Home > Paid Traffic Sources > Facebook & Instagram

---

ad account learning (9)

---

I have an ad account that has been spending about 100 a day for about 5 months, 2 days ago i dulicated an ad set and set the campaign budget at 500 a day (so i only had the one brand new ad set running in this CBO campaign, so this ad set had a 500 budget behind it since its creation) and the performance of this ad set was very poor, i had ran many other dupes of this ad set and the performance has always been good. Im aware this just could have a dud or it could have been something else but im wondering if you think that the raise in budget could have caused the bad performance since the ad account is not used that that high of a budget? or do you think that it was just the campaign that was wasnt used to this budget and performance would have been good if i duplicated this ad set into a new campaign and set that campaign at a higher budget?

Thanks in advance!

---

One possibility that immediately comes to mind: What is your audience size? If the audience size isn't large enough, and/or your ads/funnel has saturated the audience, then setting a higher budget can result in such drastic decrease in performance.

Basically what the algo does is it would try to target what it feels are the people that would be most likely to convert well, first. When you increase the budget, the algo may run out of the most suitable audience to target, and resort to targeting people that aren't as good a match - resulting in a drop in performance.

I don't have a magic formula for setting a good budget. It's an art trying to find a good balance between having a low budget that restricts data collection and a high budget that forces the algo to resort to targeting people that aren't the best fit.

The only advice I have would be to set up another campaign at a lower budget to see what happens.



Amy


Sent from my iPhone using STM Forums

---

Also: Am I correct in assuming you only have one adset running in this new CBO?



Amy


Sent from my iPhone using STM Forums

---

Appreciate the reply! My audience size is 63 million so i dont think that saturation is an issue yet, and yes i only have one adset running in this new CBO. So you dont think that the algorithm is under performing because its not used to higher budgets?

---

One thing you can try is duping the adset exactly the same as the one thats performing well and increasing budget by 15-20% a day if it's performing consistently. From my experience the magic number is around the 20% mark.

---

thanks for the reply, do you find that keeps it out of learning limited even if its not getting enough conversions to equal 50 a week?

---

Why not duplicate that adset a few times within the CBO? That way you hit more audience pools, and the algo would send more budget to the better-performing adset(s).

There's really no point in putting a single adset in a CBO. The whole power of the CBO lies in the automatic adjusting in budget allocation across adsets.



Amy

---

Yeah thats what i was doing before but i was having trouble getting out of learning limited so thought that having just one ad set would help get out of learning limited, do you think that that would help?

---

I want to ask more questions because I'm a bit confused:

1)Does this adset contain one or more ads that have been TESTED and PROVEN to perform well? (Usually in an ABO - before being put into a CBO for scaling.)

2)If you want to help an adset to get out of the learning phase in time to avoid learning limited - why not just set the initial budget accordingly? If you've tested the ads in an ABO previously, you should be able to estimate this initial budget.



Amy

---

Home > Paid Traffic Sources > Facebook & Instagram