it works

The ASK3 Pro surveys each consist of 21 questions. We always ask an overall satisfaction question - "Did we meet your expectations?" Then we rotate through all of the other questions, asking each visitor a new set of three. You get to ask them about who they are (male or female, how often they visit your site, how long they have been online, how much they spend online, etc.) and what they think of your website (graphics, content, prices, etc.) In the end you get a complete set of professional survey results, your visitors are not inconvenienced and actually come to think that you care. It is lke having a store clerk on every page asking the visitors, "Did you find what you were looking for today?"

How to Install

ASK3 is easy to install. Just add the following code to your web page:

<applet archive="ask3survey.jar" width="420" height="290" code="com.ask3.survey.Survey" codebase="http://www.ask3.com"....com"></applet>

That is all it takes. As soon as you upload your page, ASK3 knows what domain you are coming from. The applet asks you to enter your email address and a password. It also asks you which of the ASK3 surveys you want to use. And that's all there is.

How to change the User Interface

You can then go to ASK3.com and log into the ASK3 Admin area using your email address and password and modify the fonts, spacing, margins, etc. so that the look and feel of the applet matches your page.

[Kim's note - JAT would like your ideas on how to make the actual survey user interface "better", "more user friendly" and in general, attractive.]

How to view survey results

Once you have collected some responses, you can go to ASK3 again and log onto the ASK3 SMART Reporting system. We developed SMART for AT&T Worldnet (Kim was involved in this part.) We handled all of the customer satifsfaction research for Worldnet customer support for seven years, during which time we developed this system. It is now available for a small fraction of the cost (free for the first 30 days).

We have five versions of ASK3 at this time - ASK3 Free will always be completely free, but is an abbreviated version. We also have versions for Blogs, Auctions, Informational sites, and Ecommerce sites. We have more surveys in the works, so if you have an idea, let us know.

[Kim's note - Inserting the app has NOT been tested in blog software, so have at it and try it in your's if you wish. It is the actual questions that JAT has ready for blog surveys. Also, note that the questions used for the example here is for a political site. Blogs]

Another cool thing about ASK3 is that you can pass on the code to anyone or put it on any web page you want. So please pass it on to your friends and your clients.

We are in beta test mode right now, so we will not be charging anything for the first 30 days. Starting in October, we will be using Paypal for subscriptions, so you can full access and use your data. (More on this later...)

Please let me know if you have any questions, thoughts or comments?

Thanks,

Steve Runfeldt
President, justASKthem.com


[Kim again: You are free to do feedback as usual, per Website Hospital standards here, and Steve is open to private PM's as well.

Testing:

1. Were you able to install it? If not, explain steps, IS, browser, etc.

2. Do you have JAVA turned off? It helps to understand known user habits.

3. If you install the survey on your site, please tell us about the experience - good, or bad. Your feedback here is vital!

4. If you use it and get familiar with it, try to imagine how much you would pay per month to get to your data.

5. Experiment with SMART. Offer all feedback on the data collection back-end.

6. Anything that comes to your mind. We're getting our toes wet, trying something new here. Ask questions and offer your advice, please.

#2 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 06 September 2006 - 06:15 PM

Welcome to the forums Steve! :wave:

This post is a test to see that you get posts on this new thread.

I plan to mess with the app and WordPress, and will report back findings.

Some other ideas for our pro's...feedback on the UI, accessibility issues, browser compatibility, installation questions. In addition, is this something you would use on your site(s)? Please explain why or why not. That information is also valuable.

Thank you!

#3 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 06 September 2006 - 09:19 PM

Steve,

For Blog Testing:

I loaded it, successfully, into a new "page" in WordPress. The page I created in my blog is here:

Blog Feedback

I did it on my laptop, in the living room while watching Sci Fi, in Firefox. (Just to give an idea of your user persona...multi-tasking female, easily distracted by barking dog, kids, and husband who sits nearby talking to HIS laptop.)

Step 1 - Once the code is dropped in, this is what we see. The copyright date needs to be updated.

Posted Image

Step 2 - Registration screen. The word "varify" s/b "verify". "Email address" can be shortened to "Email".
There is need of a Privacy Policy here, for your users, to reassure them and inform them about what happens to their email, pw, etc. The form performed with no errors. However, there are no instructions to your end users about what happens in the next step.

Posted Image

Step 3 - Confirmation. This may be confusing. If they click on "Learn more", they are taken off-site, to the JAT site homepage. My guess is they want to know how to customize it first. They want to understand what happens next. The landing page (homepage) drops them off a cliff. I would add "Let's Get Started" as another button, to get the guidance going immediately.

Posted Image

(to be continued)

#4 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 06 September 2006 - 09:35 PM

WordPress Testing Continued:

Step 4 - First set of questions appear. Not customized. I did not wait for my confirmation email (since I'm more familiar with the app.) My next step was to answer the questions themselves. I wanted to see what appears on my blog, as well, since I'm working in a "live" environment. No functional errors on answering questions.
I told myself my blog is great. heh.

Posted Image

Step 5 - The T/U msg refers to "services". I need to check if these messages are customizable. For ex., the msg doesn't fit a blog. Note to other testers - The link to "Learn More About Ask3" appears on Free/Trial versions. It disappears in PAID subscriptions.

Posted Image

Step 6 - Leave site, return to see what the survey does next. Confirm it presents a new set of 3 questions.
Remind me, again please, if there is an option to add instructions that someone can keep answering more questions if they wish (refresh page to get new questions?).

Posted Image

Step 7 - This is the survey on a browser that has JAVA disabled. There must be a message displayed, to avoid the blank page. This screenshot is taken in Firefox.

Posted Image


Additional:

The confirmation email arrived safely. However, it comes from Roger, the lead developer. He likely prefers that this is changed :)

I would link this, so users can go directly to where the email is pointing them, rather than forcing them to search for it - "Client Central Utilities."

It worked flawlessly. I need to play with customization next.

#5 bragadocchio

bragadocchio

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 15634 posts

Posted 07 September 2006 - 12:03 AM

I've installed the applet on a page on my blog, at:

http://www.seobythes...om/?page_id=292

I'm assuming, though I wasn't really told expressly enough for my tastes, that I need to make the status of this account "active." Somehow, I just can't seem to change that status though.

Help? :)

#6 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 07 September 2006 - 01:19 AM

> I'm assuming, though I wasn't really told expressly enough for my tastes, that I need to make the status of this account "active." Somehow, I just can't seem to change that status though.<


The indication for "active" and "inactive" is something we have to change. It is intended to indicate the status of your subscription. We are in Beta test mode right now and are not charging for subscriptions, so all surveys have an "inactive".

Have you gone to the SMART reports yet?

Thanks,

Steve

#7 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 07 September 2006 - 01:49 AM

I don't know if I'm the only one with this, but here goes: I hate it when normal webpages integrate Java applets. On my systems, since I don't like Java pages, it takes quite a bit of time to get Java up and running when I open a page like that, it blocks the browser while it's loading, it eats ressources. It gets on my nerves when a normal "static" page thinks that it needs java to display some small (usually unimportant) part and blocks the rest of my system to display that.

