Dubmill - TikTok (7)
First of all, I'd like to express my gratitude to forum's owner(s), moderators. You made best forum in AM. It's a big deal as it makes many people's lifes better.
Also, I want to thank @cassshy, @ScottyG, @jason a - I found a lot of value in your posts.
I will only post about crucial things understanding how boring can be reading huge posts about everything.
Okay, here's the start.
I've done some groundwork and 30 min ago started a campaign. Focus for the near future - US LeadGen. Start with autoinsurance. Analyzed over 100 ads on this vertical. For first test chose 6 angles and adapt those ads for me - add new text / voice / use same angle but with different videos I found.
Account: from agency.
Campaign structure: 1 offer - 1 campaign - 1 ad group - 1 creative.
Optimization Event: Submit Form. Comments & Videos download off.
Target: US, 25-55+ M/F.
Interest / behaviors / device target - No.
Daily budget on each ad group: $20 to start and get first experience.
Bid strategy: First start with Cost Cap. Now it recommends nearly $10, made $14. Reason: I heard from 3 practitioner usually cost cap brings more cheaper leads. Good for start. Also as I know on Cost Cap it needs more time / changes with to start getting traffic. Will see. If after 3 hours I won't recieve traffic - I will up cap to $30 gradually. If it won't work - I'll start Lowest Cost.
Plans: I want each creative spend at least $25 (approximetely x3 from payout) to make conclusion. Also, after $10 each spend I will analyze CPM, CTR, CPC on each. I suppose on my vertical and target CPM shouldn't be more than $15, CPC more than $2. If I'm wrong please let me know.
Global plans: now whole focus on making Autoinsurance - US profitable, then Solar, Refinance, Auto loan, Education (grants, language courses), Info-subscriptions, eComm (Health, Fuel Savings, Energy Savings)...
I am happy with any format of communication - I will help where I can. Also fully ready for any kinds of cooperation. Will post my thoughts and results.
See you 
Love the organization and methodical testing setup man.
Only thing I'd recommend is to up your budget per creative to $50 and only use cost cap after you've cracked an offer.
In my opinion, it's another mystery you need to dial in and waste time on. I'd rather figure out if the offer converts and then optimize shit like cost cap later to improve ROI.
I also would only run auto to learn how to dial in creatives and to learn how to innovate. It is very overcrowded and there are some other amazing low payout leadgen offers you can run instead, message me on Skype and I'll connect you.
Add me on Skype if you have any questions.
First test on Auto Insurance US done on the day of first post. Results:
3 hours after launch I haven't received traffic on Cost Cap. Decided to create clones of ad groups with Lowest Cost how Scotty recommended. Traffic started in 1 hour. After 3 hours of running campaign I got this result and turned off.
Key points:
CPM. As I am able to judge, $22 is not good CPM. I suppose reason is I took existing ads as it was mentioned a few times on the forum, TT makes higher CPM for not unique ads. Although I've done few changes to renovate them, seems it's not enough. Conclusion: Needs to create own videos. Currently I see 4 ways:
1. Find long time working approach (for example, girl is showing options with her finger with popular music). Then find on TT search another girl doing almost the same, and update those "options" in video editor. I guess it can be considered from TT as a unique ads.
2. Be creative and generate ads as a set from different videos. For instance taking offer personal loan, it could be first 2-3 seconds person who opens his empty wallet and is upset with a text "Couldn't get money cause bad credit" and then he started dancing with cash and a text like "When got money from the only this site".
3. As it already was mentioned on STM many times - find performer who creates video by your script. Some of sites were adviced: Genero, Backstage, Fiverr.
4. Create video by yourself. It's not even necessary to show your face or voice. I will let you know when I create some ideas and videos.
CTR. As I analyzed from many posts of TT professionals, CTR on leadgen should be 1%+, better 1.5 - 2%. So mine is not enough. Once again, I think reason is not unique creatives, although I haven't use very popular ones - just relatively new that I found interesting.
CPC is too high because of factors above. I think I should strive for CPC < $1 on leadgen US.
After getting this initial result and Scott's advice not to use auto for start found other interesting vertical - finances.
As I work with an agency account, I sent them 5 offers from different sub-verticals for approve and they approved 3 of them and rejected 2.
The reason is debt consolidation. My agency's TT rep didn't approve this. I know many people here have their agency accounts and are providing ads account for a percent. Maybe someone have other agreements with TT rep and through your accounts I could test those kind of offers. Please PM me. Big thanks.
So learning is moving at a good speed. See you
Originally Posted by dubmill
You don't have to publish videos of yourself but being able to quickly create UGC content and send it to an actor or actress to clone is an amazing skill that will make you thousands of dollars.
Once you can create your own creatives quickly, you can start to test offers at a lightning pace.
Offer: After last post I changed focus from Auto Insurance on Auto Loan because it may be not such overcrowded as auto insurance.
Creatives: Create 3 unique. First - me in the car telling how cheap / month and with bad credit I got the car. Second - around overview of another car I found on the internet with my voice over and second part of the video is me filming phone with instruction of filling out the form. Third has an angle of 2th with other car and approach.
First results:
CTR is much better this time (1.6% against 0.6% earlier. Not enough clicks of course but anyway. I think unique creatives and less popular vertical made their role.)
CPM is abnormally high. ($30 against $20 previously).
