Our Brief Adwords Experiment - What Happened??

[KiddieToysReview]

Would greatly appreciate if those with more experience can have a quick look at our recent, brief, adwords campaign and point us in the right direction. What should we do to get lower CPV and higher subscribers?

I've shared everything I think is relevant below. I stopped the campaign after 2 days as I think the average CPV was way too high at $0.13.

I am also not sure why the google keywords suggest tool shows these keywords at $0 but Youtube actually charged $0.12 - $0.16

Also, I thought that the campaign should result in a greater subscriber count, which it did not. This probably means the video was not enticing enough to get people to subscribe, or people who view kids channel video ads would probably not subscribe to a promoted video's channel.

On a positive note, I think the 40 earned views (40% of 100 total views) was kind of nice, and it's good that people checked another video after the promoted one. But the fact that only 1 subscribed to our channel is disappointing.

The main campaign cost $13.22. For the first day there was also the in-stream campaign (this was a mistake - that's what is showing up as the additional cost to make it $18.57) and I deactivated it as soon as I realized it (hey it was my first time with adwords).



The video we ran the ad on:



These are the keywords targeted and the average CPV that was charged:



According to the keywords suggest tool, the CPV should be 0, or close to that anyways:


The age targeting:



The parent targeting:

 

[toywithtoys]

I noticed that you excluded the 65+ age group. You don't think they watch or play with toys at this age?

 

[FamilyToyReview]

Not sure if you already did but i excluded unknown from gender too.

And if you got .13 cents a click from adwords recommended i would ignore that. And just do 1 cent a click. I dont know where you had set your bid prices. Were they set at 42 cents? Thats too high i think. Not sure where the prices were set. Theres a few areas but i would delete and replace with 1 cent for the entire ad. And if youndont get any views then would incrementally increase to 2 or 3 cents. You already have performance history on that video/ad so i think your ad will have exposure even with 1 cent. Especially since the thing is telling you cov shoukd be zero.

If 1 cent then $13 would have given you closer to 1,000 views and possibly 400 earned and maybe 10 suscribers?

By the way your first test turned out very well. 40% earned views is excellent. I read in other post someone mentionig 30% is good. Ive achieved 30 to 50% so im guessing it doesnt get too much better than that. And 1 sub out of 100 is cool too. I wouldnt think there would be any out of only 100 test case. I get a couple thousand views to my channel per day and on some days i barely get a new subscriber. But i do remember when i did adwords and spent like $5 a day so like 500 views i would gain a few subscribers per day. Maybe not 5 out of 500 tho.

But yeah need to change your ad price as low as possible. but everything else looks great.

I think excluding 65 is ok. There are probabky viewers from 65 but ive learned being as frugal as possible allows you to spend the same dollars on more profitable areas. With 65 Theres higher chance you could spend $ targeting kids temporarily using grandparents youtube account during a visit, instead of their usual parents account. And if using grandparen account then probably less likely to subscribe. There are probably just older parents too. But is rather spend those ad dollars in the age groups with highest density of parents.[DOUBLEPOST=1458494412,1458494055][/DOUBLEPOST]by the way looks like the keywords might not be too popular so there might be a natural low number to the views.

and i think modifying the thumbnail might increae how many people view the ad. If im looking for thomas accidents will happen im probably going to choose the video with thomas having accident in the thinbnail. As big as clear as possible. You can always change it back after the adwords test.

 

[KiddieToysReview]

I noticed that you excluded the 65+ age group. You don't think they watch or play with toys at this age?

Hahahaha... my reasoning was that grandparents may not have YT accounts, and the kids would bring along their own or their parents tablets when they come to visit the grandparents.[DOUBLEPOST=1458505520,1458504090][/DOUBLEPOST]

Not sure if you already did but i excluded unknown from gender too.

Yes I excluded all the unknowns from all the categories, to get as precise targeting as possible.

And if you got .13 cents a click from adwords recommended i would ignore that. And just do 1 cent a click. I dont know where you had set your bid prices. Were they set at 42 cents? Thats too high i think. Not sure where the prices were set. Theres a few areas but i would delete and replace with 1 cent for the entire ad. And if youndont get any views then would incrementally increase to 2 or 3 cents. You already have performance history on that video/ad so i think your ad will have exposure even with 1 cent. Especially since the thing is telling you cov shoukd be zero.

