[Split Test Results] How To Increase Conversion Rates By 8.47% Without Changing Your Website

Important Note: The percentage amounts mentioned in this video reflect relative (not absolute) increases or decreases in conversion rate.  For example, if the conversion rate on a given page increases from 25% to 50%, this would be considered a 100% increase in conversions (since the conversion rate doubled).

Hello everyone, Clay Collins here, and today’s data for this episode is brought to you by Juan Martitegui. Juan is a LeadPages customer, one of my favorite entrepreneurs online right now. You can find him at juanmartitegui.com. He’s the CEO of MindValley Hispano, a seven-figure information product publishing company serving the Spanish speaking market, and they’re just absolutely killing it so I just want to send a shout out to Juan for donating the data for today’s episode. Juan recently said this about LeadPages. He said these are giving me a 3X to 4X conversion against the ones I was using before. A couple of tweaks in my biz will make at least 2X the money I was making. So thanks to Juan.

So Juan recently sent in this split testing data. So here it is, and we’re going to break down exactly what this data is, and what it means, and why this is so important to you. So this data shows these pages split test against each other. So the first contestant in this split test is this landing page. All these landing pages were created in LeadPages. This one, as you can see, it hosted on leadpages.net. All of our pages are hosted on the Google server network so they’re really fast so this is the first one he tested. He tested against this one where he downloaded the page and put it on his own server, and he was testing to see whether or not the difference in server speed affected conversion, and it did. We’re going to talk about that. And then he also tested the presence of the first name field, right, so he tested this page against this page, and this one, he’s only asking for the email address, and in this one, he’s asking for the first name and the email address. So we’re going to breakdown what happened here and the statistically significant results that Juan got during his split test.

He found that the difference between asking for the first name and not asking for it is a 5.25% difference in conversion. In other words, when he asked for the first name, conversions went down 5.25%. Now this might be a big deal to some people. It might not be to others. Over the scope of a multimillion dollar business, this can have a real impact. So I’m not saying what you should or should not do here. I will say that there is what I consider to be a notable difference. Juan said that he didn’t find this to be a big difference according to him, and so, it seems like he’s going to continue to ask for the first name field. He also said that he downloaded the page and hosted it on his own server. He wrote that the difference was huge. He said your hosting is a lot faster, okay, so the hosting that people get for free with LeadPages is a lot faster.

Again, we’re using the Google server network to host all of our pages so there are servers in every major continent, and there’s incredible caching in place, and it’s always going to serve the closest file, and it’s, you know, the pages themselves are very highly optimized, but he said your hosting is a lot faster, hence, conversion rates improve. On the server with the same page, I got an 8.47% improvement. Almost 10%. And something to note here is that Juan is a total pro. He doesn’t screw around with hosting. In fact, he spends over $1,000 per month in hosting. He’s got best of the best hosting, but frankly, it’s hard to beat Google’s server network.

So the big takeaway here is not necessarily about first name field versus no first name field. It’s really that your landing pages should be fast. In many cases, you can get up to a 10% improvement just by having faster pages. This is not only important for conversion reasons, but this is important for SCO reasons. Google, back in 2010, actually came out and just said that one of the factors they used to rank their pages is one of the signals that they use to determine which pages should appear at the top of the search engine results page, and like I said, this is also important for conversion rates. In one study, 40% of people abandon a website that took more than three seconds to load, and as you can see, as page load time increases, conversion rates go down. In fact, often, a 1-second delay in page response can result in a 7% reduction in conversion. So with each additional second that it takes for your page to load, conversions go down roughly 7%.

Let’s put this into perspective. If an e-commerce site is making $100,000 per day, a 1-second page delay could potentially cause $2.5 million in lost sales every year. If you are a LeadPages customer and you want to compare the speed of the page hosted on your server versus the speed of the page on our server, you can go to whichloadsfaster.com. So we’re actually going to go to whichloadsfaster.com and run Juan’s pages on his server against ours just to kind of drive the point home, and you can do this as well. So we’re going to take

the version of the page that he has on his own server and run that against the page that’s at leadpages.net, and I’m going to click on go, and as you can see, our page was 15% faster, and again, to run this kind of comparison, you can just go to whichloadsfaster.com.

