First ever profitable campaign. Help me milk it please! (35)
Just checked the placements again.
One has a CPA of $2.33 and one has a CPA of $3.45.
Offer payout is $3.35.
Both placements have had around 70,000 impressions.
Is it too early to cut the worse placement? It is still slightly profitable.
I would never cut something that is profitable, as long as the worse placement does not compete with the better placement.
You are aiming for Profit, not ROI.
this is a business... would you rather have $100 Profit at 90% ROI or $1000 Profit at 50% ROI?
well... that does not mean to try optimizing your profit margin. of course 90% ROI is better than 50% ROI, but I would just not kick a campaign while it is profitable.
Are they competing for the same impressions?
Stop clearing your cookies so much?
On topic: quite impressive that advertiser creatives have netted you such nice ROI already! Definitely start getting out some custom creatives if possible.
on board should be shown only first time user login in his lifetime
you should not track it by cookie,
it should be in the db..
its very annoying..
Originally Posted by delash
I did basically build/code the entire thing so it was a bit outside my depth to gel it with aMember, log everything to an STM server database, do a database lookup on login, etc.
It is a priority. Anyway, let's not derail the thread more :-)
I let my campaign run overnight with just the best performing creative.
Results so far for today:
Spend $157
Revenue $170
Profit $13
Not great!
However this inclludes the other creatives which I switched off early in the day. So if I look at just the best creative then its a bit better:
Spend $90
Revenue $123
Profit $33
ROI 36%
I will try making a LP and some more banners.
Meanwhile I am going to try and get as much vol as possible for this one performing banner. THe site gets tens of millions of impressions. So Hopefully when I get up the budget the ROI will stay positive.
great stats!
congrats for this campaign.
Do you use a tracking system like
Looks good, I love those campaigns that are profitable just right from start!
erfolgskompass, I am using
JUst a quick note: I am based in Thailand but running this in USA but from now on I will use Eastern USA time to say when days start and end.
So things have gone very quickly downhill with this campaign! Typical. I was trying not to get excited but to be honest I did have visions of scaling this and banking hard. Ah well it looks like its not gonna be that easy. Anyway here is how things have developed.
Day 1:
47k impressions
0.12% CTR
19.64% Conversion rate
Spend: $20.65
Revenue: $36.85
Profit: $16.20
ROI: 78%
Day 2:
834k impressions
0.10% CTR
9.39% Conversion rate
Spend: $345
Revenue: $274
Profit: $-71
ROI: -20%
Day3: (So far, its only 6am)
87k impressions
0.10% CTR
8.24% Conversion rate
Spend: $40
Revenue: $23
Profit: $-17
ROI: -41%
So as you can see on day 2 I massively increased my budget and turned a 78% positive ROI into a -20% ROI.
The CTR and CPM is broadly similar but conversion rate has tanked from 19% to 8%.
Day 1 was not a full day as I started maybe half way through the day. I say that as maybe the time of day is a big factor? It seems weird that the conversion rate would literally half from one day or time of day to the next. This is a simple Zip submit afterall.
Really not sure what to do now. I am using advertiser creatives and there are only 2 placement positions available on this site. So there isnt much room to test other variables until I design some new banners and get them approved.
The site I am running on has massive volume so I cant belive that banner blindness or anything like that is happening as I have only been runing this effectively a day and a half!
Any ideas what I can do? Really stuck what to do now ,my only idea is to get a bit more data and then lookinto day parting.
I will also try to design an LP and some banners too.
##NEW EDIT## maybe the original positive ROI was just because of a small amount of impressions? maybe the true conversion rate of this offer on this traffic source is 9%..
Checking the stats for Day 2 again.
The 160 x 600 placement had an ROI of -9%.
So that is better but still negative.
I will continue this campaign with just this placement from now on,.
Hmm - how many conversions have you sent to this now?
One of the reasons why zip submits aren't run more is that they have a well-deserved reputation for scrubbing harder than the janitors at a CDC incident site. OK, that's a bit of a contrived metaphor, but still - they scrub wicked hard.
Hence, this sort of pattern would be pretty normal if you've fired enough leads that they might have noticed.
Can you get the same offer on a different network? If so I'd recommend split-testing the two and seeing if the other converts higher.
Hang in there, and keep trying stuff. It looks like you're close...
Originally Posted by caurmen
I might try breaking into smaller campaigns and target by city or something like that.
So what I did was check through my stats and there was a two hour slot that had been performing best of all. So I tried day parting to just that time slot.
Day3: 2 hour slot.
81k impressions
0.10% CTR
8.5% Conversion rate
Spend: $23
Revenue: $23
Profit: $0
ROI: 0%
So a break even. I will let this run one more day and see what happens.
Next I am going to break this campaign into smaller campaigns targetting probably different Citys or East Coast V West Coast. Just a thought.