I set it at 1 cent, nothing happened for several hours. I then increased to 3 cents, and nothing happened as well. Then I set it back to the recommended 42 cents, which was a Youtube recommendation it gave. I figured it would incrementally increases until it hit. Looks like it started to hit at the 12 cents mark. I though it would stabilize then perhaps decrease. Since it didn't I paused the campaign as the 12+ cents is too high.

By the way your first test turned out very well. 40% earned views is excellent. I read in other post someone mentionig 30% is good. Ive achieved 30 to 50% so im guessing it doesnt get too much better than that. And 1 sub out of 100 is cool too. I wouldnt think there would be any out of only 100 test case. I get a couple thousand views to my channel per day and on some days i barely get a new subscriber. But i do remember when i did adwords and spent like $5 a day so like 500 views i would gain a few subscribers per day. Maybe not 5 out of 500 tho.

Great that's reassuring then. I guess the campaign went well except for the high CPV price.

But yeah need to change your ad price as low as possible. but everything else looks great.

I'll restart it again with the 2 cents CPV and give it more time to work, I think I was too impatient and expected traffic within an hour. Perhaps it takes half a day or a day to work at those low prices.

and i think modifying the thumbnail might increae how many people view the ad. If im looking for thomas accidents will happen im probably going to choose the video with thomas having accident in the thinbnail. As big as clear as possible. You can always change it back after the adwords test.

True, yes. I have "brightened" up the thumbnail, and going through all my video stock to fix up any that don't look bright. The problem was this was one of the very first videos we filmed about 2 months ago and I don't have any quality shots of the Thomas accidents action. If I use a screencapture from the image editing software of Thomas in the act of falling, it's all blurred as it's filmed in 25 fps. Also, I want to keep the kids or at least 1 kid on as many thumbnails as possible, to make the channel more personal for the kids.

Thanks a lot for your help and useful suggestions. I'll restart the campaign and report in a few days how it is going!

 

[toywithtoys]

The problem was this was one of the very first videos we filmed about 2 months ago and I don't have any quality shots of the Thomas accidents action

I don't understand why the Thomas accidents thing were so popular. Do you know why?

 

[KiddieToysReview]

I don't understand why the Thomas accidents thing were so popular. Do you know why?

I watch the kids play by themselves (without cameras) and they have the most fun crashing things into each other after a few minutes of "normal" playing. Same for me when I was young I used to love to crash Hot Wheels cars together. And all those "crazy Russian crashes" videos. Everyone wants to see things crash, and Thomas is this really nice character it's fun to make him fall down. I guess.

 

[FamilyToyReview]

Hahahaha... my reasoning was that grandparents may not have YT accounts, and the kids would bring along their own or their parents tablets when they come to visit the grandparents.[DOUBLEPOST=1458505520,1458504090][/DOUBLEPOST]

Yes I excluded all the unknowns from all the categories, to get as precise targeting as possible.



I set it at 1 cent, nothing happened for several hours. I then increased to 3 cents, and nothing happened as well. Then I set it back to the recommended 42 cents, which was a Youtube recommendation it gave. I figured it would incrementally increases until it hit. Looks like it started to hit at the 12 cents mark. I though it would stabilize then perhaps decrease. Since it didn't I paused the campaign as the 12+ cents is too high.

View attachment 34699



Great that's reassuring then. I guess the campaign went well except for the high CPV price.



I'll restart it again with the 2 cents CPV and give it more time to work, I think I was too impatient and expected traffic within an hour. Perhaps it takes half a day or a day to work at those low prices.




True, yes. I have "brightened" up the thumbnail, and going through all my video stock to fix up any that don't look bright. The problem was this was one of the very first videos we filmed about 2 months ago and I don't have any quality shots of the Thomas accidents action. If I use a screencapture from the image editing software of Thomas in the act of falling, it's all blurred as it's filmed in 25 fps. Also, I want to keep the kids or at least 1 kid on as many thumbnails as possible, to make the channel more personal for the kids.

Thanks a lot for your help and useful suggestions. I'll restart the campaign and report in a few days how it is going!

Maybe you can try the 42 cents and work your way down and see if adwords still serving with lower bid prices.

Maybe one of the probs with the bid price was that the keyword used was a little too specific and didnt get a lot of hits. Only 2,000 impressions per day so 1,000 per day. Its a pretty small search and i am pretty sure no one else was targeting that keyword. The ads that usually show up in search i think are just general ads for toy category or demographics and i dont think theyre bidding on the same keywords.