Anyway, my name is Clay Collins. I hope you found this information valuable and helpful. If you want to follow Juan Martitegui, I’ll link up to his info below. He’s a great guy. Thank you personally for me to Juan, and thank you to everyone who’s watching this episode. Take care. Bye.

[0:05:32] End of Audio

  • imran md ali

    Hi Clay- Good point about speed-

    The Facebook app landing pages though, are a teeny weeny bit slower to load….

    • Hi Imran… great point.

      The facebook pages are a tad slower due to constraints on Facebook’s end. We have no control over that.

      Warm regards,

  • Clay,
    Thanks for the data.
    I want first names though as I want to be able to personalize my e-mails to my users. So I always ask fo the first names. If someone can’t give me their first name, then I’m not interested in having them on my mailing list.

    Well your pages should be fast, not just your landing pages :-).

    • Hi Kerwin,

      Thanks for your comments here.

      One point I’d like to make is that, generally speaking, the large group aggregate data bear out that populations don’t respond to the use of names in emails (and that people don’t like it).

      See here:

      –Hey startups: Don’t start your emails with “Dear [Name]” – 95% of people hate it

      –Subject line personalization depresses open rate:

      So not only does collecting the first name decrease opt-in rate, but it also decreases email performance. It’s a “lose-lose.”

      Still, we support the collection of first names because everyone seems to want this. So there’s a lot of people who think like you do.

      Warm regards,

      • Strongly agree. Everyone realizes the emails are not actually personal an instead get the feeling that you’re trying to trick them. Be transparent and increase conversions, win win.

  • Awesome Awesome Awesome Clay! Thanks!

  • Thanks to Juan too! We hung out at AFEST

  • This was very interesting. I checked the loading time of one of our LeadPages using the leadpages.net vs. our own domain (I had added the page through WordPress) and the difference was 9x – 12x faster with LeadPages. Looks like I need to switch out some links next week!

    • 9x is a big deal. Thanks for sharing your numbers, Andy.

      Would love to see any results that ensue after changing your urls out.


    • Using whichloadsfaster.com that’s about what I’ve seen as well….HUGE difference using the WP page vs. the leadpages.net page. Don’t want to blow up the comment feed here too much…but see my comment above it seems like we could get the best of both worlds with insane speed and our own domain via dns pointing to your servers….

  • Interesting. How does this compare with pages loaded through the WP plug in?

    • LeadPages published with the WP plugin will likely be much faster than your other WP pages. But they will ultimately be slower (maybe just a little slower) than pages published at leadpages.net; I’d use http://whichloadsfaster.com to compare. Would love to see your results.


  • Bill

    Great info Clay. I knew page load time was important but not to the extent you talked about. Thanks.

  • Your hosting is 20% faster than mine. I chose mine because it seemed easier to track. I need to rethink that…

  • Matthew James

    Hey Clay, what do you have for hosting for leadpages? I own a site that gets a lot of traffic, site speed is not as fast as yours. Not all pages are leadpages, so I can’t let you host all my pages, LOL. Would love your recommendation on faster hosting. Thanks bud!

    • Thanks. Matthew. Our hosting at LeadPages.net is on Google’s server network . . . so that’s my recommendation. It’s pretty fast.

  • wasn’t aware how much time – how fast page loads – affects Search engine results. And didn’t realize how much 1 second can actually impact in such a big financial way
    Thanks for the info Clay!

  • Speed on the App Engine servers & the leadpages.net urls is so awesome.

    Here’s the thing that would take it to a whole new level of awesomeness: set it up so that we can use dns to point our domain/subdomain to our leadpages.net subdomain. That way…we get the amazing speed, AND get to use our own branded domain!

    I submitted this as a feature request…people if you’re a leadpages customer (clearly you should be since it rocks)…then go here to vote this one up so we can get the awesome speed and use our own domain!….


    • Clay – would love to hear your thoughts on this…it might be one of those things a lot of people haven’t thought about and therefore might not go out of their way to request…but that doesn’t mean it wouldn’t be super awesome for tons of LP customers!

      • Ryan

        I agree