Asked my affiliate rep to provide info about CR / EPC on this offer from social traffic to understand roughly how far I am from profit with current CPC
Answer - CR 6.19%, EPC $0.74. So $2 CPC is obviously very far.
Okay, started to find ways to lower CPM from $30 to $5-10 (this should be normal I suppose). My hypotheses were:
1. Start campaign in early time by US while auction is not so busy.
2. Try Cost Cap. -20%, +20%, +40% from suggested bid (around $12 that time on M/F 25+ USA). As I saw on the forum Scotty said while working with Cost Cap it's important to test different group budgets also. As I understood with Lowest Cost budget is not so important. So to all of these bids (-20%, +20%, +40%) created also copies with different budgets ($20, $50, $100, $200). In total 12 ad groups.
3. Try different from Submit Form optimisation event. Chose View Content.
4. Try different from Conversion campaign objectives. Chose Traffic, Video Views, Lead Generation.
5. Try Automatic Placements instead of TT only.
6. Try CBO campaign.
7. Upload video to TT app from mobile, make some changes there like music, voice over, some text before create campaign with this video. I thought TT may prioritize this kind of video as UGC and make lower CPM for them.
8. Try different optimisation goal instead of conversions - clicks.
9. Try to enable comments. Because it's written: "We recommend keeping user comments on to help your ads achieve more impressions and conversions.".
Yes, I understood maybe all of these ideas don't really affect CPM but I had to test my own
Second creative gave me 1 conversion so I decided to test all of the above on only second creative (it's ok because I need to get CPM results and test different settings so all these tests should be only on one creative variable).
All of these hypotheses went testing and here's results by ad group (consider there are few campaigns because some hypotheses needed changes in campaign settings):
To sum up:
Low CPM got on 2 campaigns:
1. Traffic Campaign Objective.
2. Optimisation Goal - Click.
Other ideas didn't bring low CPM or CPC. By the way, on those 2 campaigns CTR was event higher (2%).
What's also interesting - I got 0 conversions from 174 clicks on those 2 traffic/clicks-oriented ad groups. With an average CR of 6% on this offer it's quite strange. Obviously for me, when you say TT you want traffic, it will give you traffic which clicks but it doesn't perform well.
Cost Cap campaigns didn't give traffic at all. So later I decided to make much higher bid to test them as well:
As we can see, it also didn't give desirable CPM & CPC.
Okay, move further.
Next idea I want to test:
Product category. I suppose TT may pessimize loan category at all. So I've registered at 6 aff networks to analyse best performing offers for TT and compare them. Speaking about networks I do want to highlight Perform[cb] as the best network we've worked with from our other traffic source. It's very good network basically but rep Margo makes it best of the best for me.
Finally I chose 8 offers. Got approved on almost all of them from my TT agency. Want to test them all this week. Then I will be able to judge if offer category affects CPM.
These categories are:
LeadGen: Personal Loans, Mortgage, Health Insurance, Mortgage Refinance.
eCommerce: Smart Watch, Massage Gun, Fuel Saver, Headphones (Trial). All USA.
Two main indicators I chose offers with:
1. Trending. Hot performing on TT.
2. Payout < $25 for faster and significant tests.
I'm sure these tests will be very interesting.
If someone also working with TT would like to chat - my skype is open for sure
- live:.cid.5edf57950e84d96b
Yesterday tested Personal Loan offer as I mentioned above. Results are much more interesting comparing with auto insurance and auto loans in my case.
Those are 3 unique creatives. Total revenue is $18 but one creative is almost break even. Conversions weren't passed from my affiliate network to TT correctly (only one from 6) so I check stats by SUBID.
- CVR is 20-25% depending on creative which is awesome for start. I have already created tens of ideas for creatives so I'm sure I can do better.
- Current overall EPC is $0.56. As I know it's a good EPC for this offer. So I don't have critical problems with creatives, CR and EPC - they are working nice.
- Main task to solve is CPM and CPC. They are quite high. Best CPC I managed to get on one creative is $0.91. I suppose I need to aim for $0.5 and less CPC to be stable profitable.
Next plans:
- Test those ad groups during 10 hours because this time they were tested for 2 hours only. It's not statistically significant.
- Test the same groups with different cost caps (now they are lowest cost) to see how this can perform.
- Test best creative on different interest / behaviour settings in different ad groups.
- Create and test new creatives from my ideas.
- Set up correct conversions tracking in TT for optimising. As I see right now the best way is to use tracker for this goal. Here's an amazing setup from @twinaxe. Currently I'm able to integrate my pixel code into affiliate networks directly (some with server-answer, some in hardcode) but it seems like quite unstable.
Advantages of tracker integration:
1| Should be 100% of conversions passed in TT.
2| Great possibility to save all information (costs / revenue / geo / device / ISP...) in one place.
The one thing to consider is my agency told me TikTok doesn't like redirects. Tracker clearly adds +1 redirect to a chain. But network's link also has one redirect and that's OK for TT. I think this rule is quite stretchy and unlikely they would ban an account for this. Also I see many people here use tracker for working with TT agencies so should be fine.
- Test new offers: next is Rent-To-Own. As I heard from several people this kind of offers doing well, going to check this out. BTW found small amount of creatives for this vertical in pipiads although tested 20 keywords and 3 offers domains.
But I came up with quite a few ideas how to approach this offer and will test them.
Thanks so much @dubmill for all the detailed updates! Very helpful.
Amy
Home > Paid Traffic Sources >