Home > Paid Traffic Sources >

---

TikTok Ads - Jumbleberry Ecom Offers - No Conversions (22)

---

Hi Guys! Just wondering if anybody has any success on Ecom offers through TikTok ads.

I tested various offers including Onnit Alpha Brain. Stats for Alpha Brain actually looks pretty promising, such as:

• 5sec average watch time
• Paid likes, comments, and shares are through the roof as well

Problem is with $60 spend so far, still no ATC's... Usually I am seeing an average cost per ATC of $5 or something.

ATC should be included in tracking, as seen below:



So it shouldn't be a tracking mistake. Creatives are also solid - based on average watch time and likes, etc.

Just curious how you guys are performing on similar offers with TT ads. Let me know!

---

Not that I'm having success with it either. Are you getting any View Contents/TOF events coming through in TikTok ads manager?

For me I noticed a huge discrepancy between clicks and View Content events on Jumbleberry which I'm working with their support team to try and figure out what's going on.

What does your frequency/cost per 1,000 people reached look like?

The offer pays out $83 dollars. Per usual I think you see this around but you should be ideally testing 3xCPA.

If your metrics look good, maybe aim for TOF events like View Content. Once reached to 50 conversion events(View Content) dupe and go ATC.

---

These are my stats so far for this campaign:



I am also not getting any View Contents inside the TT ad manager...

Honestly trying to understand this - it seems like a perfect offer, landing page, and creatives.

---

ScottyG posts mentioned not to bother running TikTok unless we have tracking down to the T. Now, I'm not as technical as ScottyG is and probably won't get tracking down to 100%. But to see the clicks to VC ratio being very off is alarming at this point. I think you're in the TT insider's group(Max Finn), I'm going to ask about this and see what his 2 cents are.

So I'm currently debugging this with Jumbleberry support team(see below) and they're looking into it. I know you're doing this for Alpha Brain and I'm reaching out to support for Javy Coffee. This very well might be a TikTok issue and not the network issue so that's a bummer.

There are a few things to consider here:

1) These pages that we're "direct linking" is probably a lot larger in size therefore taking a longer time to load for the viewer. From the time a user clicks the link and gets to the page they might've exited out the app. Hence we call this clickloss where you notice a huge variance between clicks and View Content(see point 2 below - User exits the page without it fully loading).
2) From the lead gen campaigns I've ran, most of my audience is on WIFI. That could or could not mean anything, but usually imo I would associate WIFI with being faster than 2G/3G/4G/Maybe 5G. Some audiences might be on 2G network and the page loads so slow for them that they might've exited out the page before it finishes loading. Even in SquareDance response below they say that the pixel doesn't fire automatically. It happens 1-2 seconds after all elements are loaded. So with our microwave society nowadays if we don't get something immediately we automatically move onto the next thing. I'm guilty of this as well.

The only solution I could possibly see with this is hosting these brand pages on our own servers. This allows you to change up the lander and input your own pixels. And tbh this is probably how to super affs run their stuff.

---

Every TikTok campaign I’ve run has at least 30% clickloss. Even higher when direct linking, as you’ve noticed— most offer pages are huge and slow to load and tiktokers don’t have time to wait (unless your ad sells them realllly well), they’ll back out immediately.

Do with that as you will.

---

Fuck.....I guess there's our answer. It's time to abandon ship . I thought I was the only one going through this. Thanks Jaybot for confirming

But not sooooooo quick Simon..... lol

I guess there's now only 1-2 ways to tackle this. If anyone has more ideas, please spit it out!

1) Host the landing pages on your own server. Again not a technical wiz, but if you're able to get the pages on a very fast server or host on CDN(probably not likely) then the clickloss rate can improve.

2) Generate "out of this world" creatives.

I'll go with choice 2 . I'll sacrifice my dignity and become the next island boy without the tats.

---

Seriously??

I hate having to deal with landers too, that is part of the appeal of TikTok and being able to direct link. But seriously?? Hosting landers is the easiest thing out there, and many drag-and-drop landing page platforms are built on a CDN. Go rip some landers and tweak them, or you can easily pay people to make one for you.

---

So I have reached out to my AM to get them to provide me files to upload to my own server. Is that how you would do it?

- - - Updated - - -

So I have reached out to my AM to get them to provide me files to upload to my own server. Is that how you would do it?

---

This would be the advantage of starting at the bottom, like pops or push. Learning how to use a server, rip and edit landing pages, host the pages, setup click links, get down and dirty with the tracker tokens, etc...

There's a ton of info right here on STM on how to to that. I'd say check out Amy's 40 day tutorial, but I think she's still using amazon s3 which is a pain to setup.

And me telling you to buy a dedicated Hetzner server is out of the question.

