Home > Native >

---

Still Here (36)

---

As the title says, still here.

After a bit of thought, I decided to stop fighting. My experience in Facebook was always Lead Gen, it is what I first learned and what I've always done the best with. What burned me out was dealing with clients and agents, that is what had pushed me to AM in the beginning. Sadly, I went too far and wanted to get completely away from Lead Gen. However, I realized don't fight what you are good at. And so far, Affiliate Managers are much more pleasant to deal with than clients and agents.

So, I have started up an auto insurance campaign in the US. And here I am, sending leads to a company I once bought leads from, and hated. They are what drove me to learn lead gen in the beginning.

Vertical: Lead Generation
Niche: Auto Insurance
Geo: US
Traffic Source: Taboola
Device: Desktop
Lander: Template in Landerlab, Ripped from the Wild
Ads: 3 Images and 3 Headlines
CPC: Started at $0.30, SmartBid
Budget: $100/day
Payout: Varies, prime is $14 and what campaign is optimized towards. There are 4 offers and the offer URL is a smart link that day parts as well as routes by device type.
Hours: 8 - 8 M - Th EST, 8 - 3 F EST, basically prime call center and agent hours in the US

So I started by spying to see what was out there. It gave me a good idea of what people are running, and combined with my own knowledge I came up with 3 headlines. The best performer so far is very similar to what is already out there, which isn't surprising. As far as images, I tried to look at what was running and then find something similar. So far the best performer is a lark, I decided to take an angle on a common main image on landers. It appears to be working. Yeah me!

I had prepared 2 landers, but planned to submit them separately as I understand Taboola isn't a big fan of lander/offer rotation. In the campaign notes, I did tell them what was going on with the offer, so far no issues there. Sadly, the first campaign was rejected for the lander. It was one I had found in the wild and imported to LanderLab. Taboola didn't like the image on the lander. Fortunately you can now chat with the content moderation team and they told me exactly what the problem was. He also found an issue with the Click URL. It turns out there was a script that was modifying it. I submitted it to LanderLab support and after a few tries, Ervin was able to find and remove the script.

In the meantime, I went ahead and built a campaign with template lander. It went live on Friday and I immediately noticed it wasn't getting much traffic. So I bumped the bid to $0.35 and immediately started getting some traffic. I ended the day with 2 conversions before 3pm and was slightly profitable. Very exciting so far.

Yesterday I almost immediately started getting conversions, but I noticed around midday that it wasn't on pace to spend out. Therefore I bumped the bid again to $0.40. I kept getting conversions, but sadly it started throwing money at the placement that converted on Friday. It was a let down yesterday and went very negative on that placement. Fortunately 2 other placements started to deliver and left the campaign slightly negative for the day and over all. I reduced the bid for the unprofitable placement by 20%, I have one more that I'll probably turn off as it has spent close to $30 with no conversions.

Placements. Top one has had bid reduced and will probably stop the next one shortly.



Creatives.


One thing I'm still not clear on as I've heard it a few different ways on here. For Taboola, is it a good idea or not to turn off creatives that are getting a lot of budget AND have converted, but are unprofitable?

My plan is to continue to optimize this campaign, I am pretty positive it can be made profitable long term. Once I have a better idea of placements, I plan to launch on the other lander as well and continue to add in more creatives to find the best performers. Once I start to scale, I plan to add mobile as well.

I will close with this. No matter how this Follow-Along goes, one lesson I think anyone can take away is, stop fighting. If you notice something you do well with, double down versus trying to get away from it. I had to learn this lesson the hard way. Right now, it really feels good to see a campaign with such potential, and the much lower payout makes it so much easier to optimize. It also doesn't cause near the stress to look at the dollar amounts in play either.

---

Great new start @iwanttofly!

I'm sure you're going to get far better results with this campaign compared to the previous one.

I guess you've seen the thing of "mastering one traffic source" then move to others. Not that you're jumping around from one traffic source to another, but there are some things that are a BIG NO NO for Taboola.

As long as you are using Optimized creative traffic allocation, you DON"T want to BLOCK any of the ADS, even if they are not profitable. Your screenshot above, is proof that most of the traffic is going on a single ad when using Optimized traffic allocation.

Ads are being tested against multiple sites if they find a good match or not, so you cannot block ads until you've removed under-performing sites. And a campaign can receive traffic from a few thousands of sites, so even statistically it doesn't make sense to pause ads and not sites. Don't compare it with Facebook, it is a single site/app and therefore blocking ads makes more sense in that case.

