Analysis of 50+ requests for my "Offer HN" conversion rate optimization help | Hacker News Reader

Analysis of 50+ requests for my "Offer HN" conversion rate optimization help

visualwebsiteoptimizer.com

177 pointsparaschopra39 comments

This is the kind of summary that complements "Offer HN" posts so well. I totally missed your original thread, something I would have taken advantage of if I saw it. By the time I might have seen it, the service would be long gone. Even in that case where I missed the opportunity to get a custom assessment, here is a summary that I or thousands of future visitors can reach. The result is persistent value rather than a one off. I hope future "Offer HN" services do the same.

Regarding the "your greatest enemy is the back button" quote-- the earliest citations on HN are http://news.ycombinator.com/item?id=1136272 and http://news.ycombinator.com/item?id=1136272, attributing it to "one of the founders of BackType".

Personally, I think Patio11 put a nice twist on it here, though: http://news.ycombinator.com/item?id=1686784

I personally thought PG made it but couldn't find a citation.

It could be, but I couldn't find it either.

Or, I should say, Google couldn't find it.

IIRC he mentioned that in his Mixergy interview.

"Obscurity is a far greater threat to authors and creative artists than piracy." - Tim O'Reilly, 2006. http://radar.oreilly.com/2006/08/piracy-is-progressive-taxat...

PG at Startup School, 2006:

The median visitor will arrive with their finger poised on the Back button. Think about your own experience: most links you follow lead to something lame. Anyone who has used the web for more than a couple weeks has been trained to click on Back after following a link. So your site has to say "Wait! Don't click on Back. This site isn't lame. Look at this, for example."

http://www.paulgraham.com/startuplessons.html

I'm not sure how it got its current epigrammatic form, but it's surely become more succinct in the course of being repeated to hundreds of YC companies since then.

Cut text; reduce choices; only offer one CTA button.

It's fascinating how an Occam's Razor-type force pervades successful web design.

As always, test. I'm broadly supportive of this advice, but I know people who do long-form copy and see conversion rates slip when they switch to short form, and (while it isn't long-form) the BCC home page historically performs better with lots of text than with little.

Yes, one example of those kinds of pages are the never ending sales pitch pages (e.g. selling parrot training). So, text _per se_ is not a bad thing on a page. But you have to provide proper flow and layout to generate interest and then seduce visitor to that prominent 'Add to Cart' button.

That was my initial impression too (what about SEO?). It's about balancing the text with the layout and clear calls to action.

When people have only one thing to process mentally and click on, guess what? They click it.

In an extreme example[1], when a service failure resulted in a mobile app displaying only ads (no content) -- revenue soared.

[1] http://news.ycombinator.com/item?id=1753134

Here is the original "Offer HN" thread: http://news.ycombinator.com/item?id=1842768

is the "too much text" really a problem? if too much text didn't work, none of those "buy my crappy ebook and make billions like me" sales pages would work.

regarding the designer, if you are just starting out and can't afford one...borrow concepts from your competitors. If you are selling shoes, borrow the design from zappos. Why? Because you know that they've optimized every single thing, since even .5% can cost them millions.

I'm pretty sure that quote is from pg.

And if you want to be a real jackass, you can add an exit pop to your site.

The call to action thing is very important. Most users are very uncertain, if you give them too many options, they'll get confused and won't do anything.

As far as social proof, yes I agree, but it's not that hard to fake the social proof. Nothing is stopping you from adding a Techcrunch image to your site. And if techcrunch bitches...just point them to that completely unrelated comment you made with a link to your site.

overall good post, since I find that pretty much every website will have one of those missing.

Yes "too much text" may be a problem is if it unstructured. Also, if you don't have a designer you can buy one nice-looking template from http://themeforest.net

I would personally never "fake" social proof as (apart from ethical concerns) it may have legal liabilities which a startup cannot afford.

EDIT: Typo in link

I'm not sure about the legal liabilities.

if you add "as seen on Techcrunch", and that's in reference to a comment you made...that's not exactly a complete lie.

yes it's completely unethical, but I don't think it'd be illegal

I think that link should be http://themeforest.net/ ?

