Home > Native >

---

Extremely Low LP CTR When Running in Taboola (19)

---

Hey STM Friends,

I used to hear Taboola has probably the highest quality in Native ads. So just grabbed a great offer based on Spy tool and run it in Taboola:

Offer: Ecom
Geos: AU, GB, NZ
Platform: desktop
CPC: 0.59 USD smart bid
Daily budget: $50
Creative AB test.
Payout: $40

After running for 2 days, this is the results:

Spend: 131 USD
Ad CTR: 0.32%
LP CTR: 3.25%
Conversion: 0




We could see that this campaign is running in more than 200 sites but each only has a few clicks. Also the LP CTR is super low.

In this case, it doesn't seem possible to optimize anything. Even if I want to block widgets with 0 LP CTR after 30 clicks, none of the widgets has 30 clicks.

Have you ever met this kind of case before? Is there anything I've done incorrectly to cause the overall low LP CTR? This campaign seems to commit suicide if it continues.

Thanks.

---

Are you targeting all 3 geos under the same campaign?

If so, I would recommend you split them to separate campaigns although you are targeting english speaking countries. The more countries you have in targeting the higher the testing/optimization budget needs to be.

Also, $50 daily budget for a $40 payout seems too low. Increase your budget, as well as block No LP CTR sites.

Regarding the 30 Clicks no LP Clicks threshold, how did you come up with that?

Last but not least, I would as well double check the landing page if there are any issues that might influence the low LP CTR, or even check for a better landing page.

---

Interesting...

I've definitely had many campaigns just like that...

First of all, you gotta remember that the vast majority of campaigns you start will fail, so its not weird that that is the case. It would be highly unlikely your first camp started off successful.

Ecom is also VERY difficult (though the rewards are very high if you can get it to work).

I like running multi-geo campaigns, but in this case I agree with @platinum that one geo would be better, so you can isolate the variables more. This could have shown at night in the UK for two days in a row which would be vastly different than daytime for instance.

Also did you run Fixed Bid or SmartBid?

If fixed bid, they probably sent you some bad sites you'll need to block.

If SmartBid, you gotta give it a a few days to throw the kitchen sink at you before it starts optimizing. The problem with ecom is that you'll get very few conversions for it to optimize on (unless your some wizard who can create all sorts of events with google analytics and stuff like that based on time on page, scroll depth, etc).

The biggest question I would ask though is whether those two screenshots are matching - like are they for the same campaign and same amount of time? If so, your tracker has about 200% click inflation (as does mine usually with Taboola), so your lp ctr was actually 31/254, so about 12%, which is pretty normal for ecom.

Also, the sites listed on the Taboola screenshot are all pretty decent, so its possible those ones on the tracker screenshot (if that's what the top one is) are sites that had clicks you didn't get charged for (bot clicks, spider clicks, etc).

So yeah, if that is the case then it looks like a pretty average start on ecom. I'd switch to AU or NZ only and only 7 am to 7 pm, 7 days a week, or something like that, give it another 5 days at 50-100$ a day, and then see how it looks.

With high payout ecom offers you could be profitable overall while having periods where you spend 400-500$ without a conversion, just based on normal statistical variance, which means it is expensive to test high payout offers, but at the same time, if you're just starting, you gotta test lots of stuff to find winners. So its tough.

---

Hey Platinum,

Thanks for your suggestions. Wow, didn't expect you reply so fast

Yes, I target all 3 geos in the same campaign. I thought running them all together might increase the ad ctr and lead to low cpc. For now, it doesn't seem to work so well.

I set up this rule based on some threads I read in STM and some courses about native ads. What's your suggestion of blocking those 0 LP CTR widgets?

Checked the landing page myself, made the landing pages with Landerlab. All the buttons work well. Yep, that's a good point. I'll find some more landing pages set up new campaigns that only target 1 geo.

---

Hello Jack_l,

Yeah, it's kind of weird that in Taboola, both ecom and leadgen offers have very low LP CTR, as low as 3%. I saw the similar LP CTR a few weeks ago when I was running auto insurance lead gen in Taboola. Initiall, I thought maybe Taboola users don't like Lead gen offers but now they don't like ecom either.

I'm running smart bid with the base bid of $0.59.

With Smart Bid in Taboola, usually how many conversions does the system needs to start optimizing?

I thought Taboola only optimize for conversions, but we're not passing those events(time on page, scroll depth, etc) as converions, right? If I pass time on page as an event to taboola, do you think it will help taboola to optimize campaigns for purchase?

