That ugly thing converts?


I have a friend who says, “Yes Robbin, I know that testing is great, but why don’t you just start with the suckometer and make sure all those bases are covered?

In fact, most of the things on the list (if you follow that suckometer link) are really intuitive and easy to agree with. No one likes broken links, for example.

But it is amazing the number of times that I’ve looked at a site and guessed at the conversion rate and gotten it really, really wrong. Either I thought it converted famously (but in fact, it did terribly, which isn’t something that happens very often), or I thought it did terribly (and it actually performed fabulously. I feel like I see this one a lot.)

I will never forget a site that I looked at about four years ago. It was all in default font (Times Roman), way too many links, and it looked like my nephew had designed it. It was a site that sold hunting equipment, and the headquarters was in NoWheresVille, PA. I must have said something about their awful conversion rate before they told me that it was converting in the double digits.  And a conversion was a sale!

I wasn’t in the target audience, and I was never going to be in the target audience. For all I know, their target audience loved the site because it was so genuine, or those visitors/customers were so comfortable with the site. It wasn’t some fancy AJAX thing and clearly, they liked it enough to spend money there.

I think that as “conversion experts,” this is a lesson we have to learn over and over again. When that website owner calls and says, “Can’t you just tell me what is wrong with my site, I really don’t want to do any testing,” the need for humility is amazing.  (And humility doesn’t come easily to all of us!)  Yes yes, we can tell them that their links are broken, that their site loads too slowly. But we really can’t tell them that their sites are ugly and inappropriate, because we just don’t know that.

And I guess it comes back to what a big fan I am of If I am not the customer, then let’s go buy some target customers.


Our owner and CEO, Robbin Steif, started LunaMetrics twelve years ago. She is a graduate of Harvard College and the Harvard Business School, and has served on the Board of Directors for the Digital Analytics Association. Robbin is a winner of a BusinessWomen First award, as well as a recent Diamond Award for business leadership. You should read her letter before you decide to work with us.

  • Dr. Pete

    I certainly won’t argue about testing, but I do think that best practices are important. As diverse as sites can be, we have to try to take away some learnings from all of our testing, and that’s often our best foundation and starting point. Once the basics are covered, then it’s time to dig deep into the specifics of that site, including analytics, qualitative (user) testing, and quantitative (A/B, multivariate) testing.

  • Ben Hunt

    Couldn’t agree more! It’s an amazing jouney of humiity, which goes deeper and deeper.

    Yet I do agree there must be some basic Dos and Donts that can give us a good indicator. Stuff like, if there’s no spacing between paragraphs, adding it will help, product shots help, headings help, that kind of thing.

    I’m starting a range of experiments that will hopefully culminate in a book that gives a field guide to common conversion errors and how to fix them, backed up with quantitative results.

  • Dan Croxen-John

    Hi Robbin,

    The simple fact in my view, is that we – web analysts, marketeers, SEO and PPC guys – given the industry we work in, will never be representative of our client’s online visitors. This means that we have to use others – less savvy, less sad, however you want to call us – to spot obstacles to conversion.

    Plus, the user videos that you get from are a great way of bringing these ‘obstacles’ to life for the executive team. There’s nothing quite like watching the faces of the exec team as they watch a video of a user getting increasingly frustrated by your checkout process, or whatever it is..

    (Disclosure: Applied Web Analytics is a partner of

    My two-penny worth, is that many of remote usability services currently are dominated by North American users. Now for those of us in the UK, whilst I accept that no user testing is better than some user testing – it would be better to use UK usability testers – without the hassle of recruitment.

    I have come across a service called Ethnio ( that allows you to recruit usability testers from your online visitors. Until remote usability providers offer a more localised service, then this could be a great way of signing up usability testers that represent your target visitors.

    After that, you can conduct closed group usability testing.

    Hope that helps everyone.



  • Linda

    Aren’t there some guiding principles based on age or sex or price. Like men over sixty purchasing electronics like “these” colors and “this” type of design? Or women prefer the “clean” layouts for clothing sites?

  • Robbin

    Linda, I don’t think so, and my evidence suggests not. I think demographics are too large a brush to paint with. Those women who prefer clean layouts, are they living in Appalachia or LA? Are they people who like things to be tidy (that would be everyone at LunaMetrics except me) or people who work well with a mess around them (me?) and so forth. But I would welcome other thoughts on this topic.

Contact Us.


24 S. 18th Street, Suite 100,
Pittsburgh, PA 15203

Follow Us



We'll get back to you
in ONE business day.