We’ve been testing design concepts at highrisehq.com since this past May. I want to share with you the different designs and their impact on Highrise paid signups (“conversions” for the jargon inclined).
We have assumptions about why some designs perform better than others. However we don’t know exactly why. Is it the color of the background? Is it the headline? We hope more iterative testing of the winners will help us get that information. If you have any theories please add them in the comments.
Note that designs that win for us may not necessarily win for you. I encourage you to do your own A/B testing. There are many tools online that make it easy to do.
The original page
The original design had served us well for the past year. Signups were going well, but we were worried that customers still couldn’t get the gist of what Highrise did and why they needed the product.
This page would be our baseline for the first round of A/B tests.
Long form sales letter
Ryan Singer posted a link to Visual Website Optimizer’s “Anatomy of long sales letter” blog post in our Campfire chat room one day. We were fascinated by this technique. If I remember correctly there was a heated debate about whether it would work for us.
We decided that in the amount of time we took to debate the technique we could have made an A/B test to prove it right or wrong. The original page had some long form sales letter techniques, but the copywriting wasn’t as strong as it could be.
Ryan and I worked together on the long form approach. Here’s what we came up with.
The results
Over 42,000 visitors were presented either the original page or the long form page. The results were completely surprising. The long form page had a 37.5% increase in net signups compared to the original. Amazing right?
Ongoing tests
You may notice that the Highrise homepage looks different again. We’re always testing to move conversions up. In a few weeks I’ll be sharing the results of those tests. Thanks for taking the time to read this article. Keep designing, keep testing!
Josh Henry
on 21 Jul 11I was very impressed from a design standpoint at the site. The one thing that sold me was the ability to view the videos on how to use the product. They were very informative and solidified my purchase decision. We now use it exclusively here at TheChurchBusinessGuys.com to manage our contacts and open proposals where we previously had too many falling through the cracks.
One thing I would have liked a bit more would be a shorter, “flashier” video that detailed the entire application more succinctly but gave a quick overview of the entire process.
Thanks for a great product, actually had a friend call me last week looking for a good online sales management tool and I passed along Highrise as the only option he should consider!
Wagz
on 21 Jul 11Jamie,
Thanks for the explanation, awesome stuff – testing is fun! Have you discussed doing a multi-variate DOE test on any of your home pages?
From a marketing perspective, one of the first things I’d try to isolate is that sweet new value proposition in the headline. It’s so much better and more customer-centric than the bland explanation of what Highrise does on the original page that I’d look long and hard to see whether or not it’s the better value prop or the long form format (Ogilvy always did say reading drops off after 50 words but picks up again after 150).
You could also gather a little more data by placing some trackable elements at the bottom of the long form to see how far folks are scrolling.
Matt
on 21 Jul 11Why I think this one is better. 1. the tittle is a better descriptor for what the product does. “we manage client info” 2. screenshots. There are so many of sales letters that don’t actually show me what I would buy. 3. there are real fact about what the product does, and why i would use it over another product. Again most of the time I get a lot of sales jargon and little facts with most sales stuff. 4. it is set up so you can give it a quick skim, find the high points and if necessary go back or skip over a section that you do or don’t care about. Each section encapsulates info that it presents. One section does not depend on another. I think header choices are a big indicator of importance. I’ve seen a ton of people that don’t see this choice as a conscious decision of importants. For instance there should be only one h1 header and that should be the most important thing you need to say. Also all h2 headers are the same important, but more important than all h3 headers. The size of these should easily differentiated and not easily confused.
Kevin Dees
on 21 Jul 11I have a question about the design. I find it intresting that your response went up even though the signup button is at the very bottom.
Is this the only place people had the option to signup? Or, were the conversions funneled into other pages from the links on the page, instead of the button at the bottom?
I hope I’m not asking too much.
JD
on 21 Jul 11Kevin, you’re not asking to much! The bottom was the only place people could sign up.