That said, I do have some pages with Java that I visit regularly (some of our distributers shops are made in Java, etc) - however those sites have a good reason for Java, things that can't be done in html + javascript, calculations, better user-interface, security, etc.

For this applet I don't really see compelling reasons for it to have to be in Java. To me, this would be one main reason why I would not put it in a prominent place on my sites. I'll test it and play with it a bit, perhaps put it on a peripheral site, but as long as it's in Java I don't think I'd add it to an important site. I worked hard on those sites to get them as lean and clean as possible, adding Java-Applets would be almost as bad as adding ActiveX components :lederhosen:.

My suggestion:
- Instead of coding it as an embedded applet, use a javascript loader. That way you can change the script to do whatever you want - embed an applet or inject an iframe, etc. It gives you a bit more room to grow.

- If it has to be in java, use a javascript loader for it. I haven't tried, but perhaps you can set up an IFrame with a javascript in it with the "click here to answer a few simple questions (requires java)". When the user clicks on that, the applet is loaded in the iframe.

- Or even better: do it in javascript and do the security on the server (not in the javascript). Set up an iframe, use javascript + perhaps ajax (probably not even required), let the server create the javascript on a per-user basis (similar to Adsense).

Let's see what we can do with it now :blink:

John

#8 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 07 September 2006 - 01:59 AM

PS it's cutting off the third line:
Posted Image
(On Firefox 1.5.0.1 on Windows 2000 Prof)

#9 bragadocchio

bragadocchio

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 15634 posts

Posted 07 September 2006 - 02:48 AM

The indication for "active" and "inactive" is something we have to change. It is intended to indicate the status of your subscription. We are in Beta test mode right now and are not charging for subscriptions, so all surveys have an "inactive".


Ok. So it's not me. That's good. :lederhosen:

Have you gone to the SMART reports yet?


I have now. I wanted to wait until I got a little feedback in there. I answered a few questions, and I may have had a visitor or two who provided some feedback. I did notice an error in the reports - it asked me if if I were male or female, and I answered male. In looking at the report, it had entered that answer into the female category, but there, when I click on the "Gender - Female" link, it says "Male 100.00% "

While the report interface is functional, it leaves me with a number of questions. I guess I'm spoiled by a lot of ajax applications, and other programs I see on the web.

Why show me results from "Apr 2006 May 2006 Jun 2006 Q2 2006 Jul 2006 Aug 2006"? That sort of has the feel of an ecommerce site asking for my credit card showing me past years in the date expiration field.

I'm not sure how helpful a percentage based metric is on the main report page, or what it means exactly. Putting the different choices for each of those categories on the main report page, would increase my "wow" impression, and maybe show the data in the most meaningful manner.

Why are there a separate Gender - male, and Gender - female categories? It's going to be 100% male or 100% female, isn't it? What's more meaningful to me is the percentage of each in one category.

I'll probably have more for you. Let's see if I get some more feedback, and see if it gets me to think about it.

#10 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 07 September 2006 - 11:50 AM

Thank you John and Bill! This app has desperately needed objective feedback. That's why it's here. This thread showcases the need for user testing during development, which is common debate in QA, user centered design and software dev fields.

John, I agree with your views on JAVA. From side of the usability fence, it's a critical factor. Load time, especially on homepages and top level pages of sites often determines how long a visitor will stay. As soon as plug-ins come into play, large images, and JavaScript come into play, the number of visitors falls off. For an ecommerce site, this is unacceptable, which is why many eventually settle on workarounds. JAT has an expert JAVA programmer. I hope that if JAVA remains the backbone, that a solution for usability is found.

Bill, I too, need to spend more time looking at the new SMART system. I'm very familiar with its birth mother version, which was created for large scale, global usage for the collection of customer service data. Steve can correct me here, but my elementary understanding is that the evaluation of customer service/marketing data is different than what SEO's are used to working with, such as CPU, logs, page views, etc. SMART not only collects the data, but it can "think" about this data and help the site owner form conclusions and even ask questions of the data itself. This is why I get excited about this app...knowing what it might do once it gets past its hurdles.

For usability, specifically cognitive load, there is a lot to learn about how the app works. Similar to learning and getting to know Click Tracks, WebTrends, etc. I'm finding that the site design is sorely lacking in user center design elements. JAT needs to hear about every place where someone is confused or lost, no matter how painful it is to express or hear. This is how they can work on making improvements in user instructions, and incorporate call to action prompts where they are missing (a big problem), and built-in reassurances and guidance. That stuff is difficult for any site, so all feedback is really helpful.

Steve, there's a disconnect between the moment when the survey is loaded and functioning and the SMART data collection side. An email to new subscribers is one way of handling this, as your way of communicating. Another is one I think Al will like, and that's installing a video demo on high level instructions, as an intro to SMART and a separate tutorial for customization. I know Roger has spent a lot of time on writing out instructions. These days, demos are hot. Step by step illustrations are also helpful.

Your team may benefit from checking out Text Link Ads (http://www.text-link-ads.com/). This company also had an idea, and it was confusing when they first launched it. But, they began to get feedback and did usability site work. The result has been a few versions of their website, to what you see today. They keep finding better ways to simplify, but also clearly explain and instruct what is sometimes complicated. They offered good visuals to describe steps and they communicate via email on a regular basis to subscribers. The determination paid off, as they rose up to fame practically overnight.

#11 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 07 September 2006 - 04:09 PM

Thanks so much for the feedback. (Of course I was hoping for "Great, I wouldn't change a thing. " <vbg>). This is very helpful.

I did notice an error in the reports - it asked me if if I were male or female, and I answered male. In looking at the report, it had entered that answer into the female category, but there, when I click on the "Gender - Female" link, it says "Male 100.00% "


Thanks for noticing this. This is one of those things that we always pride ourselves on. It is crucial to maintain data integrity. We would be sunk without it. We will fix this one ASAP.

Most of the questions are five point scale, so we can show an indexed mean value. We purposely put the scale at 100%, 75%, 50%, 25%, 0% because a % scale is most easily translated cross culturally. In some cultures giving an extreme rating like "The Best" or "Completely Satisfied" is just never done. In other cultures, unless there is something terribly wrong, everyone gets a "Completely Satisfied". By using a 0-100% sale we make the results more nearly universal. And a mean score of 87% means just that - people agree 87% of the time. I can go on and on on this topic, so if anyone wants to discuss rating scales more extensively, let me know.

With dichotomous questions (yes/no, male/female) in the trends we show the %Yes. For questions like Male/Female, we would get into a lot of trouble if we just showed the %Male or the %Female, so we show both. This is probably where the glitch came in. This kind of question is a special case. We may also have some fine tuning to do with the trend report for categorical responses like "How do you access the Internet?" These are much better handled within the crosstab reports.

