Home > Facebook & Instagram > Follow-along Campaigns

---

Unique Niche Promotion on Facebook (35)

---

Hey guys and gals,

It's time for us to start this follow along on Facebook!
My niche is quite 'unusual', it's: beat instrumentals that are leased for singers-songwriters/rappers to sing/rap for different licensing types i.e. pricing points starting from 30$ (depending on the artists' needs, see here for our pricing structure) and the reason I 'chose' this niche is because I work with this music producer' guy and I can do the testing for free So hopefully we can get somewhere before we blow all the testing budget

Note: Music has a diverse scope of targeted audiences/genres but the audience we're going for is specifically this style i.e. this combination of rap and country/pop music.
Also, the producer has good music (i.e. good product) and is collaborating with multi-platinum producers and stuff...

Current problems:
- Too low CTR (and hence also too high CPC)
- Might be too high CPM? I don't know the good numbers on Facebook yet. The CPM for our Remarketing adsets (website visitors) definitely seems high (see below).
- Low CR
- We might need a landing page to entice the users more, which we don't have yet


Things I did so far for purchasing intention campaigns:
Started a US conversions camp with 2 cold traffic adsets and 1 remarketing adset (30d website visitors) to compare, so each adset is with its own audience (2 saved audiences and 1 custom audience).

Audiences I used:
1) Remarketing audience, last 30d website visitors
x6 ads

2) 2nd adset/audience:

x6 ads

3) 3rd adset/audience:

x6 ads

As you can see I split tested 2 different cold audiences (assuming one might be more targeted or something stupid like that), I haven't found any other better keyword interests to target in our specific use case, I am also using InterestExplorer for my audience research btw.

Ad copies that I split tested were a square image and a carousel featuring some of his instrumentals, we didn't feel like making one promo video could be make a massive difference for the efforts it would take.
So I split tested 2 images as the main 'attention grabbers' while also testing 3 slightly different copywriting texts (mostly changing first sentence), and a carousel/image split-test. So 6 ad variations per adset x3 adsets so we had 18 ads in total (which all of them sucked btw).

So those were our 6 ad variations:








My impression thus far is that CTR was too low (hence CPC was too high). Some of the remarketing campaigns I did previously I managed on getting around break-even ROAS (though not ROI!) due to some crazy bulk deals we were doing during BF/CM like 'buy 1 get 9 free' -But so far on the recent test no conversions.
I made a 30$ daily budget on each adset/audience (hence 90$ daily budget in total) and ran it for a couple of days, IMO I overheld this test without making some changes as the CTR really seems too low and so is the CPC (interesting to me is also the low performance of the Remarketing audience...), those were the results I got after initial testing:

---

Cost are probably high because;

- your audience is quite small
- your CPA is too high for FB - try the "usual" facebook pixel warm up method; create engagement via a sick video, make lla's from that and create a low key conversion point (button click? lead?). Once you have enough data of the second step, you may now proceed to conversions. Cold audiences on "high" cpa offers only really work if your ads/angles are killer.
- your ad game isn't that good. how many images did you test? the images dont look appealling/interesing/relevant to the product I think. I am an amateur beatmaker myself but the christmas tree and singing girl are irrelevant imho.
- the copywriting I dont like because you list like 20 positive points. It looks messy, too much text/little whitespace) long sentences with heavy words. thereby you target quite a sophisiticated audience since musicmakers are basically artists (or they think).... don't use the spammy/clickbait-ish capital letter headings etc. try to act/present yourself more professional.

*additional, why not go broad and try to trigger people with simple headers like "Every musiclover can make a song or beats with.... because.... only for xx now! etc etc

** also target on genre's or even certain artist likers - and adjust your image/text etc towards that particular genre/artist.

---

So nice to see another follow-along from you @roiter123! I love how you always lay out all the details in an organized fashion. Easy to read and understand, and more importantly, easy to give suggestions.
@stickupkid's advice is on point as always! I only have a couple of comments.

1)Audience Targeting: I would definitely not test cold audiences at this point, but instead focus on testing ads for the warm audience. Focus on the website visitors, and create LLAs from that so you'll have a relevant and targeted audience that is of sufficient audience size.

Testing cold audiences using unproven creatives will be unnecessarily difficult due to there being 2 variables. Whereas testing a warm audience would eliminate one of the variables.

You can also try to target website visitors as a custom audience, but depending on how much traffic the site has been getting (I'm assuming from SEO efforts mostly?), the audience size may be too small - which would make the CPC high.

So again: I'd suggest to create LLAs from website visitors - maybe starting with 1% if that will give you a big enough audience size. If not, try 1-2% or 1-3%. Then focus on testing creatives until you see success.

Once you have some proven creatives, and want to scale, THAT would be a better time to test cold audiences.

2)Creatives: @stickupkid has given amazing suggestions. Focus on the image/video first and foremost, because that's the first thing visitors will notice as they scroll through their feed. If you can't stop the scrolling using the image/video then they won't get around to reading the text.