Keep a close eye on your landing page click-through rate and lander-to-offer conversion rate. These are two metrics that will help you get the campaign in shape pretty quickly.

Here's a summary of what I would do (considering your payout and setup):
- Start the camp with an average high bid ~ $0.6 - $0.7 Smartbid
- Let it spend for at least a week to send some conversion data to Taboola
- While the camp is spending for a week, I would start blocking sites right from day 1 if they deliver awful low LP CTR
- Also, block any sites that has spent 1X or 1.5X the average offer payout without conversions.
- After the first week, I would go deep on the collected campaign stats to identify poor performing patterns to adopt my optimization strategy.

Furthermore, if you see a site that is converting well, but above the goal CPA while having most of the conversions DON'T block it, instead slowly drop its bid. If such sites are left on SmartBid, Taboola will still play with the Avg.CPC from -50% to +200% from the current site level bid. Keep this present when dropping the bid on the site level. If after dropping the bid you still get high avg. CPCs you may want to consider switching the site to fixed bidding (make sure you have enough conversions to reliably force the site to the lower CPC).

---

Thank you @platinum, I appreciate your response.

Do you know what a typical vCTR is on Taboola? On this campaign I am average 0.26%, which seems low. Looking at my first campaign for Goli on Mobile, it was around 1%. Different vertical and device type, so I'm sure that will play a role as well.

---

Hey man - great to see another follow-along from you! I agree - I think this is the perfect offer for you! Way cheaper to run/easier to optimize, and has the potential to scale like crazy if you can get it in the green

Look forward to seeing your progress!

---

Man we do things literally the exact opposite from each other @platinum! lol

We never block the sites but block tons of ads...

You know infinitely more about this stuff than I do though so I'm going to test out your system! Starting a couple camps now with some proven ads only and am going to restrain myself from messing with them and see what happens

---

Yeah, that is what I was mentioning earlier. I've seen it both ways on here, probably from you and @platinum.

Performance seems to have died today, I'm thinking the vCTR is just too low. So I'm working on some new images and headlines.

---

I've seen all kinds of strange optimization approaches but rarely like this, now you made very curious!

Can you elaborate on the logic of blocking Ads vs. Sites? What would be the reason to block tons of ads when the average campaign gets traffic from roughly 3k-5k sites?

If we were to compare running on Native vs. Social, obviously it makes sense to focus on the ad, cause by the end of the day it's one single site/app for which one has to match the best ad that can deliver results within the goal CPA. But on native its a totally different story.

Just think of it as if you were to run redirect, push, or to make a closer comparison consider running display traffic on Google. The banners gets served on thousands of sites, some of which are reputable and some of which that are total junk. In such situation what would make sense to optimize on, ads or sites?

It's virtually impossible to match suitable ads for every single site that serves them, when there are thousands of sites. Hence this is why its more efficient to focus on blocking under-performing sites rather than ads.

Obviously the above logic may be invalid in case a campaign is running on a concentrated whitelist.

---

Well, I'm not saying its logical, and I'm certainly not claiming its the "right' way to do it at all - its just what we've ended up doing lol...

On Outbrain and Revcontent and Gemini we block sites aggressively, but on Taboola it just *seems* like we do better when we just let the SmartBid do its thing...

I don't know if this is accurate, but I had someone tell me that the Taboola SmartBid is more like the facebook pixel in that it is taking into account like 20 variables connected to the individual USER on the other end, and that those variables supercede the SITE they happen to be on in importance.

Now, obviously there's some sites that are just bad - due to the demographics of who goes to them, or due to bad ad placements causing fat finger clicks, etc- but if its true that Taboola's SmartBid is focusing more on user history (what other campaigns has this user 'converted' on in the past) and interest group/demographic data, then I think it would make sense that the site isn't as important when using Taboola SmartBid vs all with other native strategies (especially something totally manual like Revc or Gemini where optimizing by site is a huge lever).

Actually... I believe it was @thedudeabides who catalyzed me to go in this direction... as he was always saying not to worry about blocking sites on Taboola (and to a lesser extent Outbrain). I think his rationale was that there are thousands of sites, so it would take insane levels of spend to ever block all the bad ones anyway, so if a campaign doesn't work with full r-o-n targeting then its not worth it anyway.

That rationale is a little different than my rationale about how the SmartBid targets clicks, and I can't remember where I heard the latter and don't know if its accurate.