Jeremy Reeves
on 21 Jul 11So glad I could help you guys make more money!
(I wrote the blog post on Visual Website Optimizer’s site that gave you the idea to test a long-form letter)
Awesome stuff, keep it coming!
Freelance copywriter Jeremy Reeves
Aaron M
on 21 Jul 11It’s hard to see what specifically helps when you change a lot. You can always just hone in on specific areas to see what performs better. Such as reorganizing the copy placement, or changing something with a lead form, etc.
Kevin Dees
on 21 Jul 11Almost every site I have seen or helped create has the “button” at the top of the page, above the fold.
Very interesting display of design elements at work here. Well played.
Thanks for the response.
Anonymous Coward
on 21 Jul 11I suspect the long-form is working well in this case because your visitors are already qualified leads who are just looking to learn more about the product… whereas the short-form works better for trying to catch the attention of the random passerby.
Wolf
on 21 Jul 11This is good. I’ll use this when someone is going crazy about the mythical “fold” again.
Garry Shutler
on 21 Jul 11I wonder if you’d get more conversions by adding a “I know what Highrise is, I just want to sign up” near the top.
I understand how the long form may convince someone on the fence but it could frustrate someone who’s already fairly convinced to scroll through it.
Dan
on 21 Jul 11How much consideration is taken as to how much each “section” shows in the screen?
Stacy
on 21 Jul 11Here’s my guess why the long form converted better.
The long form is just simpler for customers. They do not have to think about which button to click next to “learn more.” They don’t have to navigate the site. They don’t have to worry if they are missing something behind a button they didn’t click on. Everything is staring them in the face in the right order for their consumption.
I think this is a very meaningful a|b test. Trying to produce a long form will also tell you if your product is simple or not.
Thanks 37s for the peek :)
TR
on 21 Jul 11@Jamie I just came across a site that YOU will find useful in testing different design, landing pages, UX, copy, PPC campaigns, or whatever. Well, maybe not PPC for you ;) It’s so cool because you can use your own customers (or not).
The customers record their screen as the meander through your site, and say their thoughts out loud into their mic. So you can watch how they interact and listen to what they think. THE SITE
They have a cool model too, it reminds me how 37signals always refers to the most successful people in the gold rush were not the gold panners, but the people selling shovels/jeans.
They epitomize that sentiment (panners=startups), but are also realizing the “opportunity” in shitty unemployment rates, because they pay the people to test the websites. Definitely worth a look .
PS I don’t work there I swear, I just think it’s fucking brilliant
Josh K.
on 21 Jul 11In my opinion, the new ‘Highrise is an easy-to-use-app…’ strap-line on the long form probably had more to do with increased conversion than anything else.
It says ‘Hey, this is what your looking for. Take a closer look’.
Everything after it says ‘Hey, here is how you talk your boss into paying for it’!
The original tells me a lot about what Highrise can do, but it doesn’t really tell me what Highrise is. The strap-line fixes this problem to a point, and the screenshots fill in the gaps.
A few things that could be a little clearer early on: Is this an online service or a client application that would require additional architecture and management?
Lysonne
on 21 Jul 11Ooops! Old UI is pictured on your current Highrise homepage. If I sign up, the product will look different from what was promoted. Just sayin’
Doesn’t detract from the value of these design posts.
Josh K.
on 21 Jul 11Hmmm, maybe it’s not just one thing, but the combination of everything. Maybe the secret is to just throw everything that might work on one page to improve your chances that the user will find something that resonates. Of course, I’ve argued the other way for a long time. Have I been wrong all along? :)
The UX side of me is amazed that someone would take the time to scan all that extra content. I guess those shopping for a solution and those utilizing a solution have different goals. What works for one doesn’t work for the other.
At the end of the day, I suppose us UX guys are just a little better at guessing what might work that the average joe. The final decision should always be driven but what actually works. Feedback is king.