Your ad text is not bad! Great start with the bullet points and selling benefits. Just make changes as suggested by @stickupkid, and then focus on testing images/vids.

Videos are worth testing. I know it's extra effort, but it's necessary effort. It doesn't need to be very professionally done. You could even create a powerpoint presentation, screenshot each slide, download a royalty free audio clip (or even an audio of you describing the service and benefits), and combine them into a video in adobe premiere pro. You can find a step-by-step tutorial on how to use premiere pro in my FB newbie tutorial.


Hope that helps!


Amy






Sent from my iPhone using STM Forums

---

Thanks alot @vortex and @stickupkid for your replies and feedback!

What do you consider an audience size that is big enough? Because in my audiences I tested above if you take a look the first audience got 4.1m people, isn't that quite big?
And @vortex I am already targeting 30d website visitors, if you take a look above - it was one of my adsets

How do I know which % is the "right fit"? Take a look, it doesn't matter which percentage I pick or how broad the base audience, it still doesn't show me the volume estimation


Recently, only the ones showcased above. Before that, we even did some cool videos that we tested with artgrid footage on a fairly broad audience yet still CTR was terrible/less than 1% (which I consider bad for FB based on metrics I've been hearing from others)

And this was the copywriting that we used for leadgen and the 2nd one is sales campaign (and some version of this video):




What images would you use? I have no idea what could give me a better CTR. I might be doing a terrible job but I tried sneaking in skin for our male audience with the female singer for better CTR (might be a stupid move yes) yet still obviously the CTR wasn't near what we want.

We already tested videos on FB, some version of this, low CTR, even for leadgen

---

By audience being small, I was referring to your warm/hot audience, i.e. website visitors. The other audiences are not proven yet - so better to focus on a relevant audience first.

I may have made an assumption of your website audience size - how big is it?

And yup I saw that you were already targeting website visitors - I was mainly suggesting to focus on that and the corresponding LLA instead of testing cold audiences - until you have some proven ads anyway.

For LLA: Try 1%, 1-2% and 1-3% to see which will give you lowest CPC. Yes there will be significant overlap of audiences but you only need to run this test briefly until you can tell which one gives you the lowest CPC. Then you can pause the other two and just use one to keep testing creatives.

That video is too long for FB. Try to cut out the first 12-15 seconds of relatively uninteresting content, and shrink the rest into a 10-15 second clip. Long videos can definitely convert but the ones I've seen convert well, all tell captivating stories. For your type of promotion, I think really short videos will do a better job. (And a lot of people will suggest the same - short videos tend to work better more of the time because people have short attention spans. Longer videos - e.g. that run several minutes long - can presell better but you need to make them really good in order to capture attention for so long.)

And do come up with attention-grabbing thumbnails!

Testimonials work very well and I see that it was worked into your video. If you can somehow bribe customers into giving you video testimonials those can really help conversions. You can test each video testimonial separately and also splice them together into one video and test that too.

As for what thumbnails/images work well - I don't know this niche so have no idea. Will try to do some brainstorming - maybe you can do the same? Aim is to stop the scrolling.



Amy










Sent from my iPhone using STM Forums

---

I'm looking for examples of images that can work - please note that they're probably copyrighted - they're meant to serve as inspiration:

https://encrypted-tbn0.gstatic.com/i...HvNHQ&usqp=CAU

https://socialmediaintheworkplace.fi...hed-guitar.png

https://i.pinimg.com/originals/27/28...cf9466dc59.jpg

https://ontheaside.com/wp-content/up.../05/swing6.gif

https://previews.123rf.com/images/az...othes-on-a.jpg

More ideas:

-Take an image of people playing instruments, erase the instruments in photoshop and replace them with question marks.

-Find or create a video of a bad attempt at human beatboxing - and pitch your service as the better alternative.

You can get really creative with this! Just please be careful to stay compliant.



Amy


Sent from my iPhone using STM Forums

---

Only thing that's above 1000 is Website visitors last 180d = 1,200 people
But I was also thinking about taking our email list of past customers which has like 300-400 emails and creating a LLA off of it, what do you think?

By saying 1-2% I'm assuming you just mean 2% LLA for one adset and by 1-3% 3% for another adset, right?

And thank you so much for the creative ideas! I'm going to do some brain storming myself, thinking that the copywriting should also be relevant and attention grabbing...

---

1)

Tired of your producer looking like this?

It's time for all serious musicians to get on board with BUDI and find the best beats for them.

✅ High quality instrumentals & beats
✅ A lot of them include hooks - your hit is half-way there!
✅ Take your music to billboard levels - inspire people around the globe!
✅ Great value for money - get 1 beat free when taking 2
✅ Get winning collabs, get advice, be discovered by major record labels
✅ Have the opportunity to work with a top producer exclusively and make custom projects together
✅ Have your vocals/ Songs mixed & Mastered when ready (achieve chart standards)!

2)

Tired of your music production sounding like this?

It's time for all serious musicians to get BUDI to produce their songs.