Thanks,

Steve

#12 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 07 September 2006 - 07:23 PM

Ditto on what John said. I get the same thing on that rotation of question.

I'm running into trouble in Wordpress. When I want to modify the page that the survey code resides in, inside WP software, since the app is already loaded, for some reason its not permitting me to edit the page in any way. The app is even stalling the browser (Firefox), causing a need to reboot. Two tries and I'm unable to edit the page. It's as if the app insists on running while I'm trying to edit (add content) to my page.

Screenshots I could send, but you'd have to have access to my blog to really test it. Might be worthwhile to install some test blog software so you guys can see what it does...

#13 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 07 September 2006 - 08:24 PM

The workaround for my problem was right in front of my face. Eric gave me the hint.

I need to turn off JAVA options in Firefox to edit the page. Anytime someone is editing something with a preview option, such as how WordPress works, the app has to be turned "off", to be able to edit the page.

Once I did that, I had no trouble editing the page.

#14 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 08 September 2006 - 02:13 AM

>> PS it's cutting off the third line:
IPB Image


This is a question of formatting the internal margins in the applet. The default configuration causes this cut-off with the short questions, such as this one "Are you?" with the responses Male or Female. The trick is to set the top and bottom marigins at 2,2 for the questions and at 0,3 for the answers. I took the liberty of fixing this on yours.

We have also fixed the glitch with the male/female responses in the blog survey. You are a man again <g>

Also, you have a comment about the applet overflowing into your second column. You can control this within the applet itself by changing the width and height. Take a look at the ASK3 applet on http://www.frenchtown.com/ (about 1/2 way down the page).

Thanks,

Steve

#15 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 08 September 2006 - 02:35 PM

First of all, thanks to all of your cre8tive people who have invited us in to your forum. I'm very glad to be here again.

Now, as to concerns I see expresed above. Great points, all. Thanks for taking the time to consider them and to pass them on to us. Now let's see if I can add some content of value to this discussion myself...
===========

JAVA:

On this question on the advisability of Java as a mechanism for ASK3. I understand your concerns and can reassure you that Java as the input technology is not etched in stone. ASK3 was developed in Java because at the time this project was begun, Java was the preferred means for embedding our survey code in a privately-managed page.

And we believe that Java is still an excellent technology for this application and offers some tremendous features. Of course, as we know some do come at at a price and web technologies continue to evolve. But I can assure you that this applet is well-written and stable and we have made a great effort to assure compatability with a wide range of JVM implementations. The ASK3 applet was created by our own Java Guru who has written books on the subject and has taught many classes in the technology. And, at the time we began, it really was the only serious way of accomplishing our goal.


iFrames version coming

But browsers have acquired many new capabilities since then, and we do have an iframes cgi version in the works which would eliminate Java from being an issue entirely. Of course, we still don't know what issues the iFrame version will raise, (as you may recall from my visit to the programming forum in August) but the technology looks pretty solid and broadly supported in browsers. I'm sure we'll get some excellent feedback from you guys when we release ASK3-iFrame for testing.

We are also exploring css and XML and discussing a Flash version as well. We are also thinking of other means of delivery besides web pages - cell phones, cable tv, etc. It's our idea to let the user choose the technology they favor. Although tremendous effort has been expended to develop the JAVA version, and it will remain our model for future releases, we will continue to explore other available technologies as they mature.

IT's NOT THE TECHNOLOGY

But the important thing about ASK3 is not the technology used to deliver the 3 question survey, but the concept of a portable page-embeddable 3 question survey that samples many questions over a large population and that is less intrusive to the user than other survey instruments and collects raw data that can be analyzed by our mature and proven SMART reporting system.

REPORTS

SMART analysis is the real strength of the system in that it gives the user actionable information and far deeper and comprehensive analysis of survey results than most results analysis systems, no matter how many pretty charts and graphs they employ. But don't be intimidated by that. No one needs to use all of the SMART tools to learn important things about their user responses.

SMART's design was greatly influenced by professional market research principles and has proven its worth over seven years' real-world use by ATT Worldnet in maintaining and improving customer satisfaction among the users of their Customer Care systems. SMART feedback gave them the information to improve customer satisfaction and win awards. It's proven. ASK3 analysis is an extension of that system and will also evolve as new reporting requirements and methods of presentation are recognized.

COMMENTS as a measurement tool

What all users really should understand, however, is that the most favored and interesting responses - the verbatim comments - are also the least reliable evaluation of overall opinion on any matter relating to understanding the web site visitor's point of view. After all, a verbatim comment is quite literally "just one person's opinion". Sure, verbatims are interesting and are almost a requirement in any survey environment, but they can be tremendously misleading compared to analysis of the broad numbers collected in the "select an option" questions - WHEN the options are wisely configured to ellicit an accurate and valid response to the question we need to have answered.

So. Enough from me for today. Time to get back to coding that iFrame version...

Thanks

-Alan

Edited by Alan Runfeldt, 09 September 2006 - 03:09 PM.


#16 dgeary9

dgeary9

    Mach 1 Member

  • Members
  • 334 posts

Posted 08 September 2006 - 08:37 PM

Stever, I don't have a personal site that gets enough traffic to run this, but I have a couple of clients who might like to be guinea pigs.  I understand the tool's in BETA, but can you give me a better sense of what you mean by beta?  I know you want feedback on design/functionality.  Is the backend decently reliable (e.g. can my clients rest relatively assured that data collected on their site is all making it into the backend and visible in the reports)?  

If they run it, we'll likely be making some business decisions based on the data (and as a web analytics geek, let me warn you how much feedback you're likely to get from me on the reporting back end   :rolleyes: )...

#17 Rowan

Rowan

    Gravity Master Member

  • Members
  • 142 posts

Posted 09 September 2006 - 01:58 AM

One other reason NOT to use a Java applet is because the text is badly aliased/pixelated in the applet and isn't affected by Clear Type on Win XP. Also, what if one wants to use their own font family, size and colours?

#18 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 09 September 2006 - 02:10 AM

Stever, I don't have a personal site that gets enough traffic to run this, but I have a couple of clients who might like to be guinea pigs. I understand the tool's in BETA, but can you give me a better sense of what you mean by beta? I know you want feedback on design/functionality. Is the backend decently reliable (e.g. can my clients rest relatively assured that data collected on their site is all making it into the backend and visible in the reports)?

If they run it, we'll likely be making some business decisions based on the data (and as a web analytics geek, let me warn you how much feedback you're likely to get from me on the reporting back end rolleyes.gif )...


Absolutely, yes. We have been running the ASK3 applet itself for about 6 months while we were integratging the SMART reporting system. We originally developed SMART for AT&T over the course of several years, for tracking satisfaction with their customer support. The trick was adapting the same reporting system to a very different data structure.

