My original PPV survey lander, concepts and lessons (23)
So following the suggestion Tijn made on my hello thread, here's a case study of my old ppv survey campaign. Since I came up with this a few years ago, it's been copied by every man and his dog and burnt to hell and back. I see people are still running it even now, so I guess someone is making money off it, but to me the interesting thing about the campaign was the ideas behind it. Hopefully we can continue to use those ideas but in new and more innovative ways to make some cash.
First off, my thinking behind making this campaign. We're all trying to get punters to sign up for stuff, or buy crap at huge discounts (to be rebilled later), or get huge value gift cards, or whatever. But punters for the most part aren't stupid. If you tell someone they can get a $500 Walmart gift card just for giving you their email address, most folks are gonna call bullshit - and with good reason. Most lead gen and rebill offers appear too good to be true, because, well, they are. Plus, these people see popups all day every day, all crying out for attention, all making bold claims about amazing free/cheap stuff. The bullshit filter learns to block out the pops.
Bearing that in mind, I figured I had three ways to try and get people to take up an offer:
1) Hope that there are enough dumb punters who actually believe they will get x in return for an email address, zip code, or $2.99 shipping and handling, to make a campaign profitable.
2) Presell them enough that I make their desire for whatever is on offer outweigh their scepticism.
3) Give them a good reason why I am offering them a deal that is otherwise too good to be true.
I chose to concentrate on the third option. Give them a reason why. There are lots of ways to do this, and I went with 'reward' as my reason why. You do something for me, and I'll give you something amazing by way of thanks. The 'thing' I'm asking them to do mustn't be too onerous, or they'll never do it. But it must appear to have value to me, otherwise what's the point of them doing it. Polls were a common way of giving a reason why, but not quite what I was after. I needed more interaction, and a more convincing reason for the reward than just clicking a vote button. That's what lead me to think about making surveys. I wanted the punter to land on a website and see a popup that asked them questions about that website. In return, they get free stuff. To the punter, this seems logical. There is a valid reason for the popup appearing - they will assume that it's the actual target website that's asking for customer feedback, which sites sometimes do. Of course, by carefully designing the lander it's possible to nudge their thinking in that direction too 
I set up a campaign with a single target url, and made my lander look like it came from that target url. I wouldn't say I copied the sites I targeted, but I certainly took design cues from them - colours, fonts and layout. Here's the first iteration of my survey lander for qvc.com:
That mimicked qvcs design at the time. Not a copy, just my interpretation 
It worked amazingly. CTRs were awesome.
QVC wasn't chosen at random, I started there because it's a quality target - a shit ton of traffic, and that traffic is almost exclusively people who buy crap. They probably have their credit card in front of them there and then. Heck, they probably already memorised their credit card number.
As you can see, first time round I asked for an email address. I took their email, and then sent them onwards to an offer as their reward for completing the survey. The offers were rebills, and I rotated 5-10 a day to see what worked, pulled the crap and weighted the rotation to what made cash. And on the backend, I sent emails out every day offering them more stuff. I built lists of tens of thousands in just a few weeks doing this.
My next iteration was to change up the page so the punter only saw one question at a time. Seeing a long form can be off-putting, but one question is quick to answer, and leads them into interacting with the page. CTRs went up even further:
By now I was having a lot of problems with emailing. The lists were too big, and the autoresponder people were getting shitty with spam reports etc, even though I was playing by the rules. When one autoresponder outfit upgraded their stuff and lost about 50k of my email addresses, I decided to split test going right to the offer without ever asking for an email address. Turned out to be a good idea - ctr from lander to offer went through the roof. As the email stuff was taking up so much time, and low delivery rates meant emails weren't making that much cash on the back-end, I dropped that side of the process and went straight from survey to offer.
I was still rotating offers on the back end, and showing just one offer to the punter. Now I changed the "thanks" page (the one they saw after completing the survey) and turned it into a multi-offer page, with three or four different offers as the reward. I ran a ton of tests on this, and stuff worked really differently on each site. By now I had scaled this out to a load of sites, but every campaign was a single target URL, so each lander was hand-built and tuned specifically to that url. More work for sure, but worth the effort. Most of the time I was the only bidder on these targets, so it was cheap easy traffic. What the tests showed was that some urls worked better if you didn't give the punter a choice but just showed them a single reward offer after completing the survey, whilst others worked better with a choice of 2 offers, others preferred 3, and so on. Apparently giving a punter a choice is bad for conversion, it just confuses them.
All this was going great, and I was doing good $x,xxx/day net profit. Then I woke up one morning, checked my stats, did a double take, looked at my campaigns in TV, and saw they had all turned to status "Rejected". Turned out Future Ads top brass had been in checking up on things, and my landers were too close to the limit of acceptable. Jonathan at TV told me "We love the innovation, but you can't run them like this any more". Landers had to be more generic. That was bad news. Here was the next version of the QVC lander:
I tested a ton of variations, of which that was but one. CTR dropped massively, but the campaigns were still profitable, so I carried on. At this point (early 2009) I hadn't really done much in the way of scaling to other networks, so I worked hard at that. Over time though, they all started getting tougher on landers that looked too much like the target sites (apart from Adon who didn't care, but they had almost no traffic, and the little they did have sucked so wasn't worth the hassle). Rules got stricter, disclaimers and logos were required, etc etc. I tested a ton of variations, but nothing came close to the original.
Not long after that, the whole rebill thing started to blow up big time. Networks were pulling all their best offers. It was getting harder to make surveys work. Harder, but by no means impossible, so I kept on with them. There was plenty of scope for improvement. For example, originally the questions were simply there to provide the "reason why" for the popup, but I had begun using them to better target the end offer.