✅ High quality instrumentals & beats
✅ A lot of them include hooks - your hit is half-way there!
✅ Take your music to billboard levels - inspire people around the globe!
✅ Great value for money - get 1 beat free when taking 2
✅ Get winning collabs, get advice, be discovered by major record labels
✅ Have the opportunity to work with a top producer exclusively and make custom projects together
✅ Have your vocals/ Songs mixed & Mastered when ready (achieve chart standards)!

3) https://media.giphy.com/media/fuJmaW...OlgD/giphy.gif

Tired of your music production sucking?

Get your music instrumentals from BUDI and win your Grammy!

✅ High quality instrumentals & beats
✅ A lot of them include hooks - your hit is half-way there!
✅ Take your music to billboard levels - inspire people around the globe!
✅ Great value for money - get 1 beat free when taking 2
✅ Get winning collabs, get advice, be discovered by major record labels
✅ Have the opportunity to work with a top producer exclusively and make custom projects together
✅ Have your vocals/ Songs mixed & Mastered when ready (achieve chart standards)!


---

Some Ad ideas lol! What's the worse that can happen if it's copyrighted and how can I check? I guess the most appropriate question should be... What's more likely to happen?

---

How are your new years holidays going guys and gals?

---

Happy New Year @roiter123! Sorry I forgot to reply to your posts!

Creating an LLA from the mailing list is definitely worth a test! Have you tried to upload that list to create a custom audience? Not every email can be matched.

Oh regarding LLA percentages - in my haste I made a typo - it should be 0-1%, 0-2% and 0-3%. But it's not the only way to test LLAs. You can just start by testing 0-1% and focus on testing ads for example, and then scale to 1-2%, 2-3% etc. if all goes well. Some advertisers like to test on a broader scale by testing 0-3%, 3-6% and 7-10%. Everyone does it differently.

The most budget-friendly way based on my experience is to test 0-1% first, because in theory it's the most relevant audience and should have the best chances of converting well. If I get that to be profitable I would scale to higher percentages, which in theory would be less relevant and tend to convert less well.

And yes these would all be in separate adsets.

I love your ad text and that last image you found! You completely took the ball and ran with it. Please take copyright infringement VERY seriously, because FB surely does. You don't want to get your account banned over something so easily avoidable. Try to find royalty free images that look similar (you may need to pay for those) or hire someone to create them. Fiverr for example. You can try animated/cartoon gifs too - lots of gigs on fiverr that can do that for you. Or take out a local classified ad to hire someone to go to the physical store and wave the instruments around etc.

Yes it can be a LOT of work to create good creatives! But that's where the real work is. Skimp on this part and your entire FB campaign will not stand a chance. This is what will make/break your campaign, so don't just "let's throw something together for a quick test to see what happens", and only invest more time if it pans out. You may be able to do that if you were testing tens or hundreds of ecom products to identify low-hanging fruit, but not when you're trying to promote the ONE business (yours or a client's). Make the best creatives you can - get creative with your creatives! And give it your best shot.

Eager to see some new stats when you have them!



Amy


Sent from my iPhone using STM Forums

---

This feels like first time realizing I actually need to upload landing pages in order to succeed in Push traffic a year ago. But Amy!

Okay I got it, I need to invest in creatives. Where do I find a picture of a guitar that looks like poop but isn't copyrighted? What do I tell fiverr people?
I feel like animating something could be worse for CTR for some reason? Feels like ad-corporate-spammy already when I imagine it. Unless it's something funny.
But getting some images that are similar to what I found above and knowing they are not copyrighted might work. But how do I know Facebook doesn't consider them as copyrighted? (some database they have?)
Or what sites/resources I could go to to get copyright-free gifs similar to the ones we got above? (I don't mind paying duh) And... Copyright free images?

---

This is an extremely specific product that only a limited audience will respond to and actually buy. Targeting the right people will be the breaking point.

The best would be to target members of FB groups that are related to music production. As far as I understand, it's not possible to do this directly, so I was looking for a workaround and found this: https://www.filed.com/resources/blog...-simple-steps/

And one more method here: https://rocketlink.io/blog/rocketlin...oups-with-ads/

And some lists of FB groups for musicians to take a look at, if you decide to give it a go :

https://flypaper.soundfly.com/hustle...oin-right-now/
https://blog.sonicbids.com/7-faceboo...musicians-2019
https://blog.symphonicdistribution.c...u-should-join/
https://afterdark.co/blog/latest-new...ps-on-the-web-

Most of the groups are quite "small" but that's probably normal due to the size of your audience.

---

Nice follow along @roiter123 !

The vertical looks pretty interesting and should definitely have some potential.

A few years ago I used to play a lot with audio tools, and maybe the following might be helpful.

- Your audience looks too broad for your product. Instead try to target them by interests related to the music production industry. Under additional interests you should be able to target people interested in:
FL Studio, Audio Mixing, IZotope, Sound Recording and Reproduction, Steinberg Cubase, Ableton, Adobe Audition, etc.

- Try to use video ads with a sound on play suggestion. This way potential users will be able to hear some small samples where you can randomly add audio marks to avoid abuses. AudioJungle does something similar, so might be worth playing around with such techniques.