Yep, this is something I felt very weird when checking the stats. Those 2 sceenshots are taken at the same time. Both tracker and Taboola are in the same timezone.

Does this mean I don't need to pay for those inflated clicks?

Yep, 12% LP CTR looks normal but still not so good.

What kind of offers would you suggest to start testing in Taboola? I've tried lead gen in Taboola before, spent 500 USD but only 2 conversions with 20 dollars.

Thank you~

---

Where are you getting your landers from? You mention trying lead gen before with almost no conversions as well, did those campaigns have a low LP CTR? If so, that strongly suggests your issue is the landers. Either they simply aren't very good or there is a big mismatch between the creatives and the lander.

Also, looking back at your initial stats. You've only spent 2x payout. You don't have that much data yet.

---

Taboola does have a 2 hour lag when updating stats. But.

Check the country and state in your tracker and see how many are coming from Virginia in the US.

Those are usually checks by Taboola to make sure your funnel is legit.

They do it a lot, and they'll still show up under a normal widget, but you won't pay for them.

You can check your tracker clicks and balance against the Taboola dashboard and see how they line up.

I got like 300 clicks from Newsweek in my tracker, but like 2 on Taboola.

---

Yeah, so for whatever reason Taboola seems to have this issue where your tracker will register tons of clicks that either aren't "real" or that you aren't actually paying for, or are just like bot clicks or spider clicks, I don't know, but as you can above, your tracker stats are way different than your actual Taboola stats. This is a bummer as it makes it hard to optimize based on any stats from the tracker, but it is what it is. I don't know a solution for it.

As a result, I only optimize based on the data in the Taboola dashboard. I pass the conversions from the tracker to Taboola to do this.

Usually this is a "sale" or "lead" (whatever I am getting paid for) but if you are really technically savvy, you can set up all kinds of crazy conversion events using pixels/google analytics/various SaaS's/etc. I don't do that, but its a possibility.

Also if you are running ecom stuff with DFO or GiddyUp or Jumbleberry, they'll often give you the option of firing conversions on more front-end things like click to offer, add to cart, etc. So you could certainly pass one of those to Taboola and let the SmartBid optimize based on it, instead of only passing actual sales.

But yeah, any offer you see running on Taboola/Outbrain/Gemini is probably a good one to test. Testing low cpa ones will just be cheaper.

Again though, the vast majority of things you try won't work. In my experience at least its like 92% of things I try fail, 4% breakeven for awhile and I can never get them super profitable, and then 4% work great and account for 90% of my long-term profits.

---

Hey Iwanttofly,

Thanks for the comment.

I got the lander from Spy tools and tweak it a bit.

Yep, when I run lead gen in Taboola, the LP CTR is so horrible.

Oh, that's a valid point. Perhaps I'll also find more landers to test as well.

---

That's interesting. I checked the states as well. I didn't see any clicks from Virginia. But lots of clicks from {site} when I check the widgets. Probably Taboola is checking the campaigns as you mentioned.

---

Hey Jack_l,

Thanks for sharing this hard truth. Only 4% of those will be easy to convert and be profitable, it makes me wonder I didn't test enough. Even if this offer is not converting, I'll test another.

The lowest CPA offers that we can run in Native but still have a chance to be profitable these days are probably Search Feed Arbitragy which Platinum introduced in this forum.

I'm so curious what's the other types of low CPA offers that could still work for Native:

Dating: I thought it really depends on the conversion rate. However, if we use some aggressive stuff, the purchase rate for users will be low. We'll be soon kicked out from the offers.

App installs: could be a sexy vertical. But offer ownners will also check ROAS or retention of the users affiliates generated. Also, it's so difficult to get those offers. I've never seen any Network providing this kind of offers.

What's the other verticals?

---

Yesterday, I paused the original campaign and live another new campaign with the following details:

Geo: AU
Platform: Desktop
Bid: Smart bid, 0.48 USD
Daily Budget: $70

Luckily, I got a conversion this time:
Day 1
Spent: $77.73
Conversion: 1
Revenue: $40
ROI: -51.46%

Clicks in my tracker is still inflated. But anyway, I'm going to give it more days to see if anything pops up.

---

Dating works great on native... app installs it depends, but some of them like financial app products and "casino" ones can have payments at like the $20 level on native.

Just do tons of spying, that's the best way to see what's actually working and find good offers.

In general though native should have very high quality compared to other traffic sources. Most dating offers you see on affiliate networks that take all sorts of traffic will probably earn higher payouts once you prove quality on native. Same with other similar verticals where backend quality is important.

