Ok – Ive setup 3 campaigns in the last 2 weeks that have not been profitable. So time to get some feedback and suggestions to learn from my mistakes.
This is the first one. Ill post another one tonight.
Be brutal!! I am here to learn
For the first one i used this $150 a day guide:
Kardashians (still most popular over last 30 days of all celebs)
Khloe’s change of hair colour
(Looking at todays news there are probably some better angles – drinking when pregnant should stir up some emotions)
Rotated 4 offers from Peerfly:
- Free iPhone 4 & iPad (PF1568)
- Tester-Rewards – Apple iPad (PF1426)
- Choose iPad or Kindle – Zip Submit (PF1986)
- Kindle or iPad? Email Submit (PF2578)
Targets & performance:
Basically kept it to few very focussed targets (apart from the tmz one – lesson learned). The conversion was early on in the Tmz campaign so I probably let that run for way too long.
Tester-Rewards – Apple iPad (PF1426) was the only one with 1 conversion. CR was 9%.
- Dont bid on tmz unless you have a campaign and offer that converts for a generic audience.
- Stop underperforming targets earlier
- Split test a landing page, different images, etc to try and get better CTR
- CTR overall was too low. For email subs I aim for >10% and a CR > 2%
- Good CR on the 1 conversion – should consider running the offer that converted with targets getting +10% CTR
Views welcome. As I said – feel free to be brutal. I dont like wasting money!!
Ps. One of the challenges I have is that I am in the UK using Lead Impact and controlling spend. If I set the budget at say $10 per day, it often burns through that in the first hour and therefore does not give me valid results. The option I now use is $1 per hour (thats the minimum) which is better but often the result is that I wake up in the morning, check stats, and see I spend too much. I guess staying up is the solution, but interested to know how some of you that are not based in the US deal with this?
First , imho you should choose a topic where people are more emotional, so they will react more.
About your LI question , use $1 per hour and use daypartying to make your ads show 12 hours instead of 24 ( that will cut your expenses into half ).
ah hadnt thought of day parting. nice one. which timeframe would you suggest for generic US traffic (LI is PST i think)?
actually – that probably depends on the market. Dating will be very different from Email subs based on kardashian quiz.
this is pretty handy tool – check the demographic comparison
Nice campaign gallox, I’m surprised it didn’t work better. Below should help!
– I’d stick to Kardashian related URLs only and a lot more of them (i don’t know how many you have but the super longtail ones convert a lot of the time for this type of campaign.)
– Like bbrock said try to make the landing page question more controversial
Who’s the most popular kardashian? (put them in the lander). Who’s a rival of the kardashians? Lets pretend it’s megan fox, a simple ‘kardashian vs megan fox, who’s prettier’ would be worth a shot.
– 100% Offer up a free iPad in the landing page!! Make sure it’s on the page so people know they can get one if they participate.
I didnt have many URLs – just the ones I showed you.
Ive created a further 3 variations – all along the lines of the following:
Ive also switched on day parting (11am PST to 0100am PST)
I will add some targets tomorrow.
I have now expanded to about 200 long tail URLs as well – just waiting for LI approval which no doubt will be slow on a weekend.
I have also changed the headline for the lander example I posted above to "Who’s The Biggest Drunk? Khloe or Kim?"
Will post when I have some data.
Are those boxes containing the vote buttons? If so I would try and create another lander and put the vote buttons beneath the images using a click here caption. You can name the images on top instead of using their names as buttons…Just something you could test
yes they are and they are flashing boxes trying to grab attention. Once I know which of the 3 images I am running with outperforms the others I’ll splitest the CallToAction & location of the links.
– MUCH better lander, and The title "Who’s The Biggest Drunk? Khloe or Kim? = perfect
– You need more than 200 longtail URLs, do you have a URL scraper? or doing it by hand?
Ive used an URL scraper which returned first page results for about 20 related keywords.
To expand as per your suggestion, Ive just scraped wonderwheel and all related keywords to kardashians, and grabbed results from 5 pages from google, yahoo, and bing.
I then removed duplicates, and removed pages that were not live using scrapebox. Then did a PR analysis to get the high traffic pages and checked those for relevance. I removed any obvious high traffic domains without longtail URLs.
That left me with about 3500 URLs
Ive put those into the campaign and will let it run overnight at $1 per hour max to see which URLs are actually getting the traffic.
it will be interesting to see how quickly LI approve these – i doubt that I will see any approvals before tomorrow but who knows – they might work over the weekend.
LI doesn’t work on the weekends and thats what sucks about them
Your right. They will be asleep now so I will have to wait for another 8 hours before the 4000+ targets get approved.
Here is an update on the stats.
Ive had 2 conversions – both within a span of 5 minutes. Then nothing for the rest of the day (might be getting scrubbed). I have stopped a couple of targets that had 2 and 4% ctr on about 100 views but no conversions.
LP2 sofar is outperforming the others and I have paused lp3. LP 2 has a simple "Who Is Your Favourite Kardashian" headline with a different image then the one I shared above.
I have paused the non converting offers (kindle ones) and added some top performing ipad ones from WSM and EWA:
* Get new targets approved
* Run 100 additional views to lp4 and lp2 then decide on winner
* Get CTR over 10%
* Create a new LP with a clearer submit button below the images as suggested by ppvnewbie
Ok. Lp4 was the winner. CTR on both was about 4%, but LP4 had 3 x the number of conversions. Additional targets have also been approved, and Ive stopped targets with 50+ views and no clicks.
Here is the updated lander (lp5) that I am gonna run overnight to splittest against lp4.
For the winner I will then iframe the offer with a header showing who they voted for and asking them to complete the form.
Overall I am still losing money, but did have another 2 conversions overnight.
So I will run with this for another 2 days to see if I can get CTR to 10% and CR consistent to get profitable.
That landing page is beautiful, and iframe idea is great
What conversion rates % are you getting on your submits? You don’t have to share publicly, but if it should be 4-7% or so. If its anything like 1-2%, you should switch offers.
ive only had 4 conversions and therefore this isnt sufficient data to make any conclusions in my view, but here are the stats. Note that this includes all views and landing pages:
111 Direct Link to offers – 0 conversions
1281 to landing page
0.34% CR1 (conv/views)
10.8% CR2 (conv/clicks)
With $1.30 payout per offer and 0.015 cpv the numbers I am aiming for for this to make sense would be:
ps. I am rotating a range of offers from peerfly, WSM and EWA. Its their current top performing offers. The only one thats converted sofar is Peerfly but I have had issues with scrub with them once I get over 25 to 50 leads in total.