- Maybe it's not in your current plans, but could be worth considering to reach out to cover song makers. They tend to either use acoustic versions of popular songs or just rearrange the base for themselves. This might cut down the cover production time by a lot for new songs, and on top of that potential prospects looking to get into cover songs or theirs might be interested in your product (they usually follow known cover singers).

---

Excellent advice right there! The closer you can get to people who actually make/produce music, the better. Many people are "interrested" in music, but just a fraction of them actually "makes" it and those are your customers.

---

@matuloo @platinum

That gives abit of hope!

Couple points about audiences:

I definitely considered such targeting as well but we are actually targeting the ARTISTS that will get on the producer's tracks, not PRODUCERS. Therefore independent artists are more likely to like pages like "cd baby" or "tunecore" (music distribution services), or some pages like "underground hip hop" but that's abit more broad, than iZotope which is a producer's tool. And keep in mind I'm also narrowing down then to match this country pop vibe we got with guitars.

Right now I already got some audiences which I THINK, might be enough for testing creatives (as I understood creatives would be the main goal to improve CTR atm), the audiences are:
- Website visitors pixel, about 1k last 180d (and the consequent LLAs)
- Emails of past buyers, about 400 (and the consequent LLAs)
- Email list of about 3k visitors for our relevant GEOs (LLAs...)

This is for sure. But I'll probably get back to audiences very soon. Meanwhile: Creatives!

---

You're looking at this in a wrong way, I believe Let me explain.

Even the artists have to record the music somehow and they need to use some software to edit the recordings. It's not like they head straight to the studio and pay for pro recordings. On top of that, the more serious players need a range of tools to play on a solid level.

My son plays a guitar and so far I bough him several guitars, 2 combos, a sound card called "focusrite", some pedal effects... he uses some software from focusrite too plus something else to record what he plays, he is subscribed to several websites that provide guitar tabs etc, plus some online guitar tutorials... And every single person who takes music/song writing seriously will have to use some tools.

This is what you want to target! Do not target people who like music, target people who like tools that are made to compose/record/produce music and related courses/tutorials. Thats where your audience is

---

THIS is where the work and creativity comes in my friend! Nobody said it'd be easy!

But it's not rocket science either!

This post suggests places where you can get stock images and videos:

https://stmforum.com/forum/showthrea...-Videos-Part-2

If you're worried about FB flagging your image, make edits even to images you have rights to - change the size a bit, and/or the colors or exposure/vibrance levels or similar. And remove the file meta data.

Animations being worse for CTR? I don't know about that. I DO know that there are good animations and bad animations, just like there are good and bad static images and good and bad videos. Only testing will tell.


Your wish is my command!



Don't judge - I threw this together just to illustrate that it's possible. I got the guitar from envato elements (let me know if you want more guitar graphics - they got a ton on there!) and the cute turd image from here:

https://pngimg.com/download/57949

And here's just the guitar image by itself if you need it:

https://imgur.com/BePD3Mc

Wanted to slap a "WTF??" beside the turd but obviously FB won't like it, so reduced it to just a question mark.

You can test variations of this - how about instead of one giant turd, have multiple smaller turds?

Or how about: A music score but using turds instead of notes!

So yeah...get creative!



Amy

---

Couldn't help myself...



Drew the staff in MS-word using lines. Same turd image as above. Treble clef from here: https://pixabay.com/illustrations/tr...-clef-1279909/

Again - I just slapped this together for illustration purposes. You can probably do better.



Amy

---

@vortex Thanks ALOT Amy. I'm literally going to test those in my Ads.

How does Facebook finds the copyrighted stuff? Can't I just claim that such guitar gif is mine (and I just uploaded it to imgur or whatever giphy)

Also, let's say I'm going to use the "pain solving" headlines (in Primary Text) for the creatives, for example:

Should I just continue with the CTA with the same vibe? Like "Get rid of sit music" in the Headline, and then something like "Produced by..." "Best beats" in the Description of the ad.


...

What do you think about my audiences?
I got you @matuloo, but don't you think that such interests (like FL Studio, Audio Mixing, IZotope, Sound Recording and Reproduction, etc...) are much more likely to include producers than singers/rappers? Our target audience is really less guitar players but more singers/rappers. Most indie singers/rappers go to studios from time to time and they might have a mic and a DAW on their laptop and some stock plugins.. but a music distribution service likers audience can have more of those singers/rappers because almost every singer/rapper needs those And distribution services are less (although still, but less) used by producers

---

Found a couple of more creatives from WikiHow, but I don't know how copyrighted are those.

---

Your audience targeting is a great start!

The ad text looks good, but I'd definitely split-test with shorter versions. Trying picking 3 TOP benefits from the list and massively shortening each. Pretend you had to pay a thousand bucks per word - and try to be as concise as you possibly can but still be crystal clear. Can test using 5 top benefits as well.