Natives are mostly served on news sites, so you get lots of Boomers with disposable income, very good for quality.

---

After running this campaign for 1 week, I only got another 2 conversions. The conversion rate of this offer is too low to continue. It also gives me lots of stress. Every time when I wake up, I do nothing but check the stats in the tracker, hoping the 1 or 2 conversions might pop up. However, Most of the time, this offer just doesn't convert. I decided to try something else. Please see below for what I'm running now.

---

Hey Jack_l,

Thank you for the suggestions. These days I saw lots of search ads while doing the manul spy in some sites. I've decided to give the search ads a shot.

So here I set up a search ads in Taboola:

Platform: mobile phones
Geo: India
Bid: $0.06, smart bid
Creative A/B test
Daily cap: $50, then added to $100
ROI(until I wrote the post): -30%

I run a special niche in India under the recommendation of my AM. While writing this post, this campaign has already been running for 1 week. All the optimization I've done is blocking the worst widgets via some rules in Optimizer.


In Taboola dashboard, it looks like the actual CPC is lower than the bid I set in the campaign. And the campaign usually spends the daily budget in 10 hours. Does it mean I bid too high for this campaign? If I lower the bid, will it affect the performance?

For those widgets which converts well and ROI positive, is there a good method to let them running more volume while maintain the ROI and not disturbing Taboola's algorithsm?

I knew @jack_l has experience in running search ads. Could you please give some suggestions? Thank you

---

So I have mixed feedback...

To be totally transparent I have stopped running search arb (at least temporarily), HOWEVER, if I was starting native from scratch right now, search arb would be my top recommendation for what to run... just because its so incredibly cheap to test it, and because you can run it in ANY geo, which is a HUGE advantage...

Also, India is GREAT for search arb. I had some really profitable India search arb camps on Taboola, however the search arb outfit I was using kept getting complaints from Google about 'bot widget' type traffic, and I was having to spend like 30 minutes a day manually blocking their continually updating blacklist... so I stopped running there... India/Pakistan/Indonesia were the only places I really had that problem though... everywhere else was totally fine...

But yeah, when I was running search arb my strategy was to start at like 10-20$ a day and evaluate the potential based on conversion rate and payout per search... if it looked doable I'd continue it, otherwise I'd turn it off after a few days. In my experience only a small minority of search arb camps ended up profitable, so the key was to only lose a small amount of money on the losers, so the winners could make up for it. And also to test tons of ads. Different ads and headlines can make a HUGE difference.

I did have some really profitable search arb camps too, I think I made like 6k profit on Canada Credit Cards one month on Taboola... and maybe 2k profit one month on Australia Funeral... but starting in May I just could not get anything to work so I turned it off...

With that said, again, I think if you're just getting started on Taboola search arb is absolutely the best thing to choose... it will let you learn the network really well, explore lots of geo's and languages, etc. So don't be discouraged by the fact that I stopped it. My rep at Taboola said she had a client spending 100k a day on search arb profitably, and that was also in May/June, so its possible I just had bad luck or was doing something wrong, etc.






Edit: re India I was spending like a penny a click... although it might be more expensive now...

---

Hey Jack_l,

Thanks for the suggestions.

Search arb is a really interesting niche for native. I'm going to run it for a while.

Btw, when setting up campaigns in Taboola for Search arb, do you usually use Lead Gen, Purchase or Website engagement?

Do you use some external tools like ahrefs or semrush to check the search volume and CPC for some words?

Thank you.

---

I always did Lead Gen. I use Lead Gen for basically everything. And I passed conversions straight from the search arb provider to Taboola and then had Taboola on SmartBid in order to optimize based on all that conversion data. I also used the same conversion rule for all the search arb campaigns, rather than setting up different ones based on niche. Not sure if that is best approach or not though. Also I optimized for the 'click' rather than the 'search' (the middle point in the funnel rather than end point I mean) so that there would be as much conversion data as possible. Again not sure if that's optimal though or if optimizing for the 'search' would have been better since that's where you actually generate the commission.

And then yes, I did that a ton at first, but I don't think it really helped to be honest.... I think the best approach is to just copy what's working for other people, as far as niche selection I mean...

And your reps at the search arb agency will know what's doing well so if they suggest something its usually worth trying too

---

Hey Jack_l,

Wow, that's so inspiring. Thanks a lot.

I also use Lead Gen as campaign objective.

As for optimization goal, I'm using "search" for now. But I also asked network pass land, and click back to Taboola no matter which niche I'm running.

Not sure what's the most optimal way but we'll probably do some tests in the future and share it here.

---

Home > Native >