Our main interest in the Beta test is to get opinions from developers as to the usability of the installation and the admin system. We have lots of client feedback on the reporting system and it has proven itself quite well. AT&T used the system to get the highest ratings for support among all dial-up ISPs. We even helped thier new Indian call centers to go from 60% satisfaction to 80% in about three months time.

SMART was always a work in progress. It is somewhat modular so if grew substantially over the seven years we ran it for AT&T. If you are into analysis, you may like the driver analysis feature we are planning for V2.0. Basically, the data structure of ASK3 enables us to do some things statistically that are problematic with normal data sets. We have a white paper on driver analysis and ASK3 statistics on the site.

We also want to make sure that we haven't missed anything - like the male/female glitch in the reports that Bill found. But, it would be great if you were to install it for your clients. We are offering a comissiopn on sales to others too. You can contact me offline about that.

We can also do custom versions of the surveys if that is desired. But the standard ASK23 Pro surveys are free for the 1st 30 days.

So, please go ahead and set it up for your clients. If they don't like it, within 30 days they can just not subscribe. That is also one of the best aspects of our marketing plan, people can install the applet at no cost and run the surveys as long as they like. We only charge to view the results. And we are keeping the price as low as we dare.

Thanks,

Steve

#19 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 09 September 2006 - 02:24 AM

One other reason NOT to use a Java applet is because the text is badly aliased/pixelated in the applet and isn't affected by Clear Type on Win XP. Also, what if one wants to use their own font family, size and colours?


Yes, we do offer a few options for fonts - a full range of sizes, serif/sanserif, full range of colors for both the fonts and backgrounds, italics and bold. Each of these can be set for each question and/or answer set and for the header and footer panels separately. Also, the size of the applet can be changed.

When you log into the Admin area at http://www.ask3.com/login.html you can click on the Fonts tag, select a panel consisting of any combination of banner, footer and question and answer sets and then choose the font characteristics for that area of the applet. You can also change the margins and borders by selecting "Borders". If you have more than one page under your ID (email address) you have to choose which page you want to change.

I recommend changing the Top and Bottom margins so that the Top/Bottom margins for Questions are 2/2 and the Top/Bottom margins for Answers are 0/3 (where 0 is the top and 3 is the bottom). We have a glitch with short questions like "Are you?" "Male or Female" so that the space allocated for the question is slightly truncated in the default mode. Setting the margins as I describe above takes care of the problem.

If you want to change height and width other than the default values you will also have to play with the margins and borders to make sure everything fits.

For custom versions I am sure that we can do much more, but we also have to charge more too.

Hope that helps.

Steve

Edited by stever, 09 September 2006 - 02:26 AM.


#20 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 09 September 2006 - 02:35 AM

I would recommend setting up an instance of ASK3 on a demo page - a test.html - and then playing with the admin system to become familiar with the features. Then you can answer some questions (just refresh to get a new set of questions) and add some test data. Then you can also go into the SMART report and see how the results are added and some of the ways you can slice and dice the data.

You can also view our ASK3 Free demo at http://www.ask3.com/demo.html Enter some answers and then view the results, enter some more answers and see how the results change.

For more on the SMART reporting system you can also review the demo at justASKthem.com This is the site for our more traditional Internet surveys. The SMART demo there has over 30,000 test cases (based on some actual customer data), so you can see how the system responds with a little bit of a load.

Steve

#21 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 09 September 2006 - 03:26 AM

What is the preferred way of setting up something like this on a site? I have thought of the following, but I'm sure you have more experience and can probably tell me more :)

Perhaps I should add that I don't want to add it to every page viewed, but I would like to get as many people to use it as possible. Idealy I would love something as simple as the "5-star one-click-rating" to add to all pages - I realize that this tool would be much more than that and that it does require a bit more on the side of the user.

I was thinking about:

- randomly adding it to all pages viewd (perhaps in place of existing Adsense-blocks) - what percentage would be good in your experience?

- randomly triggering a javascript messagebox "I'd like your opinion, very short questionaire, only quick 3 questions. Is that ok? OK/CANCEL" on all pages, sending them off to a questionaire page (perhaps with a short writeup with the reason behind the questionaire, that it's in java (in case they don't see it ;) ), what I hope to learn, etc. Again - what percentage would be good?

Some questions that just popped into my head :) :

- When talking about the percentages here, I assume I should make sure the visitor only gets one questionaire per session, is that correct? (meaning I can't use a percentage of page views, but rather of "unique visitors" / sessions) or would it make sense to trigger multiple questionaires per session?

- Could the "random display" perhaps be controlled on YOUR end (I don't know if you can track sessions, but page views should be possible)? That would make it much easier for the user. Just add it to some of the main pages, it'll trigger automatically when required, no need to do any server-side code for that.

- Do I need to "cloak" it away from the search engines? I would assume not, but you never know how they look at things like that. (or perhaps someone else knows how search engines handle java applets? Do they index text-snippets out of it? or leave it alone?)

- Looking back at my first point... how about a simple system that alternates your questionaire with Adsense blocks (for the webmaster) or at least some user defined image? That would help relieve the "ad-blindness" and at the same time offer a good place for the questionaire. An Adblock 250x250px should have enough room for that, perhaps you could even make it work with most of the other standard Google Adsense-sizes. I think that would be neat :).

John

#22 yannis

yannis

    Sonic Boom Member

  • 1000 Post Club
  • 1634 posts

Posted 09 September 2006 - 09:08 AM

Steve

I tried to install it on a blank page first and run it using localhost. The registration process went smoothly. I filled one survey and did a couple of refreshs to try and get a second survey and see how the system reacts. But there was a problem with the java applet both using FF and trying it also with a non-standards compliant browser (Internet Explorer 6.0 :( . I guess the localhost must have cause the problem.

One comment on the confirmation e-mail you should try and provide a link as to where to view your results. What happens if you install it on more than one website will it be able to keep track of all 'channels' using one account?

I like the idea and wish you well. My thoughts is that it may do well for certain clients. In the mass market you might face resistance both with the price (after-all you get free surveys with most blogging and CMS software). Your 'free' offer is not really free, if one cannot view results within the thirty day 'free period'. Size of block and survey questions should be left up to the user to create.

Yannis

Java Plug-in 1.5.0_06
Using JRE version 1.5.0_06 Java HotSpot(TM) Client VM
User home directory = C:\Documents and Settings\User


----------------------------------------------------
c:   clear console window
f:   finalize objects on finalization queue
g:   garbage collect
h:   display this help message
l:   dump classloader list
m:   print memory usage
o:   trigger logging
p:   reload proxy configuration
q:   hide console
r:   reload policy configuration
s:   dump system and deployment properties
t:   dump thread list
v:   dump thread stack
x:   clear classloader cache
0-5: set trace level to <n>
----------------------------------------------------

java.lang.NullPointerException
	at com.ask3.admin.beans.Attr.getValue(Attr.java:52)
	at com.ask3.survey.Ask3Survey.getMainPanel(Ask3Survey.java:90)
	at com.ask3.survey.Ask3Survey.getMp_Base(Ask3Survey.java:688)
	at com.ask3.survey.Ask3Survey.initialize(Ask3Survey.java:70)
	at com.ask3.survey.Ask3Survey.<init>(Ask3Survey.java:61)
	at com.ask3.survey.Survey.getAsk3Survey(Survey.java:214)
	at com.ask3.survey.Survey.init(Survey.java:96)
	at sun.applet.AppletPanel.run(Unknown Source)
	at java.lang.Thread.run(Unknown Source)


Yannis

#23 cre8pc

cre8pc

    Dream Catcher Forums Founder

  • Admin - Top Level
  • 13454 posts

Posted 09 September 2006 - 02:50 PM

John, I like that idea of switching the survey on/off with AdSense...esp. if the applet can size down enough and still fit the questions in.

