Last week, Roberta Rosenberg at Copyblogger authored a piece that offered some terrific advice for SEOmoz's premium membership landing page. We wrote about that entry and mentioned that we'd like to be able to improve the page's current, relatively dismal conversion rate. But, that's not the whole story...
Here's how it really happened:
- Brian Clark, Seth Godin & Darren Rowse launched the Copyblogger Copywriting Contest back in May
- Despite an excellent competition with some very cool results, they were all a little disappointed that the contest couldn't actually test and measure the effectiveness of the various submissions
- Enter Rand & SEOmoz - on a phone call with Brian & Roberta several weeks back, we all stumbled onto the idea of launching a competition that would test, under real world circumstances, the quality and effectiveness of landing pages - a real life, multivariate test on a live website with everyone's submissions in the running.
- Roberta first reviewed the site to illustrate the current problems it faced, and now, we're opening it up to anyone and everyone - ladies and gentleman, start your keyboards :)
Now, here we are, ready to launch a competition 3 months in the making - a chance for SEOs, marketers, copywriters and landing page specialists everywhere to test their mettle against one another. Jeff Pollard, one of SEOmoz's talented developers, has put together a submission page where you can go to create or link to your own landing page submission for the competition:
To whet your appetite, here's a few of the details:
- Prizes! - The winning entry will receive $1,000 (by check or Paypal) and our first corporate membership (which provides access to SEOmoz premium content for up to 5 members of the same company). Second and third place will also receive 1-year free premium memberships (if you're already a member, we'll add this on to the end of your stint).
- Respect (& New Business) - Your fellow marketers will be able to see at the end of the contest whose entries had what conversion rates (for the top 5 entries) and we'll provide links and information about each of the top submissions. This is an amazing opportunity to promote your company's services to lots of folks who are desperate for exactly this type of assistance. As testimonials go, they don't get much better than this.
- High Quality Testing - the contest will feature multivariate testing using software from Offermatica. We'll judge on conversion rates, not number of conversions (to make it absolutely fair) and test over 3 weeks. We've also got some advanced analytics on the page to ensure there's no cheating or manipulation.
Big or small, well known or relatively anonymous, this contest is a great tool for any consultant or business seeking notoriety and recognition for their excellence in landing page optimization. But... it's also a great tool for those of us on the sidelines.
As an example of real-life testing, landing page authoring and what makes for great conversion rates, it doesn't get any more transparent or fascinating than this. The contest is a real opportunity to observer the conversion process in action and the results should be educational to online marketers of all stripes.
We'll be accepting submissions from now until Tuesday, July 24th (8 days away), then running the multivariate testing from July 25th to August 14th (3 weeks). The winners will be announced at the end of that week, along with all of the data from the competition including impressions, paths, conversion rates, etc. We plan on pulling back the hood all the way on this one.
Many thanks to Brian, Roberta and my hard working team (who helped put this together over the weekend).
BTW - The contest submission page has many more details including access to assets, information on rules, how to get access to premium content, etc.
UPDATE: Bryan Eisenberg from Future Now (and Grokdotcom) emailed to say that they'd like to additionally provide the top 3 winners with signed copies of their book on landing pages & conversion funnels - Call to Action - as well as one seat at their upcoming Call to Action seminar (to the 2nd place winner). It's hard to say no to such generous offers :)
Hmmm, $1000.00 isn't that much for a quality landing page layout, design and testing. Don't get me wrong, I see the value of all the services you are offering along side the cash, but you plan on actually using the page that wins on this site, right? Plus any of the others that may perform well in addition to all the stats from each submission.
Sounds like you have the most to gain out of it, that's all.
It's a brilliant idea and contest, but are you going to share all this with the rest of us and provide the exact stats? Top 5 is kinda lame, what if there were 100's of entries. I know you have a transparent business model, so I am hoping you will not let us down and keep all the SEO goodness to yourselves.
It's not the prize money, I get that. It's more for a chance to break into the industry with your landing page design skill or to gain more exposure, but can't you ask others (we know you know people) to provide some $$ donations or other prizes to add along with the current ones?
I would consider donating $$ or an xBox or something if I knew you would publish all the data from the entire contest.
Don't mean to harp on your idea, just my 2 cents.
I am sure it will bring lots of link love and a great converting page for you! You deserve it though, for all the hard work you have put in over the years.
Those are excellent points, eCopt. We are certainly going to be benefiting quite a bit from this - I didn't mean to attempt to hide that fact. The contest may not be entirely altruistic, but I do think it will be of fantastic value as a reference for those doing this type of work in the future and pitching this kind of service to their clients.
