Watching your visitors

[dickbaker]

With all of the discussion about user engagement factors, I decided to search for a service that would allow viewing visitors on my site. I found one with a free trial, allowing me to view 100 visitors. It was a humbling experience, to say the least.

It was also educational, as I could see where users were having difficulty figuring out things, or trying to figure out how to click on something. I found two or three things about my site that I need to change immediately.

Then I could see the thin content, and how users came, scrolled down a few lines, and hit the back button. Not good.

None of the first 100 visitors in the free trial went to any of the pages where I've added things such as videos, performance charts, model XYZ history, or other beefed-up content. I'll gladly pay to subscribe to see how visitors are interacting with the changes I've made.

When Panda first hit, I was huffing about how good my site was. Now, watching visitors actually on the site, I can see that I have a long way to go to hold many of them who are now leaving too soon.

[tedster]

I know how humbling this kind of experience can be. It's something like finding out your child really is [ugly, stupid, cruel, clumsy, etc] despite your parental love and naturally biased viewpoint. This is why I so value the expensive reports published by Jakob Nielsen of Nielsen Norman Group. It's not just his opinion, it's the results of actual user testing. Invaluable guidance.

I bought E-Commerce User Experience a few years ago for $450. Paid for itself almost instantly, once I got past my excuses for doing things that he advises against.

[Planet13]

Then I could see the thin content, and how users came, scrolled down a few lines, and hit the back button. Not good.

Are you saying that they went to the page and, without actually reading the text, determined immediately that it was a thin page just by gauging the amount of text on the page, and then left without reading it?

Or do you mean that there were only a few lines on the page in the first place, so when they "scrolled down a few lines" they had finished reading all the content?

[Shatner]

Interesting timing since I've recently started using a service which allows me to do the same thing. In my case the service lets me see all my users in real time, so I'm not just looking at a limited number.

My experience has been the opposite of yours. I got it hoping to find problems, to explain why I am pandalized. Instead what I found was that users were doing pretty much exactly what I thought they were doing, using the site well, scrolling down on pages often, clicking where I expected.

I guess it's nice to have the confirmation but I really would LIKE to find flaws. At least then I'd have a direction to take.

Maybe I need to find a different service, and this one just isn't giving me the right info.

Any suggestions on good tools to use for this? What are you using dickbaker (assuming we're allowed to say here)

[lucy24]

Are you saying that they went to the page and, without actually reading the text, determined immediately that it was a thin page just by gauging the amount of text on the page, and then left without reading it?

Or do you mean that there were only a few lines on the page in the first place, so when they "scrolled down a few lines" they had finished reading all the content?

It means that when people google "what does 'my hovercraft is full of eels' mean" they're looking for a page that says at the very top, in 36-point type, "It means 'please give me a box of matches'." If it isn't there, they leave without ever reading your masterful treatise on new fishing techniques.

... And if it is there, the sentence will be right on the SERP and they'll never go to your page at all. There is a moral in here somewhere.

It means that the relationship between what google thinks is the best match for a given search string, and what people really want when entering the said search, is ... umm ... unpredictable. So the problem may not be in your page at all. Unless you want to take the effort to sabotage your own search results so people looking for information on X don't blunder onto your page just because you've got a series of words that looks promising to the computer.

[Robert Charlton]

Shatner - Just to jump in here, we don't allow specific tool discussions in the public area. From our
Forum Charter [webmasterworld.com]...

We request that you not mention or post links to any specific "Google tools" or "SEO tools". There are hundreds of them, and many are owned or operated by our members. Basic webmastering tools that are in general use and not associated with a particular SEO company can be mentioned, when appropriate to the discussion.

[dickbaker]

Are you saying that they went to the page and, without actually reading the text, determined immediately that it was a thin page just by gauging the amount of text on the page, and then left without reading it?

Or do you mean that there were only a few lines on the page in the first place, so when they "scrolled down a few lines" they had finished reading all the content?

No, they read the text, although many read it quickly, others took awhile (maybe they're slow readers, or maybe they were trying to take in all of the limited information). There were a couple of visitors who quickly scrolled up and down the page, then left. Maybe they were looking for widgets to buy and, since I don't sell these types of widgets (just provide photo, spec's and description), they left right away.

I could see some users who were uncertain as to what to do, hovering over different type to see if it's clickable. In my pages that list retail stores in particular states, I have a red star next to the store's name for people to click to see the store's ad. It was obvious that some visitors didn't understand that they had to click the red star, so I'll have to make a hyperlink on the store's name.

There's other things I found I need to change, too.

[Leosghost]

I suggested how you can do this "watching" waaay back in a panda thread..just set up your one or two person "research unit".. physically somewhere ( like rent an office for a day or two ) ..get/ hire someone to waylay passing strangers ..and offer them $10.oo for 10 minutes of their time ..make it look good ..dont make it look like the scientologists etc...

Explain it is to see how people interact with websites..

Sit them down in front of a screen with sites available including your site ( don't tell them its yours ..and some others that haven't fallen in serps for what you have as KW(s) etc ( as ehow is often sighted ..use it as a baseline/ control ) ..
give them a printed page with 10 simple tasks to do on the page ( if you have a shopping site you'll have to allow for not using a real credit card etc ..you work it out ..OK ;-)..they need to do them ( or try to do them ) on each site ..