I haven't played with the new SMART system yet but will. I'm a huge fan of how it works, though I never fully was able to grasp it all. I could tell, by the feedback from AT&T employees, that it was amazing.

Steve and Al, your answers are really helpful. Thank you!

Btw, I ditto what Yannis mentions, about the confirmation email pointing to setup links, SMART, etc. I know I've mentiond it before...can't remember if here or privately.

#24 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 09 September 2006 - 04:17 PM

... I'm currently setting this up on my oy-oy.eu site in rotation with an Adsense block ...

- Is it correct that you have to register the survey applet for each page you have it on? How does it react to querystrings? I was planing on showing it randomly all over the site in the navigation. Since the user can enter his own URL (which is then placed into the querystring), I'm worried that they could register a few of the URLs as their own queries (instead of them all being tied to my site). Suggestion: allow the user to enter a top-level URL and use the same settings for all URLs in that path (eg http://domain.com/ or http://domain.com/users/joedoe/ ).

- Related question: what happens if either 1) I add a new site/folder/page and put the code on it and a visitor spots it for the first time? (eg when I move from my staging domain to the live one) or 2) you happen to lose my registration (for whatever reason) and a visitor spots it for the first time? Or in other words: How can I be absolutely *certain* that nobody else is reading my results if I don't pay attention? Suggestion: I would add a website owner verification similar to Google Sitemaps to make sure that only someone with real access can set up the survey.

- What is "Survey Category"? I don't understand it (not initially and not after re-reading this thread :(). Is it the category of service (eg free / paid / different paid levels)? or is it the category of website that it's on (eg determines which sets of questions come up)? If it's the type of website then I don't understand "free" :).

- You sent me my password via email in plaintext. Yikes!! Don't do that.

- Where do I get access to my "reports or the Client Central Utilities"? Do I have to search for that page on your site or will it come via separate email?

(probably more to come, if I can make sure that the setup routine doesn't trigger with visitors :D)

John

more .. :)
- the survey category doesn't react to keyboard input (eg "I" for "Information") (I'm wondering how many times I need to set this up :D)

- the login is ... strange :D

- the login is a single email-address / password pair: however I have registered a total of 20+ pages with that email address and could have used different passwords. Where would they end up? Or is just the first password for the email address valid?

- (tell me when to stop) the login / admin page / applet sends my browser overboard (100% CPU - all fans starting up, ready for liftoff). The survey applet is not problematic (I don't have a clue about java).

- what's an "alias URL"? Is that a simple way to assign values for all pages? (I don't even have that many, but I would hate to have to assign all values for the pages separately...) [update: I think not :( ]

- as you already explained, my query status is also "inactive" -- if I click on "change" it tries to open a popup which my Firefox correctly blocked ;) - is there a way either around the popup blockers or wouldn't it make more sense to just open a new page/window or even just redirect the current one?

- is there any way to reset the results after playing with them?

- the "how often do you use the internet?" question has some replies that might be hard to understand: "several hrs/day, LT 1 hr, 3-6 times/wk, 1-2 times/wk, LT once/wk". Someone without Fortran experience might have trouble understanding "LT", I also don't know if "wk" is a standard short form (not being english native). Perhaps writing them out and putting them on 2 lines would have been better. (I assume this also applies to some of the other questions I haven't seen yet)

- how can I change the category for my pages? I played with the settings and it does determe the questions that come. Now I have an auction, blog, ecommerce, information and free site all in one but in retrospect it would have been smarter to keep it at "information" (I think, not certain).

Edited by softplus, 09 September 2006 - 04:45 PM.


#25 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 04:53 PM

1) dgeary9 - First - about the data. No worries. We've been collecting and managing data in various formats for ten years. We're pretty good at getting it and keeping it. In seven years with ATT, we never - I repeat - never lost a single survey response, and experienced almost zero down time even though we migrated between systems three times as the system grew and we updated servers. In fact, none of you have seen it (yet) but we actually have two report systems. The one I put together is totally independent of SMART and has been used for well over a year for internal testing of data collection and integrity.

We do read from the same SQL database, but mine is a simple frequency-count (basic) analysis which we used to compare results with those presented in SMART. When the issue of Bill's male/female question came up, the first thing we did was turn to the test report to assure that we did indeed have the correct answer stored in the database. That's how we narrowed the issue down to the data display issue for this particular question which, due to it's nature was handled uniquely in SMART. The normal either/or report option was not appropriate for this particular - and unique - only-two-options question.

2) Rowan - Java does indeed handle fonts differently than browsers - as does the Flash implementation we are considering. Integrating your custom font into the system is something beyond the scope of the current testing, but I will discuss it with Roger and get back to you with a complete explanation.

Our iFrames version, on the other hand, will display only locally available fonts via css. The applet comes with a default css reference from our server, but it can be overridden in the ASK3 code you embed in your page.

Note also that we are experimenting both with basic css font calls as we know them as well as the new panose intelligent font system.

3) Yannis - Ah, so you are the mysterious "Late Night Engineer" who installed their applet at "http://localhost/lat...er/survey.html" :( yup. Can't send a java applet to localhost... :D Seriously, though. I'm glad this issue came up. Disabling localhost installations is already on our to-do list. Thanks for reiterating our need to do that.

REPORTS

As I mentioned above, we actually have two reporting systems in place for ASK3. Our in-house version does not offer the features of SMART, and does not currently include the login security module. We may make that one available to you all in the near future.
In the meantime, the SMART reports are available at http://ASK3.com. Select "ASK3 Login", which will give you the option to log in to the Admin System (Where you can configure your the spacing, fonts and colors used in your applets' displays, edit your staff accesses, change passwords and other features) or take the short cut to "Fast SMART Login" which takes you to the bookmarkable http://www.ask3.com/...TrendsReport.do

And, if enabled on your system, the SMART login page will accept your system-supplied login info to fill in the blanks automatically and make life easier for you.

4) Confirmation email: Yes, it will include links to the pertinent pages. It, too is on the to-do list. Thanks for reminding us.


5) And, last but not least, to Kim - glad you like seeing the photo. Now you can remember how serious I look when working... :D

- Alan

Edited by Alan Runfeldt, 11 September 2006 - 12:23 AM.