With the prizes - yes - obviously we're shooting more for folks to contribute to get the recognition than the cash or membership. We had thought about asking others to toss in prizes, but felt that it would be leveraging our relationships a bit far - after all, we're the ones benefiting, so we should be paying. As you said, it's really about showing off your ability to create a fantastic landing page.
As for the entries, we used Brian's competition as a general guideline and are predicting fewer than 20 entries at this point. If that changes drastically, we'll definitely open it up more... But, then, testing 100's of entries would also be extremely challenging.
And, we'll be showing off ALL the stats - every one we can find/dig up/create on the fly. The goal is to display absolutely everything we can and make this a very transparent experiment.
Sweet, glad to know Rand. I just wanted to be clear on what you wrote so I could understand it all. I am mainly interested in the stats and what we can learn from testing all the different designs, not actually participating, so I guess my goals are kind of selfish too. lol.
Your last paragraph = eCopt is a happy camper!
Oops forgot, edit:
I am glad you decided to keep the prizes "in-house."
Even though I suggested it above, after I thought about it more I decided it may not be a good idea as it may take away some of the luster from the services YOU are offering as prizes. I didn't want to make it sound like what you are offering is crap, it certianly isn't.
Cool idea, million thumbs up!
Of course, after I responded, Bryan Eisenberg pinged me with a very nice offer for the winners, so I think I have to throw that in. :)
Haha, go figure. Maybe it sparked a whole flurry of offers and now you'll have to add in others too!
But, the good thing is, more prizes should = more participants.
Any chance of a file with the actual underlying data collected? Would be fun to keep Offermatica on their toes and play around with a few other techniques on the data. If something interesting came up, might make a good education post.
David - I think it would be very challenging to keep data consistently updated, but we may try to do 1-2 interim posts with progress.
Wo! yes, no didn't expect the data constantly on-line! But a couple of posts, even one right at the end would be great.
Btw, did you know that being Rand is worth 1694.34 points all on its own...? Will explain more about that in a few days time.
Add another selfish expectator here.
I think everybody will benefit from the data Rand will be collecting from this clever experiment.
This is not multivariate testing. This is ABDCEF.... N testing. Possibly problematic, possible not a problem at all.Why I say that: if it were MVT, you would give all entrants an opportunity to opine about the headline, ditto for the first paragraph, in fact, each area on the page needs to get its own mBox (since you are using Offermatica.) But the rules as I understand them (and maybe I misunderstand!) are much less strict, it seems like everyone who enters gets to submit a page. So you have to rotate the traffic among the pages sugmitted, not among the mBoxes. (Although technically, each page probably becomes an mBox, it is still not MVT.)Which means that you can't use the Taguchi method that comes with Offermatica, since it is not a question of which combination wins; there are no combinations here. And once you have to drop Taguchi, you have to have enough traffic to that page, based on your current conversion rate and the number of entrants, to get a sample size large enough to declare a statistically significant winner in three weeks.
When I look into my crystal ball, I see that you will be okay. But since SEOMoz is such a special blog with a unique opportunity to teach, you might want to teach the difference between testing methodologies and sample size implications.
ooh a chance to show off page design skills? this i can handle! i think...at least better than link building...
should the landing page fit within the current general seomoz.org design parameters or is it a free-for-all-anything-goes kind of thing?
Anything goes, dude. We're looking for the best-converting page, and if you think the best way to get conversions is with flying monkeys and marquee tags, then have at it!
Flying monkeys convert like nobody's business. :)
This is very exciting!
I'm hoping I can round out a team in time to enter the competition. I'm finding it difficult with it being vacation time.
At the very least I am looking forward to seeing the results!
too late to submit a landing page? i'd like to submit one tonight.
Fun idea. And smart. I'm in also, not to win, just to participate and too see how my idea works out. :-)
I've been hitting up the premium sign-up occasionally to check things out.
The last one I saw, I had problems finding how to sign-up and I was looking for it.
Can't wait for the final results.
Please, where are the results?
This is SOOO exciting!
See this post. :)
Just to let everyone know, we have a winner! It won't be announced yet, but I turned off the competition and put our old landing page up for the time being.
has a winner been announced yet? can't wait to see what page won and read your synopsis of the competition.
We should be announcing a winner shortly.
I agree there is an innate flaw in the competition unless you are able to make sure everyone gets a shared amount of visitors for each ppc keyword, each referred site, people that come through organic results etc.
It is easy to do a landing page test when you are being specific about the traffic tested - such as PPC or banner ads etc. which can be isolated and tested... but most times you tweak and the competitors would need analytics on their traffic specifically to do a couple of tweaks and make huge improvements.
Choice of words on the page are influenced by the source of the traffic.
