Home > The Newbie Zone > Questions and Answers

---

Results consistently dropping off? (9)

---

Hey everybody,

I keep running into the same problem and I'm wondering if anybody out there can help me.

Basically every time I start a NEW campaign I'm almost always pleasantly surprised to see that I've made some conversions right out of the gate (nowhere near profitable but still).

The problem I'm having is that these conversions seem to just disappear after a few days.

I understand the idea that even if I get ONE or TWO conversions that that doesn't necessarily mean my campaign will be profitable it could just be a fluke.

But this seems to be a consistent trend?

Even with high payout offers (like 45$ USD) I can get some action with a brand new camp.

I'm wondering if there is something obvious that I'm missing here...

Thanks

P.S. I've been running ClickBank offers on FB if that is relevant.

---

I don´t run FB but I also see such behavior for pop and push traffic.

In my case it also depends a bit on the bid type I use.

For example when I run a pop campaign on CPA then I mostly receive high quality traffic and higher bids in the beginning.

This can lead to several early conversions.

Then the algorithm starts working and adjusts the bids and volume step by step depending on the results.

When my campaign then isn´t converting good enough = not making enough money for the trafficsource the volume drops and will result in less conversions.

When then additionally also the bid for the placements get decreased it can also result in lower quality traffic.

Don´t know if it helps but you see that such behavior happens others as well

---

I've seen this trend across different traffic types and platforms. It's like the platforms want to send you a burst of good traffic first to 1)access how well your offer converts (then send you lower-quality traffic if your offer doesn't convert really well) or 2)get your hopes up so you'd spend more money instead of pausing right away. Of course I have no proof of this - these are my personal speculation.

The only advice I have is to make major campaign decisions by observing more long-term performance, i.e. not just based on performance on the first day alone.

On some platforms in the past (not FB) I would keep setting up new campaigns just to take advantage of that initial burst of good traffic. But I doubt FB would like that approach very much.



Amy


Sent from my iPhone using STM Forums

---

Should you discard the first few days of results, since they may be artificially skewed by the traffic source?

---

I think there are two options here:

1. Something along the lines with what Amy (vortex) wrote above. I've seen it happen so often that I also tend to believe new campaigns are getting some kind of a boost. Be it from the traffic source's side in order to make you spend more, or from the advertiser's side, again to make you send more traffic. I know it's just a theory, but it has happened to me so many times that there might be some truth to it.

2. This one is more likely IMO. The funnel simple wasn't strong enough. It was not bad, so it managed to grab the interest of the most interested buyers, but it was not strong enough to get the interest of a significant part of the audience. I like to use the fishing analogy... any lure can catch some fish, the most hungry or aggressive ones, but to really catch a lot of fish, you need to have the perfect rule and master the technique.

To put it short, you either need to use better creatives or find a better converting offer.

---

You could do that yes - decide where the drop off is and just exclude the data from any major campaign decisions.

Another thing: If you're finding this to be a consistent occurrence, raise your bar when it comes to how good a campaign needs to be to "make the cut" for further optimization.

E.g. Let's say you want to test a new offer on a placement whitelist/blacklist, and your original criteria was that an offer+lander needs to do at least -20% ROI in order for you to even consider optimizing further. So what you can do here is raise the bar to 0% ROI.

Alternatively: Just run that -20% ROI campaign a few days more to see if there's a drop off, and go from there.

(Note: I pulled that figure out of my ass - please don't use it as your cutting criterion. Every person and their methodology is different.)

To summarize: You can anticipate and factor that decline into your methodology.



Amy


Sent from my iPhone using STM Forums

---

Let me tell from my experience with CPA/CPA Goal campaigns once more

A usual progress is like this:

- Receive expensive and high quality traffic in the beginning
- Get early conversions to check if campaign is converting good enough
- Drop in performance when optimization starts and bids get adjusted
- In the next few days performance increases again and campaign keeps running more stable

---

This is interesting.

I'd been using CBO on FB and just flicking off ad sets after I'd spent enough money on them. I thought this was a pretty low risk way of doing things, but truth be told I don't think I'd put much thought into how this would effect the quality of traffic that I'm bidding on.

Thanks twinaxe, given me something to think about.

---

I see it like this, when I want to test a campaign I can choose between two approaches:

1. Bid lower on mediocre-good traffic to don´t overspend and keep the test running for longer
2. Use only high quality traffic and bid high for faster results

Both approaches can work but mostly I prefer the second approach.

Yes, it can be more expensive but I want fast results and paying high prices for really good quality is often still cheaper than bidding less for lower quality.

And I think trafficsources are doing it in a similar way.

They have data from thousands of campaigns, tens or hundreds of thousands of conversions and when they send traffic to test the performance of a campaign they also send expensive but high quality traffic.

A difference between trafficsources using their conversion data and me using my own data is that their data is from many different affiliates, different campaigns, different performances.

They basically only see that a placement converted many times so it´s good to use for a test.

When my campaign has much lower payout or isn´t performing as good as other campaigns that received traffic from that placement then it happens that I also receive conversions but the bid is too high for my campaign.

Nonetheless such tests can tell enough about the campaigns and if they have enough potential to keep them running or not so that then the optimization process for my specific campaign can start, bids get adjusted and so on.

In most cases I first have to eat some losses but longterm it´s still worth it.

This wouldn´t work when a campaign lasts only 2 or 3 days or so, then it would have to be profitable right away because otherwise you have no time left to compensate the first losses and bring it into green.

But when I have to take some losses first, then optimization goes for few days and then the campaign settles to more stable performance and lasts few weeks or so then it´s absolutely worth it.

---

Home > The Newbie Zone > Questions and Answers