Home > Paid Traffic Sources > Mobile

---

Profits always drop by 3rd or 4th day, need help! (21)

---

I've been running several campaigns, amongst those that are profitable, I've got some hitting 300% roi, 150% roi on a number of traffic sources.

However I realised a pattern, by the 3rd or 4th day after running, profit just tanks. Could literally drop to 10% roi or even red.

Landers don't seem to be the problem (I think?) as its ripped as always and everyone knows them. Not like it has some unique element or a brand new look.

It's as if someone knows exactly which sources I'm running, landers I'm using and bids like 0.5$cpm above me (not exactly, but literally feels like it)

Traffic volume don't seem to be an issue. One geo has tremendous amount of volume and I haven't even hit 3% of it yet. And it's not like I set and forget, been blacklisting placements and all that daily.

What could be causing this? Frequency? Ripped? Backdoor script?

Really struggling to understand, thanks for any advice!

(PS: pops traffic )

---

is it the merchant or the network?

You should try split testing different offers on different networks.

Some merchants or networks may shave if you have poor quality traffic. . . and sometimes, they will never even tell you about the quality.

---

This is a pretty good point. I'd like to add on top of this. The network/offer may only give a cap on the amount of leads you send per day. For example, if you send around 50 leads a day that may be okay under the radar, but if you send 100 then the merchant would come back to the network to discuss your lead quality. Perhaps, they discussed that your lead quality isn't that strong, therefore they decide to shave some of your leads.

It's important for you to develop rapport with the network/Affiliate Manager to discuss your lead quality and potential payout bumps. This scenario happened to me today: I was trying to ask for an even higher payout bump from the advertiser, but the merchant discussed with my AM that my leads are solid and they want me to generate 100 leads consistently before they payout bump me. This is important for me to know because that means I have a goal to reach 100 leads a day to get a payout bump. It may not be a significant payout bump, but a little goes a long way and that's what I would like to hit profitable campaigns.

---

I've just asked my AM, let's see what happens.

One of the sources I'm using is Zeropark. I used to start out with 300% ROI on the first day. It starts dropping on the 2,3rd day..and today it's effing RED (negative).

WAY more than enough volume on this geo, what's going on? Is there someone out bidding me or something?

---

@alfiss - what else changes? Does CTR drop? Does your minimum bid increase?

And what's the volume you're running here? The potential causes are very different if you're running 10 conversions a day vs 1000.

Finally, if you're running ripped creatives, I'd definitely try recreating those from scratch just to eliminate the possibility of rogue code. Sounds very unlikely, but worth ruling out.

---

Hey Caurmen, I realised the amount of traffic clicks dropped too (same bid), by around 15-30%
Bid stays the same, since in this case it's Zeropark.

CTR dropped from 16% to 11%, not too sure if I'd place TOO much emphasis on this since I dont think data is significant enough. (around 25k visits)



Approximately 44 conversions on a specific traffic source dropping to 33, then dropping to 20~, while spending around the same a day.

All codes has been tested and checked for any backdoor javascripts, manually tested by refreshing too, nothing so far.

---

Slowly with the conspiracy guys Im not saying shaving doesnt happen, but thats most likely not the case here.

I see this happen a lot, especially when running banner traffic. You are running pops, so the most probable reason is you are being outbid indeed.

What happens is that you create a campaign and step into the face of someone running an optimized campaign in the same spots. It takes them a day or two to realize you took their traffic, they start to optimize, raise bids slowly to figure out where you are, hence the slow decline and not a sudden death.

The tricky part is, they dont have to have the same targeting as you do. So your traffic might not go down so much, and you will keep wondering what happened when you are getting almost the same traffic but the CVR went to hell. Chances are, they have whitelisted sources, so you will only loose a part of the spots. And they can have carrier targeting, device targeting etc in place. To sum it up, you will loose a relatively small % of the traffic but a large % of the profits, because your competitor is bidding high for the best traffic available, taking it away from you and leaving you with the crap

---

The more I'm seeing my stats, the more I agree with you.

Traffic has been in decline (no reason to since the geo is so big), and conversions is acting like i'm just next in position after them.

What would you suggest from this point?

"they start to optimize, raise bids slowly to figure out where you are, hence the slow decline and not a sudden death."
How do they determine or estimate what exactly my bids are?

---

Imagine that your competitor has a select amount of zones that he wants to buy from and he buys from those for some time already, so he knows how much he can get there per day, more or less.
You took the traffic away from him, so he raises the bid slowly, to see when he returns to the previous volumes. The he knows he took the first place again and how much approximately you are bidding. This is where experience kicks in.

You should analyze your data and see if you see a drop somewhere, like no traffic anymore from a zone that used to do great ... if you find them, target them with a whitelist campaign and bid higher.

---

Jesus, this reminds me of someone literally staring at the computer the entire day?

For someone who's relatively new in this industry, would you recon fighting it out and trying to 'win' this war? Or just move onto easier/greener pastures? (Or are those extinct already)

---

Its not about staring at comp all day long When you have a ton of campaigns, you can only optimize THAT much per day You do the changes in steps, you raise by a bit, wait some, check again, raise by another bit, wait ... thats why it takes a longer time to figure out where the competition is bidding. On some sources, its the only way to go tho. Im playing this game on one very large spot with 2 other competitors for about 6 months already, sometimes I win, sometimes they do ... classic bidding war

Hard to say whether you should fight the war right now, after all it might not be the case at all here, but whatever skill you gain now might come in handy later. I would at least try to play with the bid for some time to see if what I suggested was the reason or not.

