pop bidding strategies (9)
What are everyone's bidding strategies for pops? I usually kill placements that don't convert at 1.5x offer payout spent, however if I bump my bid up to spend my daily budget, I will reach my kill rule for a placement much quicker, and may kill a placement that is potentially profitable at a lower bid. How do I take this into consideration?
I am currently bidding $11/k
thanks!
It depends on how the traffic source works and how profitable my campaign is.
But usually I start with whatever I think the average bid is and I make a note about the quality. Then I go very low and see how shitty the quality is on a low bid. Then I try to raise the bid and see if the traffic improves. I keep raising and lowering the bid until I find a good spot that I feel like is a good balance of volume and profitability. Basically I just keep playing with the bids until I find something that works.
The above is strictly for pops and not ppv.
if you dont know, follow the mobile guide and go 3-4x payout.
i think it depends on the nature of the traffic, maturity of campaign.
usually if the campaign is somewhat mature, the cuts are tighter because we already know it works.
if the traffic is good quality usually u can afford to be tighter. but if you are buying pretty crappy traffic, you will need to gather more information, but the cost where you cut should be at about the same point
i do my cuts based on relative placements, i have a group of placements that are performing better that i compare to, and decide whether this placement is worth keeping or not based on its potential to increase in ROI.
I don't like cutting placements unless it's plainly obvious that the placement is losing money -- and will always be losing money.
Let's say you cut a placement that is delivering 10,000 impressions and generating $0 revenue.
If that placement represents 10%+ of your total traffic, ok, maybe you'll see an improvement in ROI.
But it's still only 10% of your campaign.
You'll never be able to cut the 10,000 placements offering one impression per day and generating $0 revenue.
That's another 10% of your campaign. And it can't be blacklisted.
The only way you can influence 100% of your placements is by improving the funnel itself.
(Or by sacrificing scaling altogether and running whitelists.)
If you start spending your days combing through rows of data, reacting on autopilot: "bad placement, bad placement, bad placement" -- we might never see you again.
You can have much more impact by improving the funnel exponentially, or by using a sweeter angle that cuts through those 'bad placements' and squeezes a few conversions out of them.
That's not to say you shouldn't check performance of individual targets.
But I would personally only bother removing the stinking corpses.
Not certain if this is a Pop bidding strategies, rather pop campaign strategy, either-way i have a question that relates to this when it comes to cutting placements and bidding.
Example: I am running both Vouchers Sweeps (Ex. Supermarket) and Gadgets (Ex. Win iPhone/Samsung Galaxy/Macbook, etc). Offers could be LeadGen, CC submit or PIN submit. Started out running the 'voucher' offers on 2 campaigns (wifi v.s. 3G) and rotating the different offers and landers in it to see which one stands out. Same goes for 'gadgets', leaving me with 4 campaigns at start. (In the case placements work great withing a campaign i am creating yet another breakout campaign to monitor costs in
-My idea behind splitting Vouchers and Gadgets was that the conversion behaviour could vary per placement. (?)
-My idea behind splitting 3G and WiFi was that the conversion behaviour could vary per offer (PIN vs Leadgen)
-My idea behind grouping different offers with similar promotions in 1 campaign was that placment quality would overlap.
All this leaves me with a lot of campaigns-->data-->optimization-->time spend
How would you guys approach this? Are you running a campaign per offer, putting all kinds of offers within same campaign, or s'thing in between ?
Usually one would run a mix of traffic to multiple landers and offers in rotation and let the data do the talking.
This is what drilldown reports in trackers are for after all.
I cut every placement that gets to 10,000 impressions and zero conversions + Increase the bid with 10% on every placement that has a ROI of 150% (i.e. spend 1$, make 1.5$). We always aim for a total margin of 25% with max. volume we can buy. As said before regarding campaigns that work EXTREMELY well on pops are offers with a big brand name where you can get conversions to happen on the cookie - you need to be lucky enough that the advertiser allows pops though.
in the end we are going for max profit, all these placement cuts, bid change are just intermediary metrics to try to get max profit. there are so many things you can test, but as a rule of thumb i try to keep everything in just 1 or 2 campaigns but work alot on it. i think our campaigns usually has less than 10-20 placement cuts, its just the really crappy ones.
for testing bids, i like to use the "computer search model". lets say the minimum bid is 10c, and the top bid is 0.50. i like to go 0.60, which usually fails, and then i go halfway at 0.35, and then using 0.35 data, i try to figure whether the next test will be
1. between 0.35 and 0.10, which is around 0.22
2. between 0.35 and 0.60, which is around 0.48.
this method usually misses on the first high bid, but quickly finds u the "sweet spot"
Originally Posted by Finch
Home > Paid Traffic Sources > Mobile