So my next idea was to try some different website placements. I had thought I had struck gold with my placement but its not as good as I hoped.
So I cloned my campaign and then chose 4 new sites with around the same CPM as the site I am already advertising on.
One of the sites ate up 13k impressions in a few minutes with zero clicks so I put that offline. Another got zero impressions so I binned that also.
The other 2 look more promising.
The first website (actually a website bundle) is showing a -35% overall.
However the 160 x 600 placement had zero conversions.
The medium rectangle had - 8% ROI. So I can continue to work with this placement and see which creatives are doing well.
The second website bundle has these stats.
1600 x 600 has 61% ROI
300 x 250 has 4% ROI.
This is great news so I will keep on with this placement too. Looking to maybe cut the medium rectangle in the future.
Just goes to show that I made a mistake initially thinking I had lucked out and found the holy grail originally. Even if it was a very profitable placement I still should have carried on looking for others regardless.
My AM says he might be able to bump the payout a little. Also the advertiser is going to asses my traffic after I have sent $500 worth of conversions and see if I can be bumped further.
I have also asked for more demographic info to help with my targetting.
So now I am going to continue to test some more sites and try to improve what I am running already.
All looking good! Good luck on the bump!
OK so I ran campaign one all day and it ended up with a big negative ROI.
It seemed weird to me at first why the campaigns would start great and then fail miserably. Then I realised that the campaigns did well at certain times of day only so I dug into day parting. I had about $400 worth of conversion data run over a couple of days so just about enough to get some good info.
What I did was make a list of all the time slots which were profitable overall AND profitable (or break even) on each day they had received traffic.
I ended up with about 5 hours per day when the campaigns were making positive ROI.
As I am still running direct linked and with advertiser creatives I didnt have many more tweaks I could make so I went with this.
Campaign one stats:
275k impressions
0.11% CTR
11.4% Conversion rate
Spend: $80
Revenue: $113
Profit: $33
ROI: 42%
OK so that is more like it!
So I wont start celebrating yet. This time I will slowly increase the spend and see what happens. I will double my spend today or something like that. Part of me wants to up the budget massively but thats probably not wise. I do have $600 worth of data though for a $3 payout so I am HOPING this time I stay in a positive ROI. Ill take it easy today and increase slowly I think.
Day parting for different timezones?
So what I am thinking to do (would appreciate some feedback) is to split my campaign into 3 or 4 new campaigns and adjust the day parting accoridng to the 4 major timezones in the USA.
So lets say 9am - 11am (EST) is my best slot. Should I be running a seperate campaign geotargetting the west coast cities an 9am - 11am PST?
I have cloned my campaigns to be approved (why do you have to wait for cloned campaigns to be approved!?!?) and I think I will test this. Seems like a good idea to me. What do you peeps think?
2nd Campaign
My second campaign ended up -16% ROI. That was 24 hours of data though. I have now (hopefully) found some good candidates for day parting now and have added that to my campaign. Hopefully can get this one profitable also.
Just thinking. Is my reasoning on the day parting by timezone flawed? I think so as I don't know which time zone my current conversions are coming from. What I should probably do is run mini time zone targeted campaigns with no day parting at first to gather data.
Yeah that's what I'll do.
So I am running 4 timezone geo targetted campaigns (PST, Mountain, Central and EST) to see if the profitable time of day changes according to timezone and also if there are parts of the country that aren't profitable or if there are hot spots. Got incomplete data so far so I will wait until I have the full day of data until I act on it*.
Learnt a lesson which luckily wasn't too expensive. When you clone a campaign in sitescout, all the target sites go back to ONLINE from OFFLINE! Lost about $100 this way overnight as I have 6 placements in the campaign I am cloning from but only one is profitable at all. Good job I was testing with a low budget!
I could really use some advice now. So far I have been testing on weekdays and have plenty of data. Its now the weekend. Should I start again and test all the timeslots? Or should I run the same campaign as I have been running midweek with just 2 hours receiving impressions?
I am tempted to drop the budget right down and test twice payout for each hour.
Would really appreciate some advice!
*Just realised that I am going to end up with half my data from a Friday and half from Saturday which isnt ideal. Id prefer a full day of either weekend or weekday.
OK so what I have decided. Is to switch off the timezone targeted campaigns and run them again on Monday. I don't want to mix up my weekday and weekend data (i could be wrong on this but seems sensible).
Then I have changed my main campaign so it runs 12 hours of the day for the weekend. I also dropped the budget down to just 2 x payout per hour.
My reasoning is I don't know how weekend traffic will perform so I want to collect some data for the whole period.
On sunday I may up my budget if Saturday performs well and has some obvious good performing hours.
On monday I will resume my smaller timezone targetted campaigns and hopefully find some killer geos and or be able to optimise day parting further.
Really hope this offer stays live for a while as I am starting to feel positive about the campaign again.