All I know is that I at first I thought @thedudeabides was crazy, and I used to block sites left and right on Taboola SmartBid, but once I followed his advice and started ignoring the site section my profits/volume went way up.

And then re: ads, it seems as though the ad is the biggest variable in most funnels (outside the offer). I know in Search Arb and low-cpa dating offers we run on Taboola, there can be a wild difference between ad performance just based on small differences in headline, let alone the image you use. So that's why we test so many ads and block the losers so mercilessly. Indeed 90% of all the profits we've made on Taboola have probably come from just a tiny handful of all the thousands of ads we've tried.

Again though, I was only bringing up the difference for interest sake - you probably know more about the inner working of native ad networks than practically anyone alive at this point lol- so certainly wasn't suggesting your approach is wrong!

Its very possible we've succeeded DESPITE the above strategy rather than because of it, and maybe it just takes longer to get in the green our way vs blocking sites early on. The SmartBid does block sites on its own that do badly, and it does quit bidding on bad sites - I can look at any of our campaigns that have spent like 40k+ or more and none of the top 30 sites in volume will be really poor performing ones like you would see at the start, because the Taboola SmartBid quits sending the bad ones traffic at some point. But I guess the question is whether its worth it to leave that up to the SmartBid entirely, or better to block sites manually early on.

I will say that on Outbrain's equivalent of SmartBid I do not notice the same thing - it does seem better than r-o-n, but it doesn't seem like it quits sending traffic to bad sites the way Taboola's SmartBid does. So on Outbrain's SmartBid equivalent we do block sites very aggressively.

Its also possible our data is skewed by the fact that we do so many low-cpa campaigns like Search Arb and Dating, because that makes it far easier for the Taboola SmartBid to do its job, vs running ecom with big payouts... and indeed we've NEVER had a high-payout campaign work on Taboola, so perhaps the above strategies are what has caused our struggles in that regard.

---

Indeed most native ad networks are trying to go in the direction of something similar to FB pixel, but I personally think they are still far away from that. The main reason, is that they don't have control on the publisher site like FB or any other social network might have on the layout and content category served to the user. Plus "Content discovery" is a bit different from social networks and UGC in general.

If I'm not mistaking Taboola used to call SmartBid "auto-blocker" or something like that. Obviously it does add some publishers to the campaign suppression list, but at the same time it focuses on delivering clicks at the optimal rate with the aim to back up things for the publishers, network and the advertisers. To put it simple it's the supply side earning in a RPM vs. the demand side buying in a CPC model.
With that said, the network will deliver traffic from a specific publisher until it reaches a certain significant amount of clicks before moving the publisher in the suppression list, and the level of optimization aggressiveness is far from the one used by direct response advertisers.

Getting back to the pixel, all conversion categories used when creating conversion events should be definitely connected to an audience and the data is used by the traffic source algorithm. But again I think that when new publishers are introduced to an ad network inventory, factors like audience, site, bid range and offer type enter in a test to discover if such combination works. Whenever this happens such tests can cost a lot before the traffic source automatically removes a site from delivering to a specific campaign, hence that is why Taboola is providing a block list of 1500 sites per campaign + 1500 sites on the account level.

If there was no need to block sites I guess they would've been pretty happy to narrow down these limits and force campaigns to run using SmartBid optimization only.

There are a lot of factors influencing the success rate of a campaign. Starting from the Ad (image+headline), presell, then the offer itself.

Offers with lower conversion friction are definitely easier to optimize, especially when the conversions are reported back to the traffic source. It spends less time and money to tell the algorithm what works for the promoted content, but... when trying to scale things up it the traffic source will still need to enter some kind of a soft learning phase IMO, and yet there are factors that the traffic source cannot take in consideration.

Let's consider these three scenarios:

Scenario A: Search Arb Campaign | Payouts: ~ <$2 (depending on the vertical, etc) | Roughly 15% - 35% CVR
Super simple conversion flow with virtually no friction at all, considering that the user won't have to enter his personal information in an unknown website form, or take out his CC to pay for something. Also it has no presells to influence the decision making of the user (at least in terms of what we use landers for).

Scenario B: Lead Gen Campaign | Payouts: ~$8 - ~$20 (still depending on the vertical and post-conversion flow) | Roughly 8% - 10% CVR
Simple conversion flow. The user just has to answer a few qualifying questions then enter his contact information. After this point depending on the offer, the advertiser chooses to either send the converted user to a 3rd party offer or smoothly end the flow.