I find that sometimes (not always), people respond better to shorter text. If you have a strong landing page, there's no need to do ALL the selling in the ad. Pre-sell a bit and get that click and then let the landing page close the deal. (By the way DO you have a good landing page? That would be another crucial funnel element to test!)

As for image/video copyright: I'm not sure if Facebook actually has automatic detection or filtering in place to catch these, but they do have a form people can fill out to report if someone's using their creative work without rights or permission. For more info see:

https://m.facebook.com/help/1020633957973118

SOME Wikihow images may be shared but each case is different - here's more info:

https://www.wikihow.com/wikiHow:Creative-Commons

Does it mean you'll get caught by FB if you use images/videos you don't own rights to? Maybe, maybe not.

Years ago, I briefly ran some campaigns with a group of blackhat affiliates. They were just grabbing copyrighted images and pulling videos from youtube and slicing and dicing them to make ads. Accounts were being banned but probably not due to the use of copyrighted materials. And with more ad accounts lined up, nobody cared anyway.

So I don't have any conclusive evidence that FB was banning accounts due to the use of copyrighted materials in ads.

You MAY get away with using copyrighted stuff. I wouldn't know though because since I left the dark side I've only used images and videos I had rights or permission to use. I have no idea how good FB is these days at identifying infringement or whether they even have any automatic checks in place.

And then there's still the risk that FB may receive a complaint from the image owner. Yes that risk may be small but it's there.

It really boils down to how much risk you're willing to accept.

I'm of the "better safe than sorry" camp when it comes to running whitehat on Facebook. You are of course free to decide for yourself. But if I was running camps for a legit business that has spent time nurturing their Facebook page for months/years, I wouldn't risk getting an account ban or a page ban just because I can't be bothered to make creatives I have rights to.

The choice is yours to make.



Amy




Sent from my iPhone using STM Forums

---

I also checked Canva, it also has poop emojis and guitar elements you can use. You would need the Pro version, but it is still quite affordable.

As to copyright, everything I have seen suggests Facebook is extremely passive about it and only cares when it is reported. So while I wouldn't advocate being carefree, you are likely to only run into issues if the image belongs to a major brand that is proactive in protecting their rights.

If you do get a copyright violation on an image you own, you can dispute it and presumably get the strike removed. Although, it is Facebook, so no guarantee they'll believe you.

---

I'm pretty sure platinum mentioned this particular ones just as examples. You can definitely choose different products/tools that you believe are more popular with the audience you are trying to target.

But anyways, there can be quite a bit of an overlap... I have a friend who's creating beats for rappers, but at the same time, he's rapping/singing and performing on stage himself. And on top of that, as a producer, he knows plenty of other rappers who might be in need of some beats other than those that he produces. This is a closed circle... you might happen to target some producers with your ads and those might recommend your product to performers that they know. One way or another, it's definitely better to target people who are involved in music procuction, than it would be to target people who just like music in general.

---

Another potential audience for your product could be video ad producers. Some of the tracks I listened on your website look suitable for video ads, and your pricing is pretty much in the same range of AudioJungle.

In regards to the above suggested targeting categories, I'd say that almost every singer that wants to create its own beat at some point in time will definitely look for options on doing that. And torrent sites tend to be quite packed with cracked versions of some of the above-mentioned producer tools.

I think music distribution services, producers and professional tools might have a considerable overlap amongst each other. Try to give all options a proper test, then make decisions based of the results you will get - making assumptions won't help much.

---

This is a great tip too!

I'm currently involved in a project that's about shooting video content and indeed, we need background music for all of them.

For now, we're using music tracks from envato elements and I'm sure there are more platforms like that out there. So envato and other "music" content sources/services customers might be another great audience to target.

---

So,

After going back and forth with the producer/offer owner about the ads design ("I don't like this poop guitar!") we finally found ads that we could test.

Reminder our website: https://budi.beatstars.com/

I've setup the campaign's conversion event as Add to carts on website to potentially help FB for optimizing better with my 30$ pricing tag.

I chose 2 illustration images from WikiHow and 1 funny dog video playing a piano (in total 3 ads for testing).

those 3 ads over 3 adsets (each adset with a different test audience):

1) LLA (2%) of All email leads to date
Country: US
Age: 18-30
Genders: Both


2) LLA (4%) of All sales leads to date
Country: US
Age: 18-30
Genders: Both


3) LLA (1%) of All Beatstars Visitors last 180 days
Country: US
Age: 18-30
Genders: Both

Ads:
1) Illustrative 1;


2) Illustrative 2;


3) Dog video/"gif";


So this dog gif had quite a big positive difference in CTR and was much better than the Illustrative ads by CPC/CTR across all adsets and on all adsets the CTR of the doggo was around 4%.
After quite a short while of noticing the quite huge difference in CTR I left out only the dog gif running to gather more data on (hopefully) conversions.

The results were quite horrendous, see below:




Yes I double checked that there was actually no conversions after this much of clicks, according to FB - not even an ATC!
Yes I was running DL (direct linked) but no product should perform this bad IMO!

Google analytics (according to them - avg. session duration is 6 seconds ??!!):


Well I think that's all the update for now.