On my first campaign, i had too much expectations from adwords and i customized the crap out of it. Made it very specific, super targeting. And i must have overdone it because i would not get any impressions or views even if i considerably increased bid prices. I think i made the pool too small for targeting.

So now i dont make it too tight. i make the pool reasonably larger, but still do some filtering to save money. You should experiment adding more keywords. Perhaps someone searching for thomas train will still be interested in an accidents will happen video? Or a thomas train toy keyword. The search volume shoukd be larger and probably easier to have cheaper bid rates. Hopefully you can still maintain a decent earned views and retention rate or might not be worth opening that pool too much.

But yea maybe the 1 cent 3 cent doesnt work all the time. maybe there are some keywords that dont go that low. My personal preference is to stay below 5 or 10 cents a view.

 

[KiddieToysReview]

Maybe you can try the 42 cents and work your way down and see if adwords still serving with lower bid prices.

Maybe one of the probs with the bid price was that the keyword used was a little too specific and didnt get a lot of hits. Only 2,000 impressions per day so 1,000 per day. Its a pretty small search and i am pretty sure no one else was targeting that keyword. The ads that usually show up in search i think are just general ads for toy category or demographics and i dont think theyre bidding on the same keywords.

On my first campaign, i had too much expectations from adwords and i customized the crap out of it. Made it very specific, super targeting. And i must have overdone it because i would not get any impressions or views even if i considerably increased bid prices. I think i made the pool too small for targeting.

So now i dont make it too tight. i make the pool reasonably larger, but still do some filtering to save money. You should experiment adding more keywords. Perhaps someone searching for thomas train will still be interested in an accidents will happen video? Or a thomas train toy keyword. The search volume shoukd be larger and probably easier to have cheaper bid rates. Hopefully you can still maintain a decent earned views and retention rate or might not be worth opening that pool too much.

But yea maybe the 1 cent 3 cent doesnt work all the time. maybe there are some keywords that dont go that low. My personal preference is to stay below 5 or 10 cents a view.

Yep running many experiments at the moment. I plugged in all the tags from my videos (running experiments on 5 videos across various categories). My tags are pretty optimized I think on the videos I'm running, and so far seem to be getting 4-6% click through and 1-3 cents per bid, so going good I think.

A problem I'm seeing is everything is bunched into 1 campaign. I think I need to separate each video category into its own separate campaign and run it like that. So 1 campaign will be Thomas, 1 playgrounds, basically a separate campaign for each playlist. Each campaign would share its own unique keywords, in addition to common channel keywords across all the videos. It will be easier to manage. I'm thinking of running each campaign Thu-Sun (high traffic times) with a different campaign each week.

 

[FamilyToyReview]

Yep running many experiments at the moment. I plugged in all the tags from my videos (running experiments on 5 videos across various categories). My tags are pretty optimized I think on the videos I'm running, and so far seem to be getting 4-6% click through and 1-3 cents per bid, so going good I think.

A problem I'm seeing is everything is bunched into 1 campaign. I think I need to separate each video category into its own separate campaign and run it like that. So 1 campaign will be Thomas, 1 playgrounds, basically a separate campaign for each playlist. Each campaign would share its own unique keywords, in addition to common channel keywords across all the videos. It will be easier to manage. I'm thinking of running each campaign Thu-Sun (high traffic times) with a different campaign each week.

Yup thats what i ended up doing, separating campaigns. Although you can also separate into ad groups and have diff keywords for each ad group. Ive tried both groups. Either way works i guess.

And yeah after i had enough data i started using the scheduling to only run the ads during the days and hours i was getting most performance. And to me i was scheduling for periods of most subscribers, or other activity like a lot of playlist additions or likes. But mostly around subscriber growth. So i didnt waste my views during other hours. I think i ended up turning off a weekday or two for optimal efficiency.

Aside from that i guess not much else to optimize. And unfortunately cannot just set it and forget it. Because one day you might start getting very low retention rate (i dont know why) or your earned views might drop by a lot. And then youll discover that youtube resorted your suggested videos so youre not earning as much earned views as before. And then i would pause that ad for a while and use another good video that fits my adword preferences. To get most bang for buck.

 

[TZJINZO]

I tried adwords and got a lot of views, but they were such low retention I had to delete my video. How is your retention looking?