Then watch them ..shut up, don't prompt, don't sigh with frustration etc etc ..don't encourage ..just watch ..( you can set up the watching via another machine ..thats what the kind of service dickbaker used usually does) ..or via a "two way" mirror ..which is how lots of ad agencies do it ..or both ..or whatever..

That is basically how it is done and has always been done ..for decades ..

Works 99% of the time ..I suspect with Tedster's background in retail ( standard procedure in large ad agencies ) that he will have seen it in action on varying scales too in real life ..

You can get software ..or sign up for heat map click tracking ..or real time like dickbaker ..but real life observation beats all.

btw..its deductible ;-))

for around $1000.oo or so ..priceless.

[netmeg]

There's a service where you pay around $35 for a usability test; you set some action for the user to do (can be searching in Google for one of your products, completing an order on your site, whatever) and you get a video of the screen action as well as his running commentary. VERY useful. I had one of my clients sign up for it, and he was amazed at how that navigation system on which he was so insistent confused and befuddled the user. The client said to me "I kept wanting to break in and nudge him over to the right place!" Can't emphasize enough how valuable this kind of feedback is.

[Planet13]

There's a service where you pay around $35 for a usability test;

Thanks for the heads up. A $35 usability test sounds like a STEAL to me.

[dickbaker]

The trial of this service gives you views of the most recent 100 visitors. After that, you buy visitor views at a cost of five cents per visitor.

Leosghost, I remember your suggestion about that months ago. It seemed like a good idea. This is more impartial, since the visitor doesn't know he's being watched. What's more, it's easier to implement.

...it's deductible

Sorry, but I'd rather have the entire dollar in my pocket than the 25 to 30 cents off my taxes.

[Leosghost]

That sounds like a pretty good deal,( your sticky mail will be busy ;-)) ..If the visitor doesn't know they are being observed then I presume its "live screen" or "screen tracking" but without camera ?.

There are services with camera and/ or commentary like the ones netmeg has used..all very very useful.

I used to organise this kind of thing when I worked in the ad business ..you get into the habit of observing people very very carefully afterwards, every day in real life, without them being aware of it ..plus very detailed observation is something that most artists do naturally..musicians are the same with sounds , actors blend it all, to some degree.

A lot of psychology is involved in design / advertising ( and on website interfaces ..we are talking design for usability as opposed to design for aesthetics only ) ..well worth the time , the money and the study , do enough of it and you can see what "works" and what doesn't nearly instantly on anything, and where the adjustments need to be, ..

Do a lot of it you may get good enough to be a consultant ..

The money is good ;-))

[johnhh]

The one I tried was 5 cents per page visited with a trial of 100 pages. Which doesn't sound much but even post panda it would be hundreds of dollars a day.

It was clever and interesting to watch, amazing the range of screen resolutions used.

But you need to add code to every page/template.

[dickbaker]

johnhh, it sounds like the tool you tried is the very same one. I had to install a few lines of javascript on all of my pages. I don't know how it's done, but it's similar to a service I used when I was doing SEO work for a hosting company, and we needed to demo our pitch to them in real time showing their website. It's like having a camera pointed at the user's monitor.

I have a lot of ideas for beefing up my content. With this I can see how visitors are interacting with one of the ideas on a small scale before spending time and money going full-bore.

[Play_Bach]

A few years ago, I made $5 along with a friend for watching a 15 minute commercial and answering a short survey about it at the local mall. I remember the pitch as we walked by the store went something like, "want to make $5?". Worked.

[tangor]

No, they read the text, although many read it quickly, others took awhile (maybe they're slow readers, or maybe they were trying to take in all of the limited information). There were a couple of visitors who quickly scrolled up and down the page, then left.

Not picking on this comment or the poster, but seems most appropriate for my remark that if one is in a competitive niche and is offering nothing REALLY NEW (similar re-hash regarding red widgets) as regards other sites, that back (your lost buck) button is highly exercised.

What engagements can be made in page display, content, or presentation to get the visitor to NOT click away? That's the burning question for every site out there... and the one that needs to be addressed. And how does any of the above factor into Panda?

[dickbaker]

What engagements can be made in page display, content, or presentation to get the visitor to NOT click away? That's the burning question for every site out there... and the one that needs to be addressed. And how does any of the above factor into Panda?

Yes, that is the burning question. As I watch visitors look at a page and leave, I search in vain for ideas for features that would keep them on the page. I have some, but all of them would require an enormous investment of time and money to implement over a couple of thousand pages. I may try one and see how it does.

How does any of the above factor into Panda? If what Brett Tabke and others are talking about in terms of user engagement, etc, is true, then this "spying" on visitors is a window into part of what Panda is evaluating.

[Asia_Expat]

I've only scanned this thread, so apologies if this was mentioned already... but there is an open source alternative to Google Analytics called 'Piwik'. One of the plugins is a live feed of your site's visitors. You can watch them and what they're doing in real time, or from the archive.