---

@matuloo, I'm currently running a day-parting campaign due to budget restrictions that I have. But, I'm keeping it consistent and running at times where I've seen the best conversions at a specific time. I'm not running it like 2PM-6PM and then 9pm - 12AM. I'm running it more like 5PM-5AM.

The thing is I started this day-parting campaign on Monday and ran it again today and will be continually running it for at least 1-2 weeks straight. Yesterday I had 32,527 visits and today I had 23,685 visits. This tells me that I'm competing with someone else I think. I'm not sure if I should up my bids to get more exposure to the campaign because I don't want to skew any data that I currently have. I'll probably bid .003 or .004 higher.

The difference between my campaign today vs yesterday was I added a sound file and a translated page. So, therefore it shouldn't mess with the "visits" volume as much since nothing significant has changed about my loading speed.

For tomorrow's campaign I'll be rotating offers to see any drastic changes in conversions or adding 2 different landing pages to see if I can spot any differences as well. What do you think about all of this?

I've also noticed that I ran the same campaign with the same bid price and one placement was doing very well until it just disappeared after I ran it the next week. Did this placement disappear because a competitor found it to be a profitable placement and tried to bid the highest on it so that it wouldn't appear under my radar?

---

A CTR drop of 5% absolute and 33% relative (5% is approx. 33% of 16%) across 25k visits is certainly enough to be statistically (and otherwise) significant. It'd also fit very well with matuloo's hypothesis of what's going on - as you're being outbid you're getting crappier traffic.

It's also worth remembering the "champagne CTR" rule - over the first couple of days of a campaign CTR and profit will usually be higher because it's new and people haven't seen it before. Less applicable in the case of ripped landers, though.

---

I wouldnt be so fast with assumptions here again, monday can be a totally different day than tuesday. I usually see higher traffic volumes on mondays compared to tuesdays in the spots Im in, it probably has to do with the fact its the first day of the work week and people dont yet concentrate on work fully, or maybe they simply surf more at work... You need to run at least 2 or 3 weeks nonstop to really see these patterns. The fact that you are day parting is going to make it a bit skewed and more complicated as different parts of the day can be less or more traffic heavy. I would suggest to wait more to see if the traffic comes back or not, the difference between days of week can really be big.


If you think you have enough data for the original setup, then yes by all means, test different stuff.

Could be true, but in that case you would most likely get at least a few hits from there - not many sources give all the traffic to one bidder, even tho its the highest one. There are other possible reasons for that : the placement might have been sold on a flat deal, it might be gone from that particular traffic network, it might have lost its traffic ...

---

That makes sense now that the CTR of a campaign would usually be higher in the beginning stages of a campaign where I narrowed down the campaign to a winning lander and offer.

I'm running the dayparting consistently within a specific timeframe and thinking about running it for about 2-3 weeks to get significant data. I'll be also possibly changing bids day to day to obtain placements that are proven winners and then blacklist those placements and put them into a whitelist campaign. That's the most I'm worried about in terms of skewing data, bid adjustment. At this point, I'm running one lander and offer, but reduced the bid at least by .0010 today. Should I be bumping my bid back to what is was originally?

In terms of getting enough data to test offers and landers, I have found a winning lander and offer in the beginning stages. Is that what you mean "If you think you have enough data for the original setup, then yes by all means, test different stuff"?

I started a whitelisted campaign to see if I can get any traffic from those placements that disappeared. Awaiting for the traffic source to approve campagin

---

Are you launching on weekends?

---

I'm planning to.

---

What I mean is, on weekends, you can usually get a bump in conversions. Then, during the week it declines. This can account for the drop in sales. I've had campaigns that are +200% ROI on weekends, but breakeven on Weekdays...

---

I remember running this campaign on a Friday and was getting crazy conversions on a placement, and then on Saturday it just disappeared. With that being said though, I'd probably have to increase the bids since on the weekends the conversions are higher and I'll be competing with more advertisers.

---

You will always need to play with the bids a little bid, its pretty much impossible to let a campaign run for 3 weeks and not to touch it Dont worry about it too much, just dont do chaotic changes, like double the bid in the morning, then get back to the low one again and then freak out and triple it in the evening again Take it one step at a time.

Yes, if you are sure you already identified the best combination from your current setup, then move on and try to modify it to perform even better.

---

This is the action plan that's taking place right now:

1) Winning Lander and Offer running on a moderate lower bid from the bid I had before just to get a full day of traffic.
2) After three days of full traffic(as well as blacklisting bad placements) on that one particular bid. Create a whitelist campaign to separate the profitable placements to bid higher on those targets.
4) On the 4th day, I'll test various landers to see if the winning lander still stays on top being the winning lander.
5) On the 5th day, I'll raise the bid to 20-25% from the previous bid for another three days and blacklist any placements that aren't good. On top of that blacklisting profitable placements and putting them in my whitelist campaign.
6) Repeat steps 4,5 continuously until I reach a campaign that I'm confident about. So this is probably gonna be a month's worth of process.

---

Home > Paid Traffic Sources > Mobile