I also split test different starting questions, and different question sets. It was important to finish with at least a couple of questions where the answer was "yes" so as to build a yes-ladder (I think that's what psychologists call it). The idea is when you ask someone questions and they answer yes, subconciously they get into the habit of agreeing with you, and so when you say "and would you like this lovely teeth whitening kit for no money now and a hundred bucks a month for the rest of your life" they say "Yes!". Sounds stupid but it works.
Over time though, ctrs fell away further and further. At first I had no idea why. I set up a virtual windows environment to start spying a little and see what others were up to (yeah, I was naive, had never spied before!) So you can imagine I was surprised to see my landers being run all over the place. Not just landers based on mine, but 100% copies. That explained the fall in ctrs, punters were seeing exactly the same popups running everywhere, and I mean everywhere - folks were running these run of network - I could see them on the most obscure pages. The funny thing was, they were running an old version of the lander, one that didn't convert anywhere near as well as my later iterations. I guess someone copied an old one, then someone copied the copy, and so on. Most never bothered testing varations.
In the end it became too much effort for too little reward, and I moved on to other things. My key lesson from the whole campaign though, was the reason why. Give people a reason why a) they are seeing a popup, and b) you are offering them something that seems too good to be true, and ctrs and conversions go up.
The other big lesson was engagement. If you can get the punter engaged with the lander before they hit the close button, you're 90% of the way there. I used both these ideas in other variations, some obvious (surveys highly targetted to the end-offer instead of generic questions, quizzes etc) and others much less so.
There you have it. That's a campaign that made a good four figure profit per day right off the bat, was almost infinitely scalable, and ran for more than six months with virtually no competition. Now the lander has been burnt to hell, but even today it looks like most people are still running an old unoptimised version. No doubt there is life in the old dog yet, just needs some creative minds, and it looks like there's no shortage of those round here!
PT
epic post and some seriously good tips.
The reason why really hits home and is something often overlooked in AM in general. Great stuff man.
Great read, thanks.
One of the best posts in STM! Epic! Thanks for sharing
Great post polishedturd!
I have never tried surveys to push rebills but I have had great success using surveys and plugging a coreg path as the backend.
Performed 2x-3x better then normal "You won" landers because made the whole thing more credible.
Great post... thank you for that insight.. Common theme in all of the successful campaigns I've seen posted/had all have "the reason why" ... never really caught that haha... Thanks for the huge insight polishedturd!
- D
polishedturd? polishedawesomeness! Thanks for sharing such an epic story and your thinking behind it!
Badass post, how did you transition from the survey to the offer I often find people leave this out.
It seems to me it would be an important step of the process.
I have done surveys and quizzes myself but I always feel like the transition to the actual offer is a weak link.
I'd love to hear what your experiences have been with that aspect of the process.
AWESOME-TASTIC post!
Thanks for letting us peek into the way you think and came up with the idea.
That's crazy man, very creative you certainly reaped the rewards!
This whole idea makes sense. People that allow toolbars to be installed on their computer doesn't mean that they are stupid. They simply DON'T KNOW MUCH on computer stuff. So treating them in this way - give them a reason why they can win the freebies makes a lot of sense to them and will definitely increase the conversion in the end.
Great post I must say!
this is great stuff any time you can engage the user conversions go up....way up....thanks for sharing
Originally Posted by bbrock32
Originally Posted by markl
The next iteration was to dump the email side and just show them an offer on a thank you page. That looked something like this:
Then I moved into multi-offer thank you pages like this:
There were a ton of variations. All of them started with a "by way of thanks, here's an exclusive offer we've arranged for you" type paragraph. After that, there were loads of different layouts, different numbers of offers to choose from, some with a presell others just the price of shipping, others no text at all. Different layouts or number of offers worked better on different targets. And obviously the images were all split test as well, because sometimes a gleaming white smile works better than a picture of the product box...but not always.
All the offers were rotated to see what worked best. Later iterations used the survey questions to target the offers. That way I knew if they were male or female (so I didn't offer male enhancement products to women, although maybe they would have bought them anyway, never tested that!), over or under 18 (no point offering rebills to under 18s with no credit card), and with their IP I knew were they lived so they only saw offers that worked in their country.
Oh yeah, on the multi-offer pages I tested out some funky stuff with adding in a really BAD offer as well...it can help nudge the punter towards the offer you really want them to take when there's a choice involved. An idea from Predictably Irrational (amazing book, highly recommended).
(BTW, to the person who sent me a PM, I will reply as soon as I have enough posts to be able to do so)
Great reply thanks!
BTW I spotted someone running almost an exact copy of your QVC lander yesterday.
I don't remember seeing that in cashvertising, I have the latest science of influence on the way right now.
Originally Posted by markl
http://www.amazon.com/Predictably-Ir...dp/B002RI9QJE/
PT
Awesome follow-up, PT!
Not trying to hijack the thread but awesome book tip you gave there, just found it somewhere and the table of contents already makes me want to read it immediately.
In case anyone wants it, here you go: http://www.fileserve.com/file/QqveAsZ (Respect the author and buy it if you like it).
One of my favorite posts in STM!
Predictably Irrational is awesome!
This is Dan's 2nd book: http://www.amazon.com/Upside-Irratio.../dp/0061995037
Very "polished" post and thanks for the share.
Those landers should be framed and on the wall of anyone serious about PPV. They are classics and if you do some spying you see that they are still used!
now the Main question is" how to build a survey lander?"
Originally Posted by rafael3000
PT
yeah just find an example one (theres plenty of survey style offers out there) and use bolt to rip and edit the page.
Home > POP / PPV / Redirect >