#26 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 05:31 PM

- (tell me when to stop)


Softplus. Yikes indeed! :( but don't stop. This is why we are here. You guys are the best :applause:

Seriously. You have posted many very good and very important questions. Many of them are answered in our help pages, many obviously need to be added.

But, before I get into that in depth with detailed replies, I would suggest that you cruise our site and see if the information we present there answers some of the questions you have posted. http://ASK3.com Besides, if you read as fast as you install applets, you'll absorb it all in a short time...

(Note: I also receive cc's of the installation messages, so that I can keep track of new installations.)

In addition, I have sent you a private message via email - from "survey-admin (at) justASKthem.com". Perhaps my note plus a direct chat with Steve will fill in the blanks quickly.

In the meantime, we will "process" your comments and compose a hopefully brief, certainly complete response to the flurry of questions as soon as we can catch our collective breaths.

One problem with having lived with the evolving code for so long is that there are many assumptions we might make about how someone else might understand the system. That's why we're here - to get a fresh view from a variety of new users. Your questions are great and are already helping us to do that.

more later (soon)
- Alan

... I'm currently setting this up on my oy-oy.eu site in rotation with an Adsense block ...


Softplus - No worries.

I just checked one of your installed pages. First load showed the ASK3 survey applet, three subsequent loads showed the adsense content. Next load showed the applet again.

Looks like ASK3 is quite compatible with the Adsense block rotation. Keep an eye on it and let me know if my quick assessment holds up.

BUT. Keep in mind also that ASK3 is hungry for page views as well. Sure adsense pays money, but we give you feedback on your site. You decide the value of each and adjust your rotations accordingly. Also - not everyone will respond. The more often the ASK3 applet appears, the more often you will get information.

But the good news is that it sure appears to be compatible to me.

- Alan

Edited by Alan Runfeldt, 09 September 2006 - 05:59 PM.


#27 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 05:57 PM

(eg when I move from my staging domain to the live one)

Ooh.. Watch out here. An applet is registered to a url. ASK3 is designed to be registered on line - on the page it is intended to monitor. The first time the applet apears on a unique url, it displays the registration page. You can install the code off line or on a staging server - no problem, but just don't register it until it is in place on line. Once it's registered, the url displayed in your location bar - the referring url sent to our server - becomes the default url for that page.

Be sure to register it ONLY when the page is online in its permanent home.

However - there is more to this...

ALIASING urls

The system does allow you to register aliases for an installation, and automatically enters a group of default aliases for you.

ex: http://ask3.com is automatically assigned aliases of
http://ask3.com/index.html
http://ask3.com/index.htm, etc... - we supply a half-dozen common aliases, to make life easier for you, and the alias list can be edited.

I think you might want to spend some time studying the aliasing options in the Admin Center.

Please log in to the Admin page and take a look at the aliases set up for one of your pages to see what I mean. "Select Aliases to review the different URLs that refer to your Web page."

One more thing. You can assign an alias from a totally different domain to an existing applet. In other words, you might have an applet installed (as I do) at http://NoDeadlines.com/vietnam/. But now we have a domain name for that site. The old url is still valid, but I have pointed Apache to show that page also as http://8thmob.org or http://www.8thmob.org. Using our Aliasing feature, I could have added the 8thmob urls to the NoDeadlines.com/vietnam registration.

ASK3 really is quite flexible that way.

But, to make life easier for all, install the code where ever you edit your page, but only register the applet when it is at home online in its permanent location.

Thanks
- Alan

#28 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 09 September 2006 - 06:17 PM

Thanks for the quick feedback, Alan! Now I've got something new to play with :(.

Perhaps an interesting value to show would also be # views (or a simple conversion-rate from viewer to answerer). That might help webmasters work out which location is best (where they get the most answers).

John

#29 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 06:22 PM

- Do I need to "cloak" it away from the search engines? I would assume not, but you never know how they look at things like that. (or perhaps someone else knows how search engines handle java applets? Do they index text-snippets out of it? or leave it alone?)

Softplus - No worries. No need to cloak anything. Google ignores java applets.
I googled the main code line of the installtion code = "ask3survey.jar" and the only 3 places it came up were on the pages where we display it as plain text designed for cutting and pasting it into your page.
- Alan

#30 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 06:35 PM

Perhaps an interesting value to show would also be # views (or a simple conversion-rate from viewer to answerer). That might help webmasters work out which location is best (where they get the most answers).

Softplus - yes. Another good point. We do track views in the Tomcat logs, and we do intend to analyse views vs responses. It's a pretty simple report to pull, but that's a utility that is still in the works. We figured we better get some live action going first and see if we can't identify some rough edges from the user perspective.

In fact, I think that might also answer some one else's question of "how beta is this test". It's not really very beta. This code has been pounded and shaped and run through all the wringers we could come up with on our own. We've had our alpha installations running for quite a while, but wanted to get ASK3_SMART ready for public use. So maybe it would be better to call this 'gamma' testing... ;-)
- Alan

Edited by Alan Runfeldt, 09 September 2006 - 10:04 PM.


#31 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 06:48 PM

- the login is a single email-address / password pair: however I have registered a total of 20+ pages with that email address and could have used different passwords. Where would they end up? Or is just the first password for the email address valid?


REGISTRATION, PASSWORDS, ACCESS:

Softplus -
When you registered the first page, you became the "Administrator" of your "Organization" - like the registrant of your domain. You can register unlimited pages in your organization. And, you have ultimate control over all of the pages in your organization. Your organization can span multiple pages, multiple domains - what ever.

You will be able to assign "Controllers" to specific pages if/when you chose to. They will have plenty of control to "control" the pages you assign to them, but as Admin, you're the boss and you need only one email address and one password.

We're also working on a third level of access - well, we have it now, but we're changing it to allow a third level of access to only view the results. That way, you will not have to give out your login and password to your client. It's not activated yet, but that's one of the first new features that will become available very soon. It's under development now.

And, by the way - you can change that 'plain text' password we emailed to you in the Admin Center. That way only you and "we" (me and the dev team) will ever have access to it...
- Alan

#32 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 07:03 PM

One comment on the confirmation e-mail you should try and provide a link as to where to view your results.

Yannis - Good point. we'll be doing that soon. For the moment the easy answer is to log in to your acoount at http://ASK3.com and follow the links.
You can also go directly to http://smart.justask...t_ask3/Login.do and bookmark the page for ease of access.

What happens if you install it on more than one website will it be able to keep track of all 'channels' using one account?

Yes. That's the beauty of the admin system. A lot of thought went into the Administration and Staff structure. When you register your first page, you become the Administrator of your Organization. You can track all pages using one account. You can also assign Controllers to groups of pages. (see my last post - REGISTRATION, PASSWORDS, ACCESS: above.) We're also working on a comprehensive FAQ on the "Organization" of ASK3 Admin. I think we'll be adding your questions to it...

