- By calcoastwebdesign
- 1 Comments
- 20 Aug 2008
Tim Ash is the only speaker on this panel. Who did he have to hack to swing that? Kudos though. I didn’t feel like like Angie’s laptop all day….plugged. His is energy is infectious. I really enjoyed the track.
“Good Morning.” he shouts. I was feeling good today. Wasn’t hiding in the back not talking to anyone this time. I promptly replied back with a “good morning” myself. No one else, just me. Not loud enough though. “Once more time folks” he demands. Finally, a few less drowsy voices pipe in. Guy in the front row gets $20. Damn. Shoulda gone even closer. Damn you 20/20 vision. LOL.
He then jokes about his yelling. Mentioned he had to get louder than Sage. LOL. I think I found a niche I can piggy back on…the loudest female SEO. LOL. If Sage can do it, SO CAN I!
How many of you have any ugly baby? Ok, ok, How many of you have pages on your site that your users don’t respond to.? They aren’t getting the conversions you want??
If you are getting traffic but not leads, you have an ugly baby.
Acquisition – Get users to your site.
Conversion – Get users to do what you want them to do.
Retention – Increase the lifetime value of the relationship (send to a friend, etc.)
Why Should You Care?
STOP FOCUSING ON THE TRAFFIC – Stop aiming a power hose at a keyhole, just open the door.
CPA=CPC/CR – Cost per acquisition = Cost per click / Conversion rate
If you are not improving conversion as fast as a CPC rate increases, you will never be able to keep up.
Quick question…Who should design your site?
You Users or website visitors of course.
- Size of the A/B Testing
- Test Configuration restricted vs freeform – (headline, CTA, ect)
- Data Collection (random vs purpose)
Quick FREE and easy way to do this is with Google Website Optimizer (In walks Google rep…creepy….Google really can read our minds!)
- Even simple changes have a large impact
- Leverage variables for interaction efforts. Fractional analysis would have failed.
- A/B Split Test (Must have at least 10 conversions/day)
- Test multiple variables at 1 time
- Sends equal traffic to all versions
- Simple to implement and track
Parametric Multivariate Testing & Non Parametric Tuning were mentioned briefly for larger scaled sites that don’t pertain to CCWD as of yet.
Avoiding The Pitfall – 7 Deadly Sins:
1. Awareness Squandered
- Focus on user goals
- Stop Screaming
- Eliminate clicks
- Unclutter the remaining
- Does your landing page have too many places to click? Are there too many options? Is there too much going on? I get overwhelmed with The Cheesecake Factory menu. Imagine what a disorganized site can do for you.
2. Don’t Frustrate Users
- Ever seen this… “Sorry, No results Found” and then immediately following they request your info for a lead capture. The alt for this particular CTA should just simply ask, “Would you like some spam with that?”
- Guide people safely to their goal
- Dont make people feel stupid
- Dont paint users into a corner
3. Invisible Risk Reducer
- Lower User Anxiety!
- Use tools, icons and other content to put a users safety first.
- Make your “HackerSafe” and “Verisign” logos prominent
- Keep it above the scroll
- Use it on all pages
- A buyer needs to know your safe before they consider finishing the buying process.
4. Lack of Client Proof
- Trust factors like big name clients help a user feel better too
- Lets user know its SAFE to fill out that form
- Ensure you list client and media coverage.
- Thats right, add a link to FOX news if you’ve had a segment featured.
- Transfer trust using larger brands and existing large clients.
Jenn when you read this, start getting big name logos for GSMU to ad to site. LOL.
He tells us, “It’s easier to ask for forgiveness than permission.”
Angie, get out of his head. LOL.
5. Transaction Interruption
- “Would you like to buy something completely unrelated.” No. If I wanted to , I would have searched for it. Thanks.
- Make sure your users can bypass that BS.
- Stop surprising and confusing your clients with this.
- No one likes an online surprise (unless its money)
- Do not interrupt people unnecessarily.
- Don’t make people feel stupid
- Identify user roles
6. Ignoring Your Baseline
- Always devote bandwidth to your current version or baseline
- Measure relative to your baseline, not optimal performance.
Ok, I swear there were 7. I will go back and review the online version once available. LOL!
Time for some audience questions…
What if your pages aren’t dynamic and you have a large site? Ooooh. Good question.
Good luck. LOL. If this is truly the case, you are going to have a lot more problems that just testing and tuning. In this particular case, all you can do it test the overall structure of a page rather than actual content.
Likewise, a similar question was asked of sites having less than 10 conversions a day. It cases like these, you simply use an industry expert to make recommendations.
Ash mentions that his company will soon have a FREE version of their own tuning software available online in a few months. Looking forward to it!
Finally, a class member requests a basic checklist of the basic elements of testing…
- Logo & Content
- Single headline
- 3-4 billet points (not text) to support your proposition
- CTA like a form with its own title
- Trusting credibility on the side
- Whatever you do, do not take them back to your corporate site to get lost, confused and overwhelmed and leave.
With those of your with funnel processes rather than landing pages, you must be testing and tuning your entire funnel and making micro sites for the landing process.
Looking forward to playing with your lovely tools Google!