Shay
on 21 Jul 11@TR 37signals is able to iterate their design (while live), and still get customers to sign up. So what is the point for paying for a site like usertesting.com when they can do it all themselves?
Chris Johnson
on 21 Jul 11How do you think a really nice video would work? We’re beyond biased, but…we’ve only gone backwards a couple of times versus a slideshow or whatever.
Casey Schorr
on 21 Jul 11This is really interesting. We “updated” our www.printfection.com marketing website with a new design, and more information and got a similar % increase in signups. Might have to try this sales letter approach next!
Casey Schorr
on 21 Jul 11I’m curious how your “original” design (in this post) compared with your previous HR marketing site design from a year or two back conversion-wise. (the design that looks similar to the current Backpack marketing site design).
Mark
on 22 Jul 11The longform provides the reader with a visual logic that is hard to classify into clickable chunks for the designer. We often learn by osmosis; and sometimes the spatial logic of seeing things “all there” affords us a sense of completeness and continuity that is often broken and interrupted when we are forced to click through for the next chunk.
I believe an apt analogy is this: Imagine receiving a direct mail letter that uses a new page fr every paragraph of info. 20 paragraphs, 20 pages. Doesn’t make much sense.
But a well written letter of 5 or 6 pages, with paragraphs clearly and concisely marked with subheads, etc, allows the unbroken flow of info and just makes more sense.
Well done.
Corey McMahon
on 22 Jul 11Wow! Great experiment. Just goes to reaffirm what “the man” aka Ogilvy said many years ago – long copy sells.
David Omoyele
on 22 Jul 11Congratulation on designing a long form sales letter that does not look a letter from a huckster. I’ve learned the reason why long form works is because people who are looking for a solution from an unknown entity want to read the details. So long form sales will typically only detract people who are not really interested in the first place, or people who hate ugly long form sales letter. Here is a suggestion, instead of removing all the 4 videos consider combining them and make on long video.
Michael R. Murphy
on 22 Jul 11Like Kevin Dees, I’m definitely surprised that the long form sales copy converted so well. Especially since the call to action signup was at the very bottom. “Conventional” wisdom has always said that call to action should be prominent and ideally above the fold. Guess that’s why A/B testing is so important ;)
Bill Seitz
on 22 Jul 11Please publish your new results before you radically change the homepage again!
It seems bizarre that you would go from short to long and then back to short without lots of incremental testing on the long (e.g. change the “Upper hand” tag which I personally don’t like, add a call-to-action link at the end of the “On this page” block, etc.).
When making such a macro/structure change like this (short to long or the opposite), would it be better to start out with minimal content changes (e.g. take a bunch of existing short pages and bring them together as a long page with the minimal editing necessary for consistency), then test that before making significant content changes?
Martin Edic
on 22 Jul 11I’m not surprised the long form form pulled better. The information flow is much more organized in a linear fashion which builds a compelling story (benefit followed by benefit followed by benefit…). The old web site-style design is basically a bunch of features. The only thing I’d probably add to the long form would be block quotes from satisfied customers. This inspiring me to do something similar with an offer we’re working on.
Kevin Bombino
on 22 Jul 11agree with David Omoyele - this is one of the best looking sales letters I’ve seen on the web.
However, I’d argue that part of the change is probably due to the much more emotionally compelling headline “keep the upper hand” rather than “know your contacts by heart”. Everybody wants to do the former, I’m not sure if I care about the latter.
A more fair split test would use the same headline in different formats.
Gary Sevounts
on 23 Jul 11Let me start by saying that I really like your “Rework” approach. I just started using Basecamp and I like it too… a lot! Below is my feedback based on my experience of 18 years in marketing software and web apps. I have worked with long letter, graphical, hybrid approaches, as well as good and bad copy. In fact, we finished a similar exercise recently and had very interesting results and web design that followed.