Scenario C: Straight Sale/eCom | Payouts: ~$15+ or even much higher (payouts vary from product to country, etc...) | Roughly <1.5% CVR
I would consider eCom offers with a high conversion friction. After all the user has to be convinced that the product that got his attention is good and will be delivered as fast as possible. Then here again we have up-sells and cross-sells that are often introduced to increase the average order value.

With the last two scenarios I guess we all agree that the presell (landing page) is a very important factor when it comes to the success of our campaigns. It is actually the one responsible for influencing the decisions of a potential buyer/converter.

Now given that when we use our own landers we tend to introduce improvements to it, in some cases, such improvements can almost double the conversion rate from the lander to the offer. And when this happens, it will directly influence the optimization strategy if we want to give a fair chance to all the publishers we are getting traffic from to get even more conversions. Same influence applies when we rely on the traffic source algorithm or on our own optimization strategy.

Related to the above, I guess it comes down to how much are we willing to risk on a new campaign before we get it into the green zone.

When leaving everything up to the traffic source by swapping ads we're somehow trying to trick the algo to deliver high quality traffic during the testing phase of an ad. Some networks do this all the time in automatic when a new ad is introduced to see if the new variation beats the previous champion, but it's not always the case.

On the other hand, we can easily start from a baseline threshold (usually offer cvr for the initial set) and work on that. By this I mean that, if we're promoting an offer paying out $8 a conversion, with an average 7.5% success rate from the lander to the offer, while paying ~ $0.45 a click, we're definitely not getting anywhere with a ~17% LP CTR. No matter how much we spend on a given set of publishers, we'd still be in the red zone.

So instead of waiting for the traffic source algo to make wonders on our campaign, we'd rather start blocking low landing page ctr publishers that have no chance of becoming profitable, and make improvements to our lander in order to maximize results.

Last but not least, we have to take in consideration that Taboola, just like other networks (Outbrain, Revcontent, etc...) assigns "inventory lists" on the account level. This mostly depending on the main verticals running on a specific account, but not limited to.

Hope the above makes sense

---

Thanks again @platinum, I really appreciate you sharing that insight and I plan to utilize it.

Considering the low vCTR, I added 2 more headlines and images yesterday, for 4 new ads. One of the images was rejected for low quality, it was basically shot through a windshield so wasn't the best. Would you recommend continuing to add new creatives until the vCTR comes up? I believe that is probably holding back the campaign at the moment.

I finally picked up a few conversions late yesterday. A negative day, but still early, still gathering data at an at -24% ROI as measured by Taboola. A definite improvement over the eCommerce.

I didn't go completely up to a bid of $0.6, but I did bump it another $0.05 to get more traffic until it is spending out each day. I also further reduced the bid for the top spending placement. It has 2 conversions, but it is sucking up traffic with little return at this point.

Placements (The first is at a -50% bid and the second has been paused)


Creatives

---

Very productive and interesting @iwanttofly

If you are looking for Interesting campaings please reach out via Skype.

live:.cid.43eae0cb8485de1b

---

Thanks for all the info @platinum - to say you are a wealth of knowledge would be an understatement

Everything you're saying makes perfect sense.

You also inspired me to dive into my data and see what is actually going on...

I have a 6.2$ cpa offer in Spain I've spent 4.2k on in 2 months at an average $4.90 cpa... of the top 50 pubs only one hasn't converted (#48 by spend) and the SmartBid has blocked 4...

So I was impressed that there was only one widget with 0 conversions in the top 50, as it implies there's not a lot of 'bot widgets' that are entirely useless.

However there are definitely a decent number of widgets that are -50%, -80%, etc that haven't been blocked by the smartbid yet... so it does certainly seem as though the overall profit could have been increased had I blocked those ones earlier on...

But yeah, fascinating discussion my friend! Appreciate all the wisdom you've shared

---

So an update and a few questions.

My AM was kind enough to put me in touch with someone at Taboola. It took a few emails and having to CC him, but got her to respond. She quickly hopped on a call and get a Rep assigned. I'll have to increase my spend some to keep him, but nothing scary. Just need to get into the profit zone or test even more. He did have some great advice, he suggested including Tablet with Desktop. I check with the AM and it pays the same, so done. That also really brought the campaign back to life. It seemed to be getting throttled before that, but now with a higher bid and Tablet traffic, it is getting traffic and conversions again.