Thanks
- Alan

#33 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 09 September 2006 - 10:54 PM

On my systems, since I don't like Java pages, it takes quite a bit of time to get Java up and running when I open a page like that, it blocks the browser while it's loading, it eats ressources.


Softplus - A Note on Java loading times. - Once a day.

Another good point. I'm glad you brought this up. Applet download time was an issue that concerned me greatly as well during the development phase, and as a result, we have looked into this matter and discussed it at great length.

In reality, the ASK3 java applet download time is actually about the same as a 30k jpeg. Any delay in display that you might notice is a function of the JVM (Java Virtual Machine) loading in your browser - which it will only do once a day as you browse the web.

If you have visited any web page employing a Java applet prior to encountering an ASK3 Java applet, the load time is less than negligible. Download time is a serious concern for us and we have streamlined the applet to load as absolutely as fast as possible. We are concerned about this and have tested it heavily. But we have decided that Java is a stable technology and it will be around for the foreseeable future.

Since we do anticipate improved support for Java as time goes on (such as browsers pre-loading the JVM, for example), we do expect this concern to have minimal impact on ASK3's success.

But, as mentioned in previous posts, we are also developing alternative versions - and will continue to do so as new technologies evolve and mature - to give the webmaster more options - specifically to address such concerns.

- Alan

#34 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 10 September 2006 - 12:26 AM

John asked:

I was thinking about:

- randomly adding it to all pages viewd (perhaps in place of existing Adsense-blocks) - what percentage would be good in your experience?

- randomly triggering a javascript messagebox "I'd like your opinion, very short questionaire, only quick 3 questions. Is that ok? OK/CANCEL" on all pages, sending them off to a questionaire page (perhaps with a short writeup with the reason behind the questionaire, that it's in java (in case they don't see it wink-2.gif ), what I hope to learn, etc. Again - what percentage would be good?


Hmmmm.... ASK3 already has a randomization element to the algorithm. Traditional survey research is designed on the principle of sampling. Because it has been physically impossible to speak to every customer, either in person or on the phone, market researchers turned early in the 1900s to sampling theory, where a random subset of people represents the entire population. The idea behind ASK3 is that on the Internet you can talk to everybody. The problem, and the limiting factor, is how many people want to talk to you? Long surveys and links to surveys put people off. Popups, especially, are intrusive.

So ASK3 was designed to be WYSIWYG. Putting in a link to an ASK3 page could defeat the purpose. You will lose people. You can certainly try it, but I think that if someone is going to click on a "I'd like your opinion..." link, they are just about as likely to be willing to complete 20 questions as 3.

The real key, conceptually, to ASK3 is that the visitor sees up front that you are only asking 3 questions and doing so will not distract them from the reason for their visit.

Randomizing the applet could be interesting depending on the number of visitors and their rate of response. One reason for this Beta Test phase is that we don't know yet what sort of response rate we can expect. We have asked lots of people "How many questions would you feel comfortable answering before you found it to be intrusuve?" and most say up to five. Three seemed to be a safe number to standardize. We are also prepared to add in a couple of extra questions in an inline frame, like "Would you like to hear back from us? If so, please enter your email address." and that sort of thing.

The secondary questions are randomized, so if you have 1000 responses, you will have 1000 people who answered the primary question (Overall expecations, likelihood to purchase, etc) and 100 people who have answered each of the detail content questions and demographic questions. If you want to look at crosstabulations between the details and demographics, you will only have 10 people in each crosstabulation cell. So, statistically it makes best sense to present the three questions to evetybody. But, if you have 100,000 responses per month, then you have no worries and do not need to have the applet appear each time.

On the other hand, if you want to get overall feedback for the entire site and don't want visitors to be asked about every page, then randomizing appearances across all pages may just work. We have been telling potential clients that they can have up to 20 instances on up to 20 pages per month. The main reason for this is for eBay sellers. The thought is that they may want to run a number of auctions per month and get feedback from several. So, in that sense what you are suggesting makes some sense.

But that's why we are testing, so that innovative ideas can be tried.

Thanks,

Steve

Edited by stever, 10 September 2006 - 12:27 AM.


#35 roger239

roger239

    Honorary Member

  • Members
  • 1 posts

Posted 10 September 2006 - 06:20 PM

This is great stuff! All of you: Thanks.

I understand the resistance to Java applets. I have the same perspective on most plug-in technologies. Nonetheless, Java was selected based on a spec we drew up in the late 90’s. That spec required all the current flexibility noted in earlier replies by Al and Steve and it required backwards compatibility to browsers running the Java 1.08 specification. A lot has changed as the development cycle moved forward. Frankly, Java itself has undergone at least two major changes and any number of minor changes. We keep our eyes on emerging technologies and exploit them where possible, but that practice must be tempered with empathy for legacy implementations. Given the spectrum of technologies representing the Internet, this seems to be a sound practice.

JavaScript was mentioned as an alternative technology. Our problem here is ease of installation and modification. Our spec mentioned earlier, required a simple cut and paste operation easily accomplished by non-programmers. Actually, it went further than that stating knowledge of HTML shouldn’t be mandatory. To this end, we settled on the simple <applet> tag as a mode of delivery rather than the more involved <object> or <embed> tags.

The choices made are not necessarily the ones we must live with forever. Everything changes eventually. The tier driven design of Ask3 is based on the likelihood of change. As we proceed with development and offer new front-ends for varying page requirements, the enhancement will easily fit into the system giving our clients a spectrum of survey delivery options.

#36 bragadocchio

bragadocchio

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 15634 posts

Posted 10 September 2006 - 06:57 PM

Hi Roger,

Welcome to the forums.

JavaScript was mentioned as an alternative technology. Our problem here is ease of installation and modification. Our spec mentioned earlier, required a simple cut and paste operation easily accomplished by non-programmers. Actually, it went further than that stating knowledge of HTML shouldn’t be mandatory. To this end, we settled on the simple <applet> tag as a mode of delivery rather than the more involved <object> or <embed> tags.


I'm trying to recall how may java applets I've seen on the web recently. I would say that the number is pretty small. In contrast, many of the pages I visit everyday have java script upon them. I think that Google has shown, with a pretty wide adoption amongst adsense participants, that even users with limited understanding of html can handle installing a snippet of javascript. Amazon has a pretty widely used affiliates program, and uses the same approach.

Of course, modification of how an ad might be formatted is done on the Google or Amazon site, and it requires replacing the code that may have been previously chosen. But it seems like something that people can handle.

When you get to the point where you're asking someone to insert a line of code into a page, I'm not sure that there's that much of a difference between one one for a java applet, and twenty lines for javascript. So, if it's a question of ease of modification and installation, I'm not sure that a java applet has many advantages over the use of java script.

We have been telling potential clients that they can have up to 20 instances on up to 20 pages per month.