I think your original page was very confusing. After reading it several times, I still had a very little idea on what did the product do and why should I care. The copy is sort of an inside-out view, instead of customer’s view in a way your target buyers can easily understand. It had a lot of useful data points, but no cohesive story that communicates the benefits of your product to a person who has no idea about it.
So, the long letter version is definitely an improvement. It tells a story. And people can understand it better. I think that could explain the A/B test success. While now there is a lot of really good text, there is still a lot of fluff, like “sleep well at night,” or “impress with preparation.” I bet your customers never use these words when they talk to each other about your products or the challenges it solves. And we have done A/B testing showing that “fluff” turns a lot of visitors away immediately and they leave the page without reading the rest of the text.
Another problem I see is that the long text is, well, too long and complicated. This sort of runs the opposite of the concept of simplicity, that “Rework” preaches. The trick for a great marketing that generates strong “conversion” leads is not to dump lots of words (even great words) on a page, hoping that the fact that here are tons of them will capture different audiences. As I mentioned earlier, our A/B testing shows that many visitors just leave even without browsing the page. Yes, you are capturing more people than before, but I am convinced it is a fraction of what you could get with the great products you have.
The trick is to come up with a balanced approach that combines design, graphics, elements (like a video, testimonials, etc.), great text that resonates with your prospects / customers and a call to action. Hook them with a very clear main message, substantiate it, then provide them with elements and paths that lead to conversion. One person may chose video, another one screenshots, etc. Keep measuring the traffic and see which of the paths / elements work the best and maximize those.
As I mentioned, we did a similar exercise recently. Here is what we inherited: http://cloudprotection.pandasecurity.com/
We ended up with a design and messaging that our customers helped us with. All the paths, exact wording, elements come directly from our customers. In fact, we get a lot of comments from prospects that they “recognize” themselves in the video and that they quickly “get” the message, leading to high conversion rates.
Here is the new site I am talking about: www.forgetsecurity.com
If you notice, under the “Products” tab, you have all the same info of the long letter, but we give the viewer freedom and flexibility of choosing where to go and when.
I think a similar approach would generate much better results for you as well – I am convinced more than a long letter.
Sorry for a long post (after preaching against long letters ;) and I hope this helps.
Piotr
on 23 Jul 11Great great blog posts, guys.
I enjoy your writing a lot but none more than your how-to posts (e.g. Ryan’s posts and videos on hand-on design were just superb).
Are you experimenting with redesigning your other products’ sales pages?
Can’t wait for the follow up
Jimmy G
on 23 Jul 11Good post Jamie. I design and build every day and I find it hard to be subjective on layout sometimes, I find you need to show loads of people your ideas before actually committing them to code.
Jimmy
Scott
on 23 Jul 11Can you share a little bit about why you find the result “completely surprising”?
Paul
on 24 Jul 11Congratulation on designing a long form sales letter that does not look a letter from a huckster.
Paul
on 24 Jul 11My above comments got cut off. I was quoting someone else who said the above and agreeing that the success of this page – not letter as such – is down to the clear design and copy. It’s a long scrolling page not a salesletter. I’d be interested in seeing a salesletter type style tested against the above version – I think the salesletter would lose.
Robert Long
on 25 Jul 11Short Pencil better than long memory
Michael Fritz
on 25 Jul 11You guys are awesome! I work with landingpages for four years now and I’m always inspired by your ideas.
Greetings from Germany.
christina
on 25 Jul 11Web Design Bolton
JD
on 25 Jul 11Scott, I found it “completely surprising” because I had strong doubts about the long form technique.
Melanie
on 25 Jul 11Great test. It certainly helps that your headline has a killer benefit—the reader gets to keep the upper hand in their business dealings. Who could resist that?
As a web writer, I was always taught to keep it short, but more of these tests are popping up showing the long copy can work in your favor if you’re really after conversion.