I also started up a Mobile Campaign. At first most of the ads were denied. Fortunately Taboola now has Moderation Chat, so they told me what was wrong, gave me feedback on my proposed new Headlines and then approved them once submitted. If you are running on Taboola and are online during US hours, I highly recommend taking advantage of Moderation Chat, particularly if you don't have a rep yet. My AM said he was sharing that with all his partners as well. Still won't say I love Native compliance, but it is far, far easier to work with when you have problems than Facebook.

And even better, as I was writing this, I just got access to the Taboola API.

So a few questions.

1. On LanderLab, is it possible to set up a Listicle page? I have noticed in the past that a lot of Lead Gen sites will open the target site in a new window/tab and then change to show several other offers. I would love to do the same to increase revenue. Does anyone know if LanderLab can set up the page to swap out or redirect to that as it opens the offer page? And more broadly, is this something that networks or offer owners would object to?

2. Kind of a RedTrack and LanderLab question. I added the LP View script to my lander so I could more easily see views vs clicks and have RedTrack compute the CTR for me. Unfortunately it doesn't seem to be working. I've reached out to support, but is it possible that LanderLab isn't adding it correctly? I pasted it into the Custom Code section, just above where I pasted in the Taboola Pixel. The Pixel is firing, so I would think the LP View script would as well.

3. Getting conversions and not a horrible ROI. $537.21 in spend and $346.50 in payout, roughtly -36% ROI. I guess I'm in the phase now of gathering enough data to find the sites and ads that will convert, hope it is enough and then scale on those?

---

Also many thanks for your insights @platinum

I take your point that you need to weed out the under-performing sites before drawing any conclusions about good and bad ads. But after that stage, surely there's going to be a 'winning ad' that converts best? Different ads will convert differently over the many sites on Taboolas network, but ultimately there's going to be a 'best ad' that on average across all the sites is going to have the best ROAS?

Problem is that Smartbid and Optimized traffic allocation does not seem to have ROAS as its top priority, whereas we affiliates do. I'm guessing the most important factor for Optimized traffic allocation is Ad CTR because that's what pays the bills for Taboola. That's why Optimized traffic allocation ends up favouring Ads with the best CTR, but which may not have the best ROAS as in @iwanttofly's example.

Therefore isn't the best approach to start a new campaign with 5-6 identical copies of the same winning ad (with different hashed images), then let Smartbid work with the ad that is going to give the best overall ROAS?

---

This is great news! A rep can definitely help you get over the default limitations and filters (as I was saying earlier)

This one is definitely a big PLUS for Taboola. Most people get bored with slow moderation and responses from the main native networks up to a certain point where they can even drop them and move on with some other networks. I just had a duplicated campaign pending for over a week on OB, that's crazy!

I guess it's time to deploy the big guns and have machines deal with machines from now on.


Let me try to answer your questions below:

You can definitely setup listicles on Landerlab. You can even start with a ready to use template for that, simply swap logos, colors and stuff to make it more personalized.



Regarding the page swap on redirect, are you maybe looking for something that after the click event sends the visitor to another page while the offer page gets opened on a new tab? If that's the case it's a script related thing, but I'm not sure if the traffic source will be happy about it. I've seen such techniques around, but never took the trouble to test it.

I would wait for RedTrack support reply about the page view pixel. If Taboola's pixel is firing but Redtrack pageview not, than I would say its a script related issue. Anyways let's see what Redtrack has to say about it.
Just keep present that when using Landerlab its like if you were to host your landers on your own static servers, other than that we've added additional configuration related settings to make sure the page loads as fast as it can.

You're in a good position now and in a few days you should be able to get your campaign into the green zone. Just focus on your campaign stats and let them drive your decision. It's really hard to make mistakes when relying on your campaigns data for optimization actions.

---

Thank you again @platinum.

As to the listicle, I've seen that behavior somewhere, I just can't remember where. Facebook, Google or Native, because I've encountered it on live traffic. I'll look into having someone write the script later. And yes, I'm sure some won't like it. Fortunately with Taboola, there seems to be little downside with trying new ideas as long as you don't change the behavior of the lander after approval. Their big hang up seems to be anything that even hints at cloaking. Which is fine, the attraction is that it isn't Facebook and every mistake isn't fatal.

Yes, I'm still waiting on a response from them as to the LP View Script. I may have to email them again.

And thank you again for all your help. Access to the machines is definitely going to speed up the process!