---

I don't think the results are horrendous at all! The way I see it: For less than $300, you've learned how to make WAY better ads than before.

From $3-$16 CPC to $0.20-$0.30 CPC, and from 0.4% to 4% CTR - that's a MASSIVE improvement!

Next - let's figure out why you're not getting conversions.


Lots of people are clicking through but not buying - can mean one of two things. Either:

1)Your ads are attracting the wrong people.

OR

2)Your lander is dropping the ball on converting the visitors.

Or both!


A few suggestions:


1)Tweak your ads to better attract your ideal audience.

Make it crystal clear WHAT the service is and WHO it's for.

I would also test more images, and seriously consider testing videos. If you could reach out to some of the long-time customers and offer them an incentive in exchange for a 10-15 second video testimonial on who they are and how your service has helped them, that would be some of the best ad content you could possibly get.

Choosing the puppy gif - nice work btw - based on CTR alone may not be the best idea. I know it looks like you're getting the best value by running only this ad because the CPC is the lowest, but what if people that are clicking on this ad, are not a good match to the service?

I've lost some hundreds of dollars to learn this one lesson. Used a gif of a monkey destroying a computer in an ad when promoting a make money online opportunity. Got sky-rocketing CTR and ultra-low CPC. Unfortunately, the ad didn't convert. Worse, I set a high budget because I was trying to test a bunch of other ads at the same time, but because this particular ad had the highest CTR, FB favored it by sending all the traffic to it. My whole budget was zapped with zero conversions.

My guess is that people just found the gif cool and may have even expected to find a full length video of the monkey trashing the computer further. The same may be what's happening to your puppy gif - that WOULD explain why they didn't stick around when they got to your site.

I would resurrect at least one of the two ads you paused out - the CPC for those isn't bad! And it would probably be good to test a couple more.


2)Test more audiences.
@platinum's suggestion of targeting by interest is worth testing. It's true that FB has gotten pretty good at identifying the ideal audience, so that in many cases targeting broad can work better. But when you have a niche product, interest targeting can still help a lot.

BTW - why the different percentages when creating LLAs? The higher the percentage, the "more different" the audience becomes (compared to the original seed audience). Why not just test 0-1% for all 3 audiences to start?


3)Create a better landing page.

Right now you're relying on the ads to do most of the selling, which really isn't ideal.

You're basically counting on the visitor to read through the entire ad to find out what the service is, so that when they click through to the website, they'd be ready to pick out a tune.

Whereas in reality, many visitors may not finish reading the ad. And/or, the ad text isn't enough to pre-sell the visitors adequately.

The landing page is such an important part of the funnel. I wouldn't hold my breath counting on making the whole funnel work by tweaking ads and audience alone.


Lastly - I have some questions for you if you wouldn't mind sharing:

-How was the service being promoted before?

-If it was getting SEO traffic, what sorts of keywords were bringing people to the site?

-What was the average conversion rate? (If you have the breakdown of CR by keyword that would be even better!)

-Which page(s) are visitors spending the most time on? Is there a typical "path" (from page to page to page) that visitors would follow before making a purchase?

I have no idea what tracking, if any, was in place before you came along, so I know you may not have some of this data, but gotta ask anyway. Insights like these can really help us improve the campaign.



Amy

---

Hey Amy, great suggestions I really appreciate you being my caring forum friend.

I agree that the CTR and CPC are completely better now than they used to be, I was rather talking about the conversions.

The service was/is promoted mostly through YouTube or the Beatstars Marketplace (https://beatstars.com/budi)
Based on previous organic data the CR on the site was 1%, usually such CR would be when people listened to the beat on YouTube and then advanced to the website (already quite warm).
I then also did Google keywords CPC tests on the site (also direct linking - pretty well targeted audience), ROI was also bad, around -50% to even -70% on average I'd say.
The usual path of a user is simply clicking on the ATC of the instrumentals they want and then buying (sometimes entering the instrumental's page itself first) often selecting multiple instrumentals.

The reason for the different % in the LLA is an assumption of mine consisting the factors of the size of the base audience and how well targeted it is (for example the "sales leads" email import only had like 300 emails imported therefore I assumed 1% would be quite small and I gave it 4% LLA; all email leads around 2k less targeted but bigger so I gave it 2%...)

But, ACTION PLAN:

- I really thought the way I could make this work would only be with such high CTRs, I may be mistaken, but I would love to if I could find such high CTRs that would still convert (given that most of the problem is in the creative) although I'm not sure how influential the creative really is in our case. It might be. Hard to tell what other ad could have such good CTR (or atleast in the area) but with higher CR.

- The landing page might play a more important role, but I am abit low on ideas for a landing page, or maybe I am just lazy to think of ones (?) The guy is camera shy. Are you going to tell me it's time to read more copywriting books?

- I really think the audiences should be a decent starting point to atleast get some conversions, I feel like I should work on the other parts of the funnel (mostly landing page)

---

Current status: Ads are on pause. Been procrastinating on Facebook.

Meanwhile running Push and Native. But I need a third baby!

