When to declare a failed native campaign and forget about it? (11)
When to declare a failed native campaign and forget about it?
Hey STMers,
I'm a native noob doing my baby steps in IM with search arb.
Running though the FA of @jack_l, I've noticed the following (not exact quote, sorry): "The main metrics of success in paid AM are Total Campaigns Started/Average Loss Per Failed Campaign".
That got me thinking about the question "how to decide from the start what is a failing campaign?".
With a clear definition of what's a failing campaign, one can plan and automate the "pause campaign" rule (TheOptimizer,
I would be happy to hear your personal definition for that, and combine your wisdom and experience into a method.
Thanks in advance!
lublub
Here is my criteria for native campaigns:
If there is enough conversion data to estimate the roi, the funnel is rather well put together, yet there are no high performing outliers (really good placements or creatives) - "Fail" if the roi is worse than -75%.
On the other end of possibility, if the campaign starts off at -25% ROI or better, we can probably reach profitability by split testing creatives and offers.
If the campaign starts off between -75% to -25%...it has life but still will need quite a lot of help. I'm looking at testing very different angles, creatives, devices, geos, etc. Honestly these campaigns usually cause the most financial loss, so I keep a close eye on them and try to make every testing dollar count. I'll also pause the worst ~40% of the traffic (measured by cost per landing page click...we want this to be low)
Other factors that contribute to my decision:
-how much potential does the campaign have?
-how are my other campaigns doing?
-how big of a financial loss is it incurring?
-How willing am I to run at a loss to keep the offer "warm"
I hope that helps!
I usually stick with 10x the payout and 0 conversions, then I cut.
This obviously depends on the offer and how much the payout is, but it's a good guideline.
Originally Posted by jaybot
As I’m dealing with search arb, 10x the layout with no conversion never happened to me.
My doubts are regarding how much money to bleed if not breaking even.
I understand that my question is too general, as I assume treating campaigns at -50% ROI or -15% ROI is different.
And in addition, the phase of the campaign (how many days in and how much optimization was already applied) also counts.
Yet, I’m managing about 10 campaigns now in different geos and verticals/keywords and without a clear plan when to give up on any and closing them down I will bleed money in vein.
This is an activity I’m running about 30 days now, so I don’t have enough experience whether optimization (taboola’s Smartbid plus recommended auto rules) will continue and improve after the 30 days mark and beyond.
Sent from my iPhone using STM Forums mobile app
I mainly sell e-commerce products through native, but lately I've been testing search arbitrage on Outbrain with a Bing / Yahoo search feed.
I only have 2 campaigns in a single niche (finance) and I have tested very little (only 3 keywords), but the desktop one is very promising, I have widgets with 100% ROI, but the problem is that it is not very scalable, perhaps for the niche perhaps for the geo. While the mobile campaign has a fantastic cpa, it will never work, because the Bing / yahoo feed is stingy with mobile traffic and the RPM in this case sucks.
If I may ask, what niche do you have the best results on? Please don't tell me that you also give free SUVs ...
Originally Posted by just_g
Then again, looking for a magic formula (or at least a first rule of thumb to build on) when to say "no more bleeding, this campaign will never work", and forget about it.
aaahhhhh this game is tough.
Thanks, @jacekplacek.
I like it that you have a comprehensive strategy, and this is exactly what I aim to build and improve with time.
Does that work for you for search arb (if you deal with it at all), or native in general?
Thanks again!
Originally Posted by lublub
And.
Ah shit. Search arb totally different ballgame.
Let's get @roiter123 and @platinum in here. They're the resident experts on Native search arb.
Originally Posted by lublub
I make it easier and faster, (probably wrong
(For e-comm, not for search arbitrage)
CPC immediately high.
Native algorithms strangely give you the best traffic in the beginning.
(Only mine impression?)
So I run the campaign in the 4 most profitable hours of the day. (morning for Italy).
After those 4 hours if I get ROI I know it will be a great campaign, if I am almost at ROI I know it will be a good campaign, if I am far from ROI or even no conversions, I discard the campaign.
Only 4 hours... Madness for many, but having experience in my "small niche" and my "small geo", I know it's the best method.
Clearly by changing geo, products, etc. I don't apply this, but more the methods described above.
Perhaps a good tip for someone who wants to start with native is to find a niche and a geo where that niche works.
If the ROI after a significant amount of spend (AND rules' optimization i.e. blocking publishers, creatives, changing bids on publishers) is still low I'll start lowering the bid gradually every couple of days and see if I can get some profitable spend going with still some volume left.
Sometimes bidding down will kill most of the volume but I might leave the camp if I see a "not bad" ROI because one day competition might decrease and you'll get some volume going again. Rules will be monitoring this campaign all the time. And I will also be checking on it every once a week or two weeks time-frame.
Home > Paid Traffic Sources > Native