Whether the visitor is a first time visitor or repeat has an impact... can you change the page content based on previous actions... what browser are they using.... which search engine or site (forget about the keyword) impacts how you pitch...
The competition is a luck of the draw result - a good start for your landing page redesign but not a thorough test of the skills of the competitors...
Now if there will be insights for the competitors to make changes etc. and a weighting according to previous ROI for traffic groups... then it could be interesting!!
Frank - we are going to do the multivariate testing in a very fair way, so that every entry will receive an equivalent amount of traffic from every source. The software basically shows every Xth visitor your landing page, where X is the number of entries.
This will insure that over time, the number of visitors to each page from paid traffic, banners, first-time vs. repeat, etc. will be nearly equivalent to each entry.
I agree that one weakness is missing the ability to tweak and tune your page, but since no one else can do it either, the playing field is, at least, level.
Fair enough... was just playing devil's advocate... want to remove as many of the confounding variables as possible
I'm all in. Getting updated data and stats, competeling side by side w/others....this is really a good idea. Fun!
Wow, this should be fun.
It is going to be very interesting to see the results.
BTW: there is a typo on item #2
Fixed.
An example would be nice, but to answer my own question here is the current premium member page:
https://www.seomoz.org/users/premium
So the contest is basically create a page with those 3 sign up buttons on the bottom that converts the best. Correct?
Chris - yeah, you can use that existing page as an example, though as we've noted, it doesn't convert particularly well (note that if you're logged in as a premium member, you'll need to log out to see the real version).
...
Offermatica does not change the content or URL structures that search engine spiders see. Spiders will record the same site they currently index.
I suppose that's cloaking? :)
To answer my own question: yes it's cloaking. I personally don't have any trouble with it but I was wondering, Rand, if you had any thoughts on this before you decided to use their software...
Google does the same thing with its Website Optimizer by using Javascript to serve up the content being tested and leaving the code there for the SEs to read. I assume that if this type of cloaking is promoted by Google that is acceptable to use, assuming you are not trying to push the limits for evil uses...
Take a look in the forums and you will see a lot of talk about this when Google started promoting Website Optimizer
I know mbarr, I've discussed it multiple times :) I was just wondering if Rand had any thoughts about this ;)
Joost - I've talked to search engine folks about this in the past and they really have no problem with it (hence Website Optimizer). There's "good" cloaking and "bad" cloaking, right?
It's all about intent - if the goal is to hide optimized copy to manipulate rankings, that's bad, but if the goal is to provide the best possible user experience (obviously, in this case, the one that converts best), that's good.
I've talked with many conversion specialists and none of them have experienced an issue with the engines getting upset over a multivariate testing program.
Thought so but it's always nice to know other people think along the same lines. I agree with you on intent, of course, but the Webmaster guidelines have a more "black and white" version of this...
Would be nice to see if we could get them to officially state it's OK to do this :)
Rand,
Can you shed some light on the sales funnel for the landing page...
What % of traffic to the page comes from PPC vs. people browsing the site? Landing pages that work well for PPC tend to be very different from those that work for people browsing a site so I thought I should ask.
I must agree that we need lots more info here to make sure our landing pages are targeting the right message to the right people.
If there is PPC (which I assume there is after seeing PPC for SEOmoz all over the place lately), can we peek at the terms (assuming the hood is being pulled back all the way)?
Todd - excellent point. I would probably do well to make a post on this very subject, not only to assist with the contest, but also to give folks an informative look at what makes premium membership tick.
Rand, what a great contest. Very SEOMoz style too, looking forward to seeing the results and the community of ideas come together.
Not sure if you are able to pull it off, but a temporary account with Offermatica for the duration of the contest would help. Or tweaking the ad during the contest is no go? That leaves much of landing page optimization off the contest.
Also, if I want to write my own HTML, but host the page on your site, what do I do?
Thanks.
$49 a Month starting July 18... what will the 6 month & 1 year prices be?
also, where EXACTLY will our content start. Is this illustration correct? 760 pixels wide that we have to work with?
I think 6 months is going away (nobody ever signs up for 6 months). 1 Year will now be $399 instead of $299 - which changes the discount from $169 to $189 for going yearly.
Yup, 760 is what you have to work with.
Actually, I'd say that if you want to create a wider or slimmer page, that's cool. You're free to keep the header or toss it out (as Roberta recommended in her clinic) and every nuance in design or mock-up is up to you. We've noted that a lot of successful landing pages have little to no relation to the rest of the site.
Just to clarify ... Where is the traffic coming from? AdWords? Yahoo! Search Marketing?
If Paid Search ... what is the exact ad copy?
Each of these factors should play significant roles in how you create the page.
I'll try to update the entry with this information later today.