---

Eek sorry I've been meaning to reply - had the tab opened but lost in a sea of 100+ tabs lol!

I hate to need to point out something that may be discouraging, but not all websites that are successfully making profits from SEO traffic, are good candidates for paid traffic. There is a possibility that no matter how good you are at media buying, that you just won't make it back out.

But let's give this our absolute best shot! I have several suggestions:


1)Collect More Data First

You said:

How many views on the videos per link click (i.e. website visit)? And what's the average revenue per click? Any idea?

When we have these metrics we can work backwards to figure out how much we can afford to acquire each click. Of course the nature of clicks from youtube vids vs. clicks on FB ads would be completely different, but at least we'd get a very rough idea.

It would REALLY help if the site has google analytics or similar installed, so we can look at what types of keywords the audience are using to find the site. I'm really not familiar with the music industry so have absolutely no idea. When people need beats, do they usually google something like "hip hop beats for download", or "Heaven Mac Miller x Logic Type download"? Knowing the answer will help to decide what the landing page should look like, what keywords to bid on adwords and other stuff.

So conversion metrics is one unknown. And another unknown is HOW customers have been finding your service - e.g. what keywords they've been using to find those youtube vids or the Beatstars Marketplace.

So I'm going to ask again: Would it be possible to get google analytics data, or install google analytics now so we can collect some data going forward?

Also: Have you checked out video analytics data for the youtube videos?



This data can be really valuable. Specifically, on the "Reach" tab, look at "Traffic source types", "Impressions and how they led to watch time", "Traffic source: External" and "Traffic source: Youtube search" to see WHERE your traffic is coming from and the percentage from each source, video CTR, and the keywords people are using on youtube to find your videos. On the "Audience" tab you can find audience demo data on age and gender.

Those FB ads - depending on how many audiences and how many ads and angles you've tested, and how much you've spent on each angle+audience, we may or may not have enough data to conclude whether we'll be able to make FB traffic back out or not. It may be good to take a step back to look into analytics and data as I've mentioned above, before taking another look at the FB campaigns - to assess whether we're using the right ad angles.

Same thing applies to landers - I feel that we're kind of flying blind here, not knowing what kinds of keywords customers are using to get to the site. Are they e.g. looking for hip hop beats for download, or they already know which beat they want? Are they music producers or video creators looking for background beats or whatever else? Knowing this information will directly determine what your lander needs to look like, or even whether or not you NEED a different lander. For example if most visitors already know what beat they want when they arrive at the site, then your current arrangement should work just fine.


2)Do More of What Works

Once you find out what search terms customers are using to reach your site from google / your videos on youtube, you can run ad campaigns on adwords and youtube to target those keywords.

I know you've been trying adwords, but which keywords have you been targeting, and are they the ones your customers have been using to find the site?

And it won't hurt to test youtube ads either.

However, before you run either type of ads, knowing the metrics I mentioned at the start of this post - i.e. how many views on the videos per link click (i.e. website visit), and the average revenue generated per click - will provide a helpful benchmark.


3)Try Cheap/Free Methods

Those beats have a free option in addition to the licensing option, which should make it relatively easy to promote via free/cheap methods!

Think online forums, FB groups and chat groups.

Again, this is where knowing your audience would REALLY help - music producers hang out at different places than video creators that produce videos to earn ad revenue and/or make commissions by promoting products.

Moreover, once you know exactly who your audience is, you can figure out other ways to reach them. For example advertise on specific websites, or pay for email blasts to relevant audiences.

Also: Once you know which keywords customers are using to reach the site, you can rank more properties on google to increase the amount of traffic you get - for example FB pages, google sites, articles on medium - these all have ranking power just from their root domain.


Again, hope that helps! Stuff like this is a whole different beast from running CPA offers. Let's keep the discussion going and try different things and see what happens.



Amy

---

@roiter123 Something else just occurred to me: One possibility would be to send a customer survey to existing customers to find out how they found the site and what they're using the beats for. You could even turn it into a contest where the winner gets to pick 10 beats for free or something.



Amy

---

Are you seeing any results with these traffic types?

What was the CR you got from google traffic? And was that just search or also the display network?

I'm asking, because google search can convert ANYTHING with the right KWs chosen. But, the clicks can be very expensive so the ROI might not be the best. So the better metric to use here, to judge the potential of the product, would be to look at the CR.

Your product is quite specific and only suitable for a limited audience, so there is still chance you won't be able to promote it in a profitable way, at least not on large scale. The google ads data could tell us more.

---

Re: Views on video per link click – a bit hard to tell since we don’t have such kind of tracking on YouTube - but simply calculating the ratio between the number of views and amount of clicks directly on the beat in the website it’s around 3% CTR (we also run pure video ads for more views exposure – non conversion objectives).
Average revenue per click is also hard to tell since not all beats have links pointing to this store we have tracking on but rather to another 3rd party hosted store (The 3rd party hosted store has its benefits because it’s like a SoundCloud for beats and can build engagement there and fans – this page), but for the beats that we do have tracking for RPC, it can range from 1$ to 5$, more inclined towards the 5$ mark.