"if too much text didn't work, none of those "buy my crappy ebook and make billions like me" sales pages would work."

You're referring to 'long copy' which is wholly different from 'throw a couple of paragraphs of text on my website because I'm too lazy to make it short or replace with graphics'. Writing long copy is a craft, and is structured completely different from the websites in the review.

For one data point, http://ourdoings.com/ got the "too much text" criticism from Paras, but that ugly wall of sentence fragments actually increased conversions relative to the few carefully crafted sentences that were there before.

For more data points, the post that prompted me to uglify my front page in this manner was Don Norman's: http://news.ycombinator.com/item?id=1659020

That said, I'm grateful to Paras. A couple of the feedback items are ones I expect to be very helpful, though I haven't tested them yet.

Yeah, my advice was based on what I thought could work best. But the best idea is to do an A/B test (and not implement best practices directly). I'm glad you liked my feedback.

>> is the "too much text" really a problem?

Yes. Even though I read the entire article, the captions for the graphics combined with the graphics themselves provided such a clear, concise summary of good and bad examples for each bullet point that I immediately understood the meat of the detailed text surrounding it.

The graphics and accompanying captions in the article itself serve as a nice TL;DR which reinforces the author's earlier point of one way of fixing the "Too much text" problem:

...Instead of extensive “How this works” (consisting of heaps of text), make a simple graphic detailing the process...

Yes. Look up a reference to "no one wants to read your shit" in the google. And it is so true, no one reads your website (or mine either). You are fooling yourself if you think otherwise (at least for getting that initial conversion).

I always liked the Git website for these reasons. The headline states that Git is a "fast version control system". There is a an example of the commands to show how easy it is to get started. There's a list projects that use Git. And finally, there is a prominent download section.

http://git-scm.com/

It's interesting that you chose to show Skype as a good home page. I happened to visit it today because I had to install Skype on a new computer. They make you jump through a lot of hoops to download the software and I actually pointed it out to a coworker.

Any thoughts on that?

Great analysis, thank you for this.

One big question - what kind of social proof can a new site offer? Without being unethical, of course. Obviously, we have no customers yet, so we can't tout any huge customers or even media attention without flat-out lying.

As I wrote in the blog post, if you are new you have to rely on ROI proof instead. You can perhaps cite a scientific study (in your area) or make up a reasonable theoretical model. Then say something like "78% companies who did A/B testing in 2010 increased their conversion by 50% or more"

So, you pitch the general concept of your startup when you are new (and don't have social proof). Later you pitch your actual product.

Regarding the "Who's using" stuff. Maybe I'm not your typical person but I could care less about who is using your stuff. "Featured in the NYTimes..." oh yeah? So what.

Like I said, maybe I'm wrong but do you have numbers to back it up?

http://en.wikipedia.org/wiki/Asch_conformity_experiments

Another really good reference is Scientific Advertising ( http://www.abenetis.com/encyclopedia-article-59.html )

You are the exception, not the rule. I choose a product first by it's merits and second by what people have said about the product on their reviews or comments. Social proof can be of extreme benefit because of this.

Which plugin are you using for that "More In" hover text? I have seen a lot of popular websites use it, and its actually quite neat.

No need for any plugin. It's fairly easy to do it yourself:

http://jsfiddle.net/treeface/Z63Yt/2/

It is actually this plugin: http://wordpress.org/extend/plugins/upprev-nytimes-style-nex...

Thanks to both of you.

Off topic here, but I find these rather obnoxious. They're almost like mystery meat navigation in that they're state isn't consistent. You have a disappearing navigation element, so a user might see it at one point, misunderstand what causes it to appear, and then not be able to find something on the page that they saw before.

On the other hand, they do grab the users attention, but I wonder what the ratio of successful clicks is to frustrated users.

Just my two cents, and pointing out that at least some users don't think they're neat.

A news site I read has them. I see it pop up while I'm reading an article, think one of the headlines sounds interesting but then can't find it when I've finished the original article and I'm ready to read something else.

The website seems to be down, anybody got a cached copy?

Sorry, should be back up.

Excellent, helpful, understandable - thankyou.

Reply on news.ycombinator.com