Is your offer time-sensitive? Is there reason to believe it would be deleterious to just leave it running 24/7 to collect data?
I already ran it 24/7 mid week and that gave me data to find profitable day parting.
But now should I keep that day parting over the weekend or run it 24/7 again?
You are wayyyy overanlayzing it and not reaching statistical significance. Stop thinking about timezones and day parting etc. Run this bad boy for a week, get some real data and then analazye. Unless your on a tight budget, which if you are, Id suggest getting of sitescout asap anyways..
Originally Posted by vidivo
I have won 3.2 million impressions so far and spent $1k.
If you've bought that many impressions you should be able to day-part from the data provided you didn't bias it initially.
Did you run it continuously for say 3-4 days and observe that say 8 AM to 4 PM was best every day by a long way?
No I didn't do it like that unfortunately. The first day the campaign ran for just a couple of hours and got a 60 percent roi.
I then ran it for another 2 days and it flopped.
I then parted it using all the data and it was positive roi again.
The day part was the same period as the first day.
The problem is this day parting only gives me two hours. I think maybe I need to gather some more data for other time slots.
Perhaps I should continue this day parted campaign and try to get more volume whilst running a 24/7 data gathering campaign with a lower budget.
I'll check my stats again when I am back at my pc before deciding.
I made some more banners, basically tweaked the ones I have by adding some text and animating the cta. Waiting for approval with advertiser to be safe before running them.
You should really run a campaign for at least 3 days at all times before considering running at specific hours only, unless the data from 1/2 days is very, very convincing.
I think you may have jumped the gun on hour-parting and it is likely making interpretation of the data very difficult now.
New campaign, RON, whitelist your profitable domains. You can get some pretty decent extra traffic sometimes! Sites usually sells their traffic through a lot of traffic sources. You need EVERY impression from your profitable placements / sites! Also, like others said, stop analysing everything, just let it run for a while! Significant data beats everything. Try to increase volume (even if it's not profitable) - get higher payouts - optimise - ??? - profit! 
Ron on site scout a bit scary on my budget. Lots of shit on there with massive volume.
Thanks for the advice though people.
I am going to launch a new campaign from scratch and do it the right way.
I will also keep my current campaign though.
My payout is 3 dollars 50. What daily budget will give me decent data? Is three times payout per hour enough? That's about 240 per day.
Right so over the weekend I ran a small amount of traffic. I was all over the place with my thinking and ended up pausing the campaign after 6 hours.
One problem was I had day parted for a twelve hour period. The thing is with sitescout if you day part it doesnt seem to deliver your traffic evenly throughout the period. I ended up with almost no budget left after 3 hours.
The campaign flopped badly. All placements all creatives and all timeslots were negative ROI.
Anyway the stats for Sunday are as follows.
319k impressions
0.12% CTR
5.43% Conversion rate
Spend: $95
Revenue: $70
Profit: $-25
ROI: -26%
Back on weekdays I continued my limited timeslot campaign again which has just finished running.
I set the daily budget to $300 and was hoping for a 30% or more ROI. Didn't happen though!
3 hour time slot:
221k impressions
0.16% CTR
4.93% Conversion rate
Spend: $64
Revenue: $60
Profit: $-4
ROI: -6%
The campaign was slightly negative overall but with one creative making a decent ROI.
The main issue was the volume I got was tiny. I had a $300 budget but managed to spend only $64 which is pathetic. The third hour I only spent $5.
THe site has a ton of volume so I am thinking maybe I am bidding too low now as I set the max bid to just abouve the average I am paying whereas before it was set much much higher.
I will up my bid tonight and see what happens.
I ran the campaign for a 3 hour slot:
First hour is my best overall timeslot and has been positive every weekday so far.
Today: ROI 48%
Overall: ROI 26%
I still think my day parting at different times for different timezones has merit and I am going to test that this week also.
Still waiting for my new creatives to be approved by the advertiser. I used the stock creatives and did things like change button colour and add a few lines of text. I could probably get away with running them but want to be safe so I will wait. Really eager to test them thogh as a higher CTR would be golden with this campaign.
Got kicked off the offer as my traffic wasnt working for the advertiser.
A bit suprised as I was running with the advertisers own creatives and running on a good quality site ( alexa top 300).
Anyway they do have a lead gen verision which i could try (i briefly tested it before with 0 conversions).
I do have quite a lot of data so its probably worth trying the lead gen version.
What's the difference in conversion point between the two?
First was uhh SOI/DOI - is the lead gen version a paid sale on the backend? What's the payout compared to the SOI/DOI?
First offer was zip submit. $3.35
THis new offer is $13 payout short form. Name address telephone number email.
I would try that then and see how it goes. Could do well could crash and burn - my intuition says it would do worse than expected due to it asking for a telephone number - I think of all our personal details, telephone number is the last thing you want to give out (amongst name, email, etc.)
Home > Display / Media Buying >