But keep in mind that this is not an accurate comparison – Those who are proceeding to this from the video already go with intent of purchasing after they already “fell in love” with the beat. But with Facebook it would be different since they are just seeing a “service offering” or an invitation to visit the catalog with some potential benefits that they will checkout and then browse to find their favorite beat etc.






Re keywords: I’d say both, but on YouTube the trend is a bit for more specific keywords like you mentioned “Mac miller x Logic type beat”, or some kind of a vibe KW (like melodic chill guitar type beat). For search paid campaigns I was doing the "hip hop beats for download" (obviously more volume).


On analytics, we don’t have any organic KW data for some reason.




Paid KW data reveals the Google Ads campaigns I've been testing. Tracking is sloppy so not all conversions were registered but let's say ROI was at the -50% mark:



But on YouTube we do have some data. This data originally made me decide on the vibe I was going for with my ads and my target audience (in example, realizing that the audience is more melodic – including acoustic guitar in the interests):



Traffic source external - has a comparatively smaller amount of views, we have been pushing new uploaded videos via our newsletter by just sending people the links and writing a couple of words:



So it’s quite simple, people who already know what beat they want usually came "hot" from YouTube or they are some sort of fans already.
Otherwise - what we’re doing here, is trying to advertise the catalog as a whole for users who are interested in general and then having them browse through the catalog and pick their favorite beat/s and make us moolah.
The ideal customer is definitely an artist looking for beats to record his sexy vocals/raps on. Not producers. Not video editors.

Tested that! It didn’t work. Tested retargeting to the store with this fancy video I’ve shared couple of posts above and also cold traffic to an overall good performing popular beat but without luck.

Seems like we don’t rank for anything useful at the moment huh?

Btw - here is an example of someone who is doing successful selling to Google Ads cold KW traffic and also made some blog posts to try and rank: https://www.luke4presbeats.com/
I was literally targeting the same keywords as he btw (see paid KW screenshot above) and using negative KW for optimization and optimizing device types etc.
He even made all kinds of tutorials sharing the exact ways he was doing that and I literally copied what he was doing, lol: https://www.youtube.com/watch?v=e-VpxVixUbA

And...
@matuloo Sorry for not mentioning that! On Push and Native I meant I'm only running my own affiliate marketing stuff, see my follow along's

And what I did with Google Ads with this Biz was search traffic (without Display) but I also tested YouTube ads (like instream ads). See CR ballpark above, quite bad despite targeting relatively close KW and working with the negative keywords to filter out.

---

Ps. Desktop is obviously converting better because the ones that purchase expect to download the files to their computers. And of course I tried the Google Ads game with mostly desktop only, CPC would increase quite a lot!

---

Sorry for replying late! Been busy with the "What's Working in Affiliate Marketing" report and didn't want to rush through your post.

Thanks for this info! I was thinking you can try running more youtube traffic to existing videos on there that are already bringing in good RPC. It sounds like you've already been doing that:


And what's the RPC on these? Comparable to the $5 from "organic" youtube traffic to the same vids? Breaking even or better? Any opportunities for scaling there?


Of course! They are different sources altogether, and the visitor's intent is different.

This is why I've been asking for data that can potentially provide more insight into visitor intent. E.g. Who are they? What are they looking to do with the beats? Are they looking for beats in general when they come to the site, or do most of them already know they want a certain beat (like what you said - they've fallen in love with the beat before they clicked to the site)?

This will all influence 1)which audiences to target and 2)what ads and messaging to show to each.


I've been wanting to ask: So beatstars is like soundcloud, right? What's your original website that your friend owns? I tried to search for it but nothing turned up.

Are most of the sales being made at beatstars?


Perhaps try some of these? https://neilpatel.com/blog/unlock-keyword-not-provided/


For adwords, can we perhaps try more specific keyword phrases? E.g. "sunflower beats for download" "avicii beats for download" or even include names of individual beats in the phrases?

Of course search volumes will be really small - but worth a shot! Maybe if you have a ton of these keyword phrases, volume would add up!

Of course the important thing would be ROI. I'd imagine a lot of people would be looking for free downloads, so if you haven't already, add "free" as a negative keyword.


Noted! But the BETTER question to ask may NOT be "who is our ideal customer?" But rather, "who are our EXISTING customers? What is their intent? Where in the buying cycle? And where do we find more of those?"

Which is what we're trying to figure out by looking at all this data.


I was thinking more along the lines of running paid traffic to existing videos. That's why I was asking for CTR and RPC on those vids. You can do a calculation to see whether the traffic has a chance of backing out. The only unknown here is whether you can target the right audience to get the same CTR and RPC as the "organic" traffic you've been getting to those vids.


Oh wow that's EXCELLENT!

How do your stats compare with his though? Product pricing? CPC? CTR? CR? RPC? This type of comparison can give us insight into where we can / need to improve.


By the way how much test budget do you have left?




Amy

---

Home > Facebook & Instagram > Follow-along Campaigns