Home > The Newbie Zone > Questions and Answers

---

How do you know when a profitable campaign is dead? (11)

---

Let’s say you have a profitable campaign that’s been running for a few weeks but you notice a drop in performance over the last couple days? How do you know when to kill it?

What about banners/landers? If they’ve been running well for awhile, how long do you let it fail before knowing it’s dead?

I get using stat calculators during the testing phase but how would you use these after the camp has been profitable over a period of time? What time period would you use in your analysis?

Curious what you guys think and what rules you have about this. Thanks.

---

Hi! Usually from my experience everybody sticks to such a rule: 1/10 part of budget is spent for offers testing. And in my opinion if CR is lower 1% campaign can be considered as dead.

---

Please refer to this post for some of the things you can do to "revive" a campaign:

https://stmforum.com/forum/showthread.php?t=39123

Try all of the suggestions.

Regarding the 2nd suggestion. Setting up a new camp with the exact same settings can really make a difference in some cases.

What happens is that when you're running aggressive landers, some publishers (placement site owners) will not like it, and will opt to stop getting traffic to your site. If you compare placement stats before and after the decrease in performance, and see that you stopped getting traffic from some of your best-converting placements, then that would be the tell-tale sign.

As for saturation, landers don't saturate quickly. Monitor Adplexity for some time and you'll see that the same popular landers have been running for months now.

Also, saturation does NOT happen suddenly. So if you see a sudden decrease in performance, it's most definitely not due to lander saturation. Look to the other 2 pillars of a pop camp: Traffic, and offer.

No stats tools are needed in this case because you're not split testing anything to find the best. There is a stats tool that can tell you what the chances of profitability are (i.e. my spreadsheet), but that won't help you in this case because you're not trying to predict profitability. Your current situation of decreased performance is already a reality.

Hope that helps! And do let us know how you get on!



Amy


Sent from my SM-G930W8 using STM Forums mobile app

---

Thanks for your input. But may I ask where you got these rules of thumb?

I actually have a different perspective: Iwould suggest spending a big part of the budget on testing offers, because that is what can make/break a campaign. Also, the more offers you test, the better your chances of finding a good offer, the less you'd need to spend on optimization, and the more traffic your offer can make profitable - basically maximizing your profits and minimizing your spend in the optimization stage.

As for the 1% CR, I have reservations about that as well. CR is only one variable in the profits formula, the others being payout and cost/bid. If your payout is high enough and/or cost of traffic is low enough, your campaign can still be profitable at below 1% CR. For lower payout offers though your rule may be suitable.

Again, thanks for your input! Discussions of any type is what keeps this community going - appreciated!


Amy

Sent from my SM-G930W8 using STM Forums mobile app

---

Thanks a lot Vortex, really great tips! Also loving the new 40 day tutorial.

That's a good point regarding landers.

What about banners (that are currently profitable) if we're running display? Any rules of thumb to know when we should switch them out?

I usually run the stats on them for the last 3 days and if they're unprofitable during that time period I'll cut them...but I'm not sure if that's the best way to do it.

---

Banners do tend to get "stale" a lot faster than landers, in general.

The best approach I know is to just keep creating new banners so that you'd always have at least a few running, and then just regularly cut inferior ones.

That way you never need to ask the question: "When should I swap out stale banners?" And you won't worry too much about potentially pausing a banner prematurely - because there are always other, better-performing ones still in rotation.

You should aim to test new angles when you find them, and at the same time, create more banners for angles that are working or that have worked for you in the past. Creating variations of your best banners can work well too.

I also liked to take the best banners and create different combinations/permutations using their images and text. (E.g. Say I have 2 banners doing well, I would take banner1's image and combine that with banner2's text to produce a new variation, and take banner2's image and combine that with banner1's text ro produce another new variation, and add them to rotation for a test.)

Hope that helps!



Amy

Sent from my SM-G930W8 using STM Forums mobile app

---

Let me chime in too

Once I have a working campaign, I try to squeeze it as much as I can. So Im not rushing to kill it when the performance drops, because as it was able to work once it was a GOOD campaign.

First thing I do when the performance goes down, is to check that everything works fine ... I talk to my AM or the advertisers to make sure offer works fine. Then I check my setup for possible server problems, domain flagging etc...
If the tech part works fine, I look at the bidding situation ... Many times all it takes to look a specific position in the bidding chain and the campaign can stop performing.
If the bidding isnt the problem, I look at banners ( I also share Amys view here, LPs last way longer) and try to test new ones.
If I still can't make it work again, I start messing with LPs and also different urls of the same offer.
If it still doesnt work, I test different offers with the same funnel (sometimes I do this before touching the banners/LPs).

I only kill the campaign if after a significant amount of work/time/tests it simply won't start performing well again. To give you some rough estimates, I'm ok to spend 2 weeks on this and even if I cant revive it, I will get back to it again after I give it some time to "settle down".

---

Thanks for your feedback, Amy. It is just my expericence.

---

Thanks so much Amy and Matuloo. You guys have given a great checklist to follow. Appreciate it!

I've always been too afraid of running a successful camp into the ground and giving back all my profits by trying to revive a dead campaign but I think I will try harder to revive them. If nothing else I'll learn something new.

---

This kind of task definitely belongs in the "pays back tenfold in the long run" list. Yup.



Amy

Sent from my SM-G930W8 using STM Forums mobile app

---

Yup, giving up too early is a frequent mistake people do in AM, but like always, there are limits to this as well. It's a bad idea to get emotionally obsessed with some campaign and trying to revive it "no matter what". When it's dead, it's dead and there is nothing you can do about it. Campaigns come and go, it's a never ending cycle.

The important thing to realize here would be that when working on a new campaign, it takes several tries to get it right, nobody is capable of putting up profitable campaigns one after another, just like that. It always takes some time. And it's the same when trying to revive a failing campaign, chances are you won't hit jackpot on the first try either, so give it the time it requires, but if after multiple attempts it's still dead, then so be it ... and move on

---

Home > The Newbie Zone > Questions and Answers