Every site has its own needs. I think I will consider testing long copy for clients whose goals are not just branding or informational, but strongly sales-oriented.
p.s. I usually check out whichtestwon.com to see more examples of A/B testing.
Peter Cooper
on 25 Jul 11I noticed the weird URLs back then and went through all the pages. I might have tweeted about it – sorry if I threw off your numbers ;-)
Anyway, great to hear the results. Just wanted to highlight a site I’ve been enjoying lately (as a big nut for this sort of data) – ABTests.com. It’s basically a repository of screenshots and results data for tests like these.
Amber Cebull
on 25 Jul 11What I love about this test is that some people like the first, some people like the second. Both functioned well to provide information and get conversions – but when people were able to read about it, engage it, they loved it even more. Agreed with changing a lot of things kinda muddies the test… The moral is, though, engagement clearly reigns. Wonder what the difference was with video vs written testimonials…
Paul Lindemann
on 25 Jul 11As a web user, I’ve noticed a frustrating trend towards simple, overly terse, cool-looking home pages similar to Highrise’s original. That design pattern can be inviting for casual visitors, but the forced hierarchy of multiple pages fails when I’ve arrived specifically to quickly understand what a product offers and judge its potential benefits.
In contrast, the long-form version covers everything I need to make that initial judgement, laid out so it’s easy to scan and skim and read detail on the aspects I care about.
Ironically, though, as a web designer I’m not sure I would have proposed something as dense and lengthy as the winning version for a client’s site! Thanks for describing your testing experience; this clearly deserves more consideration.
Richard
on 26 Jul 11Thanks for that analysis. I will try and use some of the data and information to my own advantage. Cheers SVN
Axel
on 26 Jul 11This is awesome stuff and pretty amazing. I’m always thinking back and forth if I should have more long and very detailed landing pages/sales letters or not.
Your case sounds very interesting and it might be the great combination of screenshots, good facts and a good product leading you to success.
Very interesting having the signup part in the very bottom only!
I’ll give it a try for one of my new ventures: < a href=”http://tools-for-coaches.blogspot.com/”>Tools for Coaches
Marc-Pierre
on 26 Jul 11Nice insight – btw is it intentional, that you use no hover states at all? Always thought it’s kinda important that your buttons and clickable object react to interaction?!?
Thx!
Dirk Mostien
on 26 Jul 1137.5% more sign ups seems ok. but how much is this in absolute numbers?
I had to go to the bottom of this page to make a comment!
Dirk Mostien
on 26 Jul 11Could you also give me an indication of the average age range you are targetting or getting response off? (W- X or Y generation)
Sean Kelly
on 26 Jul 11It is a shocking result, considering that “shorter is better” is a fundamental web-writing principle!
Yet many times, when I’ve been truly interested in finding out about a product or service, I’ve found myself thirsting for a lot more detail, down to minute nuances, and been disappointed when the vendor doesn’t provide it.
David Ogilvy would have liked the long-form letter, and I know that Jason Fried is a fan of David Ogilvy.
Here’s a quote from David Ogilvy (from his book “Ogilvy on Advertising”): “All my experience says that for a great many products, long copy sells more than short.” But he warns that “if you want your long copy to be read, you had better write it well.”
I think this long-form letter is extremely well written!
Dan
on 26 Jul 11If these are impulse buyers, will they stay to be billed month after month?
Nathan
on 27 Jul 11You tested an incumbent landing page vs. a rookie landing page so this may not a fair assessment. Your historical page has been live for quite some time so visitors (both customers and non-customers) have had time to build up their own personal bias against it. But when you launched the new landing page there could be no such bias by definition because it was new.
Thus, for the purposes of this test, you should have scrapped your incumbent landing page entirely and tested 2 rookie designs to see how they perform independently vs. historical data. I wouldn’t be surprised if 2 well designed (and totally different) landing pages both outperformed your historic page.
Jason Leister
on 27 Jul 11Not surprising at all guys.
This discussion is closed.