Like Kim, I installed the script on a separate wordpress page, but it's much more likely that I would get considerably more feedback, and more valuable feedback, if the questions were to run on every page of my blog. However, that would mean installing it into the template of the blog, which would mean running it on considerably more than 20 pages. Has the paradigm of how to approach visitors changed with a widespread adoption of content management systems by a fairly wide set of site owners?

I could add a more obvious link to the page where feedback is requested, but how much more feedback would I get if the feedback form fit comfortably within a one of the thin columns on my blog. I would think a lot more. I don't think that I would use it the same way on a non blog site.

#37 Alan Runfeldt

Alan Runfeldt

    Honorary Member

  • Members
  • 32 posts

Posted 11 September 2006 - 12:17 AM

ASK3 APPLET FORMATTING -

how much more feedback would I get if the feedback form fit comfortably within a one of the thin columns on my blog.

Bill -
Ah! I'm glad you asked... :huh: One of the great features of ASK3 is that you can fit it comfortably into one of the thin columns of your blog. That was in the spec and that's just what I did at http://frenchtown.com/#feedback - (I run that one at 250x400) and I think it did increase the response rate as opposed to the default format, placed elsewhere on the site.

Just change the dimensions of the applet tag in your page and then use the Admin Controls (available at the ASK3.com login page - http://www.ask3.com/login.html) to adjust fonts and spacing to make it fit and appear as you want it to. You can drag the preview to match your size and shape, and then select font, sizes and spacing to make the text fit.
- Alan

#38 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 11 September 2006 - 02:30 AM

I have modified my installation to increase my page loading; I put the applet into an iframe element and use a javascript snippet to install that part into my page. By doing that I can change the way the iframe is set up (perhaps add some general information to below the applet) and I can be certain that ask3 doesn't recognize "new" pages on accident.

I insert the following code into all my pages:
<script type="text/javascript" src="/us/feedbacks/showask3.js"></script>

and my /us/feedbacks/showask3.js looks like this:
document.write('<iframe src="/us/feedbacks/" width="400" height="290" border="0" frameborder="0">');
document.write('General user questionaire in Java');
document.write('</iframe>');

then the page I feed into my iframe has more or less just this:
<applet archive="ask3survey.jar" width="390" height="290" 
		code="com.ask3.survey.Survey" 
		codebase="http://www.ask3.com" VIEWASTEXT></applet>

I slightly reduced the width as a test -- it works nicely, but on loading it displays two scroll bars for a split-second.

Feel free to copy, Bill :huh:. I am thinking of using javascript to activate it, that might be neat: "click here for a short questionaire" - js opens a div on top of that page with the questionaire. That would move the java loading to "on demand".

Having had this on my site for a night and a day (not much, I know) I have noticed that it does slow down the page loading / rendering. I would really prefer a javascript version.. I'm still waiting for some good feedback through the questionaire - let's see what else comes up :D

John

#39 stever

stever

    Honorary Member

  • Members
  • 23 posts

Posted 11 September 2006 - 01:43 PM

When you get to the point where you're asking someone to insert a line of code into a page, I'm not sure that there's that much of a difference between one one for a java applet, and twenty lines for javascript. So, if it's a question of ease of modification and installation, I'm not sure that a java applet has many advantages over the use of java script.


This certainly is the case for everyone on this blog, but I still think that the one line is easier for many others, including the decision makers at some potential client companies.

I think that the key is having multiple options.

QUOTE(stever)
We have been telling potential clients that they can have up to 20 instances on up to 20 pages per month.

Like Kim, I installed the script on a separate wordpress page, but it's much more likely that I would get considerably more feedback, and more valuable feedback, if the questions were to run on every page of my blog. However, that would mean installing it into the template of the blog, which would mean running it on considerably more than 20 pages. Has the paradigm of how to approach visitors changed with a widespread adoption of content management systems by a fairly wide set of site owners?


None of this is set in stone yet - nor is it set in code. So, for now, having more than 20 pages is not a problem. Actually I would like your opinion(s) on ways to do this. Clearly a company who wants to have one to three instances of ASK3 on their site has different needs than one that has hundreds of pages produced by a template format. Roger can address the ways we parse the URL, by the way. At this point, I believe that we parse everyting up to the .html (or .asp, etc.) and do not parse anything past the ?.

By the way, in SMART you can pull up all of the results for all of your pages using the same survey and choose any one or all or any combination. So you can see how one page does vs the entire site. Just select the page at the bottom of the Trend Report and CTRL click or SHIFT click to choose more than one. The default will bring up all of your pages at once.

You will have to register each page independently, however, if the distinction between pages is at the html level. This also allows you to select different surveys for each page.

Steve

#40 JohnMu

JohnMu

    Honored One Who Served Moderator Alumni

  • Hall Of Fame
  • 3518 posts

Posted 13 September 2006 - 04:44 AM

Just some short feedback after a short run so far :)

- In the statistics, "Time online" gives me data on "How much did you spend online in the last 6 months?" (in the "Informational" category).

- I'm still not happy *at all* with the load times ... I feel a javascript / iframe solution would be a BIG selling point. It's almost so far that I get the urge to write a javascript solution for my sites using a similar approach (for me privately, no desire to pass it out to others :)). Almost... but perhaps I wouldn't need to ;)

- What I find missing is a feedback on the questionaire itself; perhaps rotated in and out with the other questions: things like 'how relevant are these questions', 'how do you like this questionaire', 'which questions are you missing, which are superflurious?'.. perhaps even allow a 3-value rating of each question?

- The response-rate is something I'd also like to see (mentioned already), it would give me an indirect idea of how well the quesionaire is accepted.

- I'd love a way to remove unnecessary registered page (since I have moved to an iframe for the questionaire, I only have one page). Also, a simple way to reset the results (remove test-responses, etc). Over time they'll (hopefully) get devaluated, but in the beginning it's hard to ignore those "this site is absolutely great" responses I added myself ;)

- The results in the statistics are hard to interpret. Sure "50%" with age might make sense to a computer, but it would be better to keep the average age instead. The same goes for all the other resonses. Perhaps you could internally translate the results and show the corresponding responses instead? I know it makes it look more complicated in a table layout; but as it is now, it looks simple but you have no idea what it actually means. It's "too simplified" by showing just a percentage.

- I would like to remove the exit-link provided by the applet. Or at least minimize it so that it's not standing out like that. I don't really want visitors who just gave me feedback to be transfered to your site with a big button like that -- I have no problem with giving credit to your system, but it would be nicer to have it in the "Adsense-Style", a small link in the bottom or top corner, something like "Survey provided by ask3.com".

- Your sites confuse me :). I don't understand the reasoning between using different domains for the same website. Why do my SMART reports have a header "justASKthem.com" and the URL is "ask3.com"? It could be interpreted as an attempt to do phishing... Perhaps you should submit your back-end site for the Website Hospital as well :). I think there is a lot of potential, but there are a few issues that could be cleaned up.

Hope it helps!
John



RSS Feed

0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users