190% Improvement In Landing Page Conversions By Removing Content | Hacker News Reader

190% Improvement In Landing Page Conversions By Removing Content

themetricsystem.rjmetrics.com

61 pointsjakestein17 comments

After working with landing pages for quite some time, these types of numbers don't mean much to me anymore. Sure, 190% increase in conversion rate is impressive; however, I would argue that posts like this ignore some other major factors such as conversion rates along the funnel.

I'd argue that there is an inverse relationship between the conversion rate of this form and the conversion rate to the next step in your funnel. I'm not saying that either way is better, I'm just arguing that it deserves deeper analysis.

I've worked with optimizing landing pages as well, and I am completely unsurprised, because the content that they removed was a Flash movie. Flash, especially big bulky gaudy presentations, tends to draw attention to itself rather than to what you want the user to do on your page. I would have liked to see a test against another version where the login/registration was inside the Flash (probably poorly for a video that long).

Yes, I agree. I think they could improve their entire funnel conversion rate by adding some simplified "marketing content" back into the page.

As soon as I saw that big Flash movie I thought "Oh shit, am I going to have to watch a long-winded mostly-content-free movie to get any information from this page? Run away!!"

I'm not surprised that removing it turned out to be a good idea.

Yes, I wouldn't want to let the person think themselves out of anything before a discussion can be initiated. Better to drop them as fast as possible into the sales person's lap for an intelligent discussion..

Chris, I agree. We are definitely going to follow this up with more analysis on the conversion rates on the next step of our funnel. We won't have results for that for a while though because: 1)It will be a partially manual process to match up conversations with prospects to demo form completions because the form completer often introduces us to a different person, so we can't do a simple email comparison between our form completions and our CRM 2)Conversations with prospects sometimes don't take place until weeks or even a month or two after the form is filled out.

I wouldn't be surprised if the quality of the leads goes down somewhat, but my guess is that the overall effect will be dramatically positive.

I know this pain first hand. The feedback loop for this kind of analysis is quite long and therefore much harder to actually accomplish.

It would be interesting to see a full-baked analysis on a landing page optimization. Many landing page optimization "experts" tout big gains like these but ignore analysis of the entire funnel and attributing costs of sales-reps time to deal with filtering "low quality" leads. (In a perfect world this isn't an issue because it can be automated, but we all know that isn't the case most of the time.)

Interesting analysis, though, and I would love to hear about your deeper analysis as the data becomes available.

The conversion rate went up, but the quality of these conversions is probably lower.

It's nice that the conversion rate went up, but it would be a mistake to attribute the change to removing content, when there were many improvements to the overall design of the landing page as well.

I would say this is a lesson in the value of thoughtful design much more than in simply removing unnecessary content.

Indeed, they confounded variables instead of changing exactly one in the A/B test. http://en.wikipedia.org/wiki/Confound

"Word economy" is the highest calling in my experience with landing pages. If something can be said in bullet points intstead of paragraphs, use the bullets.

Do you have data to back that assertion up?

2.1% -> 6.1% is a 290% improvement, not 190%.

Don't cut yourself short ;-)

No, this would usually be called a 190% improvement. (6.1 - 2.1) / 2.1 = 1.9

Similarly, 2.10 to 2.31 is usually called a 10% improvement, not a 110% improvement.

Are you saying 2.1% -> 2.1% (no change) would be a 100% improvement, and 2.1% -> 0.21% is a 10% improvement?

290%

190% improvement

but not

290% improvement

You guys are right. I confused myself with the wording and thought too much about it.

Thank you for correcting me.

Reply on news.ycombinator.com