Home > The Newbie Zone > Questions and Answers

---

Is there a reason why campaigns drop off after the first day? (8)

---

Hello,

I keep running into a similar situation every time I run new pop campaigns. Things will look very promising the first day and get dozens of conversions. However, then the next few days it just nosedives into red with a few or no conversions. This happens without any optimization or cutting placements on my end.

What changes between the first day of the campaign and the days after? Is it bots? It has happened enough times with different campaigns, verticals, and offer flows for me to notice and it not be a coincidence. The only differing thing I can think of is just the day of the week.

Any insights into this occurrence?

---

Hey
Numbers of clicks is dropping as well?
Did you check this traffic maybe there are a lot of bot traffic?
Have you tried to raise a bid?
You get this traffic from ad networks?

---

When this happens, there are usually 2 reasons for this. We've discussed this here on the forum quite a few times but I can't find of those threads now... @twinaxe do you happen to remember some of those?

Reason 1 and the most likely one: Your funnel is not strong enough. It could get a few random conversions with limited audience, but once more clicks go through it fails. You need stronger creatives or LP or offer.
Reason 2 : sometimes (often) it looks like the traffic source send better quality traffic to new campaigns, in order to motivate the advertiser to spend more. We can't really confirm this, but it happens very often. So you would see a few conversion soon after launching the campaign, but it will quickly cool off.

It can also be a natural cycle of the algo, depending on what bidding model you use. So for example in case of CPC, it's normal to receive better traffic initially, when the algo is trying to asses the quality of your ads and their CTR... once they have the date, the rotation changes.

One way or another, you need to improve your funnel so your CVR improves and you can afford to bid higher to get the better traffic.

If you can tell us more info about the traffic type you work with, we might be able to share more advice.

---

I´m not sure if I think about the same threads as you but as far as I remember we had such discussion recently about CPA campaigns but maybe it also was about campaign performance in general.

There we were also talking about better quality forst to keep advertisers happy and keep campaigns running.

I´ll also try to find it

---

We've discussed this a few times... you know the mysterious initial conversions when you start a new campaign

---

I've been suspecting @matuloo's "Reason 2" for years.

This is why it's so important to set the bar high when deciding whether a campaign is worth optimizing.

Many new affiliates would take a -70% ROI campaign and try to optimize that to green by cutting - meanwhile campaign performance continues to drop, so the optimizations may JUST be good enough to counterbalance this drop - to maintain the same -70% ROI.

This is why it's important to spend more budget on testing a ton of offers and landers and ads - to find a REALLY strong combo that is close to green or in green right off the bat, and THEN use optimization to get green and combat this natural drop in performance.

Another thing you can do is exclude the initial data from the "nice burst of initial high-quality traffic" from consideration altogether. i.e. Only evaluate the campaign based on its "real" performance AFTER the performance drop occurs.

It doesn't happen every time, but enough to be annoying. It also seems to be worse for some geos than others. If you ever find a solution, please do let us know! Before then, we'd just need to learn to work with it.



Amy

---

Could also be an interesting topic for a small case study.

Start campaign, grab high quality initial traffic, stop campaign, duplicate, grab high quality initial traffic, stop campaign, duplicate and so on and so forth.

---

I used to do exactly that. Then traffic sources smartened up and closed that loophole. But there may still be networks out there where this can be done.

That approach would only be worth the effort if there is night and day difference in traffic quality - UNLESS you come up with a script to automate the process.

(The same can be tested on CPA camps - some networks would give a burst of good traffic to essentially "scope out" how well your campaign can convert, before their algo decides whether it would be worth sending good traffic continually. There are opportunities to manipulate the algo here other than just duplicating camps to take advantage of the initial burst of good traffic. I don't do blackhat anymore so don't know if the loopholes still exist, and on which networks. This is as much as I will say on that topic.)


Amy


Sent from my iPhone using STM Forums

---

Home > The Newbie Zone > Questions and Answers