But you don't want to go backwards or start down rabbitholes that take awhile to climb back out, do you?

In these cases, I always suggest landerlab or convertri. Drag and drop. Hosting is taken care of. All you need is to setup your domain with namecheap and a free cloudflare account.

---

The thing is I did start out with pops. This follow along is collecting lots of dust

Correct, I don't want to take a step back and work on other traffic sources. I set a course to really try to master TikTok ads.

In terms of budget/capital, I do have a lot to work with so dedicated server isn't an issue.

I'm close to finding a solution to hosting the affiliate network pages on my own servers.

---

Good. Stick with Tiktok. Not trying to dissuade from that at all. If you've been around for 6 years, then you know to avoid shiny objects

My point wasn't jumping back into pops. It was getting good grasp on tracking and landers, which seems to be where you're stuck rn. fr fr no cap

Dedicated servers aren't expensive, I pay like $40 a month for mine. For a Hetzner linux server though, it does require some level of technical skill and being comfortable with the command line and ssh.

Sounds like you have a plan. So we'll go with that.

---

Since 2015 I been building and hosting my own offer pages. Need to get the edge from other 100k affiliates running same industry with same landing pages and targeting ...

But surprisingly, not very many host their own self-created/heavily modified offer pages, I guess this depends on niche. Why do I think that? Well, I created never done before presell/offer page for different more advanced niche(just recently) and I was thinking, hei, this could actually work really well in nutra. Haven't done any nutra for at least 8 years, but I use to be very very strong in there with high volumes - back then, JB was one of the networks with biggest caps and most of my traffic went there. I still have these guys in skype and chat time to time. So I hit them up, explained my story/needs and asked if they have some sort of API or some other method to self host/create offer pages for nutra:

JB:'' im sure its doable, just havent done it yet , but if we can start of with some steady volume, im sure we can make it happen.''
(This is direct, unedited quote)

To my surprise, they didn't have working solution for this. It's not really complicated for them to build JS form with API solution, but I guess demand for this is not there yet. Might be also, that most product owners do not allow it as it creates different kind of risk, as affiliates could build crazy BH offers with no TOS or other must be notices.

To OP, iirc, most networks can host your TT or FB pixels and launch them when CPA happens.

But advances from hosting yourself:
-Easier to host your multiple pixels without need for AM to be online AND fire pixels with different events, helps a lot with re-targeting.
-Better optimized pages(Could only load part of page at beginning and when user scrolls, it loads more).
-With API solution, NO TRACKING LINKS.
-Offer page domain rotations to avoid getting blacklisted.
-Product owners usually don't like this, but can get e-mails and do e-mail marketing part yourself.
-Element/Partial cloaking.
-Element rotation in offer page to optimize offer page for your traffic(If running TT traffic, could use some TIKTOK video in offer page).

There are probably many more positive things, but there are also loads of negatives, IMO positives outweigh negatives in all cases. You seriously do not have to be coding/server setup specialist to make this happen.

---

Thanks for chiming in desteny. I took another route and switched over to pops. jk @jaybot .

But since my last post, I was working with a dev to just kinda guide me through how to host my own offer pages. Surprisingly enough, it's not that hard to do it. Just get rid of some scripts here and there, and input some code and you're all good to go. Initially, I thought you had to use a service like HTTrack to download the files but you can just save the webpage and the files along with it.

The only caveat to this is I can only host the initial page(ViewContent) and nothing further than that(like AddToCart or Initiate Checkout pages). The upside from what it seems like(I tested this on Friday), I was able to get a lot more View Content events and a lot less clickloss. So for example out of 100 clicks I would get 60 View Content events which seemed to work better if I were to just direct link the affiliate network's offer page.

I recorded a Loom of what I did in terms of hosting my the offer page on my server. I'd appreciate if someone could take a look and see if there's anything off. https://www.loom.com/share/3a92520d1...984d0a883f0e59

With the API/JS method you're talking about, can you dive deeper into this? Is it getting access to the advertiser's server?

I can see how product owners would not like this for the reasons you mentioned especially with collecting emails and affiliates potentially redirecting to other BH offers.

-Better optimized pages(Could only load part of page at beginning and when user scrolls, it loads more). How does this work?

-With API solution, NO TRACKING LINKS. If we use API on JB, are you saying that we can't use a 3rd party tracker because the integration with the traffic source(FB or TT), JB will only feed back data to the traffic source and not the tracker.

-Element rotation in offer page to optimize offer page for your traffic(If running TT traffic, could use some TIKTOK video in offer page). This would be awesome! This builds up the idea of congruency from the ad to the offer page.

