98: TZ Interview – Lance Jones / Conversion Optimization

Justin and Jason speak with conversion expert Lance Jones about the three fundamental questions that need to be asked to optimize conversion and the tools required to answer those questions.

  1. Tyler Link says:

    FYI, Avinash Kaushik’s blog is titled Occam’s Razor but resides at kaushik.net not occamsrazor.com as stated in the podcast. I used to work for him at Market Motive. Nice guy. Very entertaining.

  2. AK says:

    @Lance, regarding writers showing off their pages: why not have a separate version of each page that cannot be voted on? It would be linked from the author’s homepage and would contain the number of votes it got.

  3. @Tyler, great catch. Shows how much I actually look at the browser address bar. 🙂

    @AK, we currently give writers the ability to link directly to their feedback page, but it’s only feedback — not the page 99 content itself. I agree with you (and Justin & Jason) that we should probably just open it up and let writers invite their own contacts to anonymously vote on their page — even at the risk of introducing some bias. Sometimes I wonder why we (Page99Test co-founders) continue to try to police such things when letting go would probably result in higher satisfaction of our users and more traffic.

  4. Neville says:

    Home run guys, amazing wealth of info in this show!

  5. Neville says:

    @Justin you do a great job editing, I’m glad you keep the out-takes at the end of the show, they are a real pleasure.
    @Jason, I think you should do a push-ups for every day AppIgnite is not in private beta. 😉

  6. Justin says:

    @Neville – Thx for the kind words and also thanks for slagging Jason

  7. Jason says:

    @Neville – Glad you enjoyed the show. Also, I hereby declare that each and every day I will do the same number of push-ups that Appignite is late (days since the new year). That means today I owe (and will do) exactly 7 push-ups. 😉

  8. Justin says:

    @Jason – That’s more than I can do

  9. Andrew says:

    Jason has competition as I’ll soon be releasing my own automated-application-factory. My product also has a trisyallabic name – ‘vaporWARE’, it’ll be out in two weeks I promise!

    On a more serious note, I remember a post on HN from a user named ordinaryman who released a tool in the same domain.

  10. Justin says:

    @Andrew – Wow. having a good look around that tool, and the complexity of it, I think it might be another year before neverignite releases…

  11. William says:

    Hi guys–an embarrassing question about your sound cloud link–I tried clicking on the button and nothing happens. I tried it on IE8, firefox, and Chrome beta. Does it require flash cookies enabled? I finally just downloaded the mp3…


  12. William says:

    I’m glad I downloaded this interview–I can see myself listening to it again. I do wonder if some of these optimization tools lead to local minima…For example, maybe something you do to optimize getting someone from page 2 to page 3 deters people who would eventually have bought. Also, how long do these tools track individuals? For example, maybe someone visits, then leaves, then comes back, then tries the product, then leaves, then finally buys….

  13. Jason says:

    @William – If I remember correctly, I think Performable might do that.

  14. @William, glad you enjoyed the discussion. Regarding your first question about doing something in a test to deter someone from buying… it’s certainly possible, but if your overall conversion rate goes up 20%, would it really matter?

    For repeat visits, a user will continue to see the same version of a page when you’re running a test. So if your test remains open for, say, 3 weeks (which is pretty typical), any repeat visits by people will be accounted for and reflected in your test results.

    Great questions!

  15. Excellent information-rich show!

  16. Since the content was pretty crucial for all of us I decided to write some of the key things down. I tried to post the result in the comments but somehow maybe due to the size or the links the post would always fail.

    so I ended up writing a blog post titled “Improving Web Conversions In 3 Key Steps”!

    Thanks again to Lance and you guys for the great content.

  17. @Justin and @Jason I suggest you to check out SaasPulse, one of the best analytics tools I’ve ever seen.

    I’m lucky enough to work with it, and I’m really thrilled about it.

  18. “@Neville – Thx for the kind words and also thanks for slagging Jason”


  19. @Udi, I had also recently heard about SaasPulse, but I see they’re still in private beta. Any ideas on when it’ll actually be available to try?

  20. @Lance_Jones I don’t know when they launch, I’m just one of the beta customers. Send me your email here: udiudi.com/contact and I’ll see if I can help you to get an invite.

  21. Yeti says:

    Very interesting show, thanks guys.

    Just small point, Google Analytics does support Events but never used it so cannot compare to MixPanel – just thought it worth mentioning

  22. Justin says:

    @Yeti – Yup. I think I’m using that by directly logging javascript ajax calls. Not sure what it gets me though… just, now, all my top pages are ajax polling calls!

  23. Very good interview, thank you very much Lance, Justin and Jason. So much useful information! And thank you Philippe for the blog post, I was planning to listen again with a notepad and now I don’t have to 😉

    Lance, I have couple of questions if you don’t mind.

    1) What do you think about KISSmetrics? You didn’t mention it on the show and I was wondering if you have any experience with it.

    2) We’ll be implementing dynamic home page that shows different content based on the traffic source (and search keywords if from a search engine) and type of visitor (first time or repeat visitor). Do A/B testing tools work in this scenario? We would like to optimize home page for each visitor group separately because they are quite different.

  24. Tyler D says:

    Great episode! Website optimization is a really interesting science, and I think a lot can be learned about good web design based on the data it gives you.

    One product I’m surprised didn’t get mentioned is Monetate. Really feature rich for getting conversions, doing A/B testing… even allows you to “geotarget”, e.g. showing the Golden Gate Bridge to San Francisco users, and showing the Liberty Bell to Philadelphia customers.

    The interface isn’t as clean as some of the alternatives, but you can do a lot with it.

  25. @Michael, to answer your two questions:

    1. Yes, I absolutely love KISSmetrics tools… especially KISSinsights. Love it, love it.

    2. That’s a great question. I know that Omniture Site Catalyst will enable you to test based on several user ‘attributes’ (e.g., operating system). And Visual Website Optimizer will let you segment tests by new or returning visitor. This seems to be a common request, so I assume most of these tools are starting to include segmentation options.

    Hope that helps!!

  26. Correction above… Omniture “Test & Target”… not Site Catalyst.

  27. Thank you very much Lance. We’re going to try segmentation in Visual Website Optimizer.

Trackbacks / Pings