In the meantime, I've spent like $40-60 on optimizing for Initiate Checkouts without any events triggering for that. So at this point I don't know if it's my creative or if it's how I have my pixel/tracking setup. Nonetheless, I'm working on getting on a call with my dev and JB support team to hash this out. I'm going to take a step back from TikTok for a little bit until this is figured out while I learn more programming knowledge like JS+Jquery+PHP. I'd like to just run ads but I've come to the conclusion that if I need to debug some technical issues I must understand the basics of what's going on.

---

Good call, these guys have been pushing TT hard:
https://www.youtube.com/c/JavyCoffee/videos
They been using every marketing method known to affiliates - weight loss, girls with big butts and everything that goes between these 2.

I haven't seen the code, but I'm positive you could place pixel under the button ''BUY NOW'', unless it's iframed, but even then you could place pixel to fire when redirect to their page is happening. Again, ask the network to place pixel to their network, it's easiest. There is NO WAY any advertiser allows you to host payment page. You have to be very well established with advertiser for them to allow you to mess with their payment processing page.

Try asking that for CC submit sweep offer advertisers hehe

Ever seen this? Way easier to do it with websites.


Not really. API is a server, that is configured by your network/product owner. It's like automated helpline for computers.

When someone is filling in the form or survey or whatever and then presses ''buy now'' or ''register'', your server will contact the API with questions:
Offer_id
Aff_id
Consumer_firstname
Consumer_lastname
etc

If all is filled correctly, API will take the information and register to network/advertiser. But leaves you option to do whatever you want, yes, you can place after lead page there(self hosted), if you want, you could even place 5 of them there. To end transition, still need to end with redirect to advertiser/network page if you want consumer to actually buy something/fill payment information. API is only good for sending data to network with no redirects and more freedom to customize the full funnel flow. I would be happy to post tutorial docs from network, but not sure if I am allowed to do that. If network has API, they will have docs about setting it up.
I seen few crypto guys, who have been placing API lead form directly to presell page and creating custom after lead flow, where user is ''motivated'' to answer phone. Doubt it works, but who knows ... Hasoffers/Cake and probably all fancy network tracking platforms usually support it, just not all networks have completed the process of setting it up as it takes time.


https://i.gyazo.com/b5821c16c0cf4e21...1253fc8ae5.mp4
This is for heavy offer pages only!
If you don't want to click on the link, most social media platforms use this. They do not load all posts when you land, only load more when scrolling, this could be used on heavier offer pages.

Everything will work normally because the end is always to network/advertiser.

There is no such thing as just running ads, it's not 2010 anymore. Need to find a way to run ads with a edge.

---

Holy shit, I didn't even know they were posting content like that. They must have internal media buyers pushing hard on this offer. I've been focusing a lot on this offer on Jumbleberry too but FML I can't get it to convert. Again, I think it's back to my tracking which I think is almost solved. **Fingers Crossed**





So, in the hosted page, a script was added at the bottom and within the tracker Keitaro, I had to set the URL parameter action=ViewContent so whenever a user enters the page it fires the VC event back to TikTok. On the page where "CTA" buttons appear, I would just put this following code nested outside the <div>

Isn't that what cloaking is lol. Not looking to go the BH route.

So is this like where a affiliate can send a customer to upsell page after they complete the payment on the offer page?

Ah I see this with facebook all the time.

Sadly the truth. Like for lead gen, from what I understand the people that send large volumes have dedicated offers with way higher payouts than the normal affiliate or they have a CRM tool to collect all the leads and sell it on a auctioning platform or something like that. My hope with my advantage is that TikTok is still a relatively new source and not a lot of affiliates are stepping into this uncharted territory. But that comes with the frustration on my part on trying to figure out how to run offers lol. A lot of people that do run ads on TikTok I feel like are either doing lead gen(not affiliate way) or ecom(dropshipping or Shopify store).

---

On your CTA just use <a href="yourclicklink/click" onclick="ttq.track('AddToCart');"> so you can have an event for going to the offer page. Also, optimize for that before optimizing for Initiate Checkout. If not getting any traffic in TT, always go one step at a time: ViewContent, AddToCart, InitiateCheckout, SubmitForm/CompletePurchase

Better optimized landers are usually simple for ecom/leadgen. Not sure you need anything more than kraken.io to compress images and lazyload.js to not load images until it's been scrolled to. Should be able to google a few examples online of how to do that.

---

So I'm not sure if this will work. Because on the self hosted page I'm forcing the ViewContent event to trigger by the S2S postback I have set in the tracker.

The idea is to have the events flow naturally from the affiliate network offer page. But, the issue we have with JB or what my dev told me was that the subid wasn't being sent all the time. Which is why him & I thought of hosting the initial page ourselves just to trigger the ViewContent event.

This screenshot below is why I had to host the page on the server. I don't know what kind of networking/server setup JB has. But a self hosted lander seemed to do a lot better than direct linking to the affiliate network's offer page.



To expound further, here is another reason why I had to self host the page. This is from the Keitaro tracking log. Prior to self-hosting, my dev told me the subid wasn't being passed through. 1) This could be clickloss or 2) This *could* be because JB's system is slow? <---- I don't know that's just an assumption of mine.



To expound even further than that, here's a screenshot of all the postback ignored. These are all essentially "View Content" events that are missed and not recorded into TikTok's system. So in a way it's taking longer for me to optimize a campaign on TikTok. Plus what I do notice is that if I run the campaign for 1 day the VC seems to come in semi-normal, but on like day 2 afternoon & time after the View Content events are not trickling in that often, if I were to estimate probably 1 every 2-3 hours.



So in the end, if the View Content events would've flowed in naturally if I direct linked the campaign I wouldn't have even thought of going the self-hosted route. But, with the method you mentioned again I don't know how that would work because once a user clicks the CTA button it doesn't really go to ATC event.

After clicking the CTA button it goes to the network's offer page which displays the event as a Lead. With JB Lead equals ViewContent on TikTok.





If I'm going one step further into the next page. It's View Content.



And one more page further, then it's Initiate Checkout.



The final step would be CompletePayment/Purchase, but I won't be making any coffee purchases as I bought 3 products from them just to film creatives So I can't arbitrarily set further events like ATC/IC because of those reasons above. I want the user to go through the normal flow and if they land on the initiate checkout page it should feed back the IC event back to TikTok. Plus, the funnel built for Javy Coffee doesn't include ATC. Please let me know if I need to clarify anything here.

---

What? Why? Just have your TT pixel code in the <head> and then you can call ViewContent on the page load. No need for a tracker or API or anything fancy at all. It's straight javascript.

So basically, in the <head> section of your html file (in your self hosted lander) you should have the code for your Tiktok pixel:

And your TT pixel and functions will be loaded, plus 'ViewContent' will be fired directly to TikTok on page load.

Then you can call the ttq function at any point in your lander, like wherever your CTA is , most likely an <a href= "yourclicklink">Click here!</a> which we would simply add the onclick javascript code to fire the tiktok pixel function with 'AddToCart' just before it sends the user to the offer page.

---

Hmm, then I don't know why we're doing it the way we've done it now. Because your method makes sense to me, I'd have to talk to my dev about this.

So to clarify on page load it'll trigger the VC event and then if the user decides to click a CTA button/element it'll trigger the ATC event. So I guess that's what I'm confused about.

Do we actually want the user to hit a page where the ATC event is actually installed on the 2nd or 3rd step of the offer page or are we making this up to trick TikTok that the button clickthrough triggers ATC?

---

s

Correct. So the the pixel gets all loaded up on your page and creates a ttq.track function (same as fbq function for FB pixel, for future reference), then we can fire that pixel directly from your landing page whenever we want. This is why pixels are nice when we have control over the page. Since it's your lander, we do. Then on any clickable object, you can put the onclick="ttq.track('AddToCart');" . You can test if it works by using the Tiktok Pixel Helper chrome extension.

As for AddToCart InitiateCheckout CompletePurchase, etc. Tiktok algo doesn't actually care as far as I can tell, as long as the steps are in order (so it's logical you'll get less SubmitForm than Initiate Checkouts, which are less than AddToCart, and less AddToCart than ViewContent, etc. At this point, you're just trying to get traffic through the funnel and help the algo get there. And considering the JB shit doesn't seem to be working well, if at all, it's good to show TT that your funnel is working and you are getting traffic to your initial pages.

---

You're a genius @jaybot. I can't believe I didn't think of anything like this. This seems very simple but this is why I implore people to throw stuff out there and have someone chime in because someone who has more or less experience can have a different perspective on how you run things. All this time I was thinking I have to match up with the funnel that Jumbleberry shows. I keep getting ATC/InitiateCheckout mixed up. I'll do multiple tests on this and will report back.

---

Just a quick update to this. In the screenshot below, here are the results. It's been confirmed that there's some reporting issues happening with JB/TT and they're looking into it. Hence the first campaign, you can see 0 VC after 54 clicks.

The 2nd campaign is where the pixel is on page and onclick it fires the ATC. This was Jaybot's method

The 3rd campaign is the one I had previously where the TTCLID is passed through a script/php file on the Keitaro server and then goes to TikTok.

Between the 2nd & 3rd campaign. The results are similar. And at this point, I'm waiting to hear back what JB is doing to fix the issue.

---

Home > Paid Traffic Sources >