---

Well, very much in the phase of getting enough data to optimize the placements while continuing to look for improvement with the ads.

A huge thank you to @platinum for all his help. When this cracks, it will very much be due to all his assistance.

---

Would anyone be willing to share how often they swap out creatives when testing? Any guidelines as far as spend or time before swapping in something new to see if it does better?

I'm torn between, am I giving it enough spend to perform or am I waiting to long to try something new.

---

If we have a good established creative getting the majority of the clicks, then we block new creatives we try at 1X cpa without a conversion, or perhaps 5x at a negative ROI (again, if the top ad is profitable so we have a benchmark to compare against).

In a brand new campaign its tougher because you don't want to just block all the new ads due to negative impact on algorithm. So I guess on a new campaign I would just try to quickly find my 'benchmark' ad and then start trying to beat it. Or, if you're close to breakeven, just let all the ads keep spending so you (and the algorithm) can get more data.

If you were at like -70% though I'd either keep trying new ads or I would just start a new campaign iteration with new ads.

There's another native ad expert named Joe Burton, and he says he NEVER uploads new ads to a profitable campaign. So the only way he tests new ads is by cloning the campaign and trying new ads in a new campaign iteration.

So yeah, lots of ways to do things. We upload new ones in the same campaign, but it can definitely derail things a bit if you get a higher ctr ad with a lower cvr, because the algo will want to send it more clicks due to the higher ctr, but it has a lower cvr so then your epc goes down.

If you're worried about it or trying to find the right system, just lean heavy on the 'clone' button. You can clone a bunch of iterations of your campaign and try a different ad strategy in each one Usually almost all our big volume winning campaigns are actually a combination of 2-6 sub-campaigns that we've cloned from the original, so its definitely a good tool to utilize in general.

---

So not a whole lot of update yet. Mainly battling moderation at this point. They have decided to change the rules on what is and is not allowed. So even duplicating active campaigns are getting rejected. What was approved yesterday is now denied. Very disappointing.

Also, hard to believe it has only been 2 weeks. On one hand it feels like started yesterday, on the other it feels a lot longer. I guess it is good to look at the calendar periodically and get some perspective.

One thing I am noticing from Optimizer, the same widgets are getting turned off in each campaign. I'm guessing this is normal and I should probably just put them on a blacklist to add to each campaign going forward? Maybe periodically test them on a performing one, but otherwise just ignore? I kind of hate this because they are huge traffic placements, so I hate to miss the traffic. But they are also sucking up spend and are a huge impediment to getting profitable.

---

Has anyone used Interest Targeting or Retargeting via the Taboola pixel? If so, could you please share your experience?

My rep was strongly suggesting retargeting, and that does look interesting, but the audience isn't very large so far. However, I do see the interests and hopefully someone can share their experience in using interest targeting.

---

Looking at my campaigns, it seems one of my biggest issues is the low ad CTR. I am looking at the spy tools and what I find with manual spying, but I'm hoping some of the experts can chime in with what works well. What tends to drive CTR, the image or the title?

---

Still here.

Getting conversions all along, but nothing profitable yet. The two biggest issues have been vCTR from the ads and LP CTR. To that end, I found and ripped a lander used by one of the bigger lead generators that is rather aggressive, and also used some manual spying for inspiration for the ads.

Right now, just waiting on them to get approved.

---

Unluckily I can´t really help with Native.

---

I appreciate it. @platinum has been very helpful. Also, some nuggets by @jack_l in some of the threads as well.

I feel like I'm going to need a LP CTR around 20% to have a chance. Fortunately the offer converts very well, over 15% and the average for Native is around 20%, so I'm sure improvements in vCTR and LP CTR will help there too.

---

Are you still running auto insurance on Taboola?

You asked about retargeting up above if I saw it correctly - that your rep suggested it - I personally have never been able to get much success retargeting on native... a couple times when I was doing listicles with huge volume I managed to get some profitable retargeting campaigns at low spend... but for the most part I've never had much success with it... then again I don't do much ecom though...

I believe in the James Van Elswyck course he says the same thing... however I think he also encourages Facebook marketers to get started in native via placing a Taboola/Outbrain retargeting pixel on their offer pages and then retargeting those facebook visitors on native... so might be worth a try.

At the same time, I imagine you'd only get like 1-5% of the traffic your getting from the main campaign.

I think 'Try retargeting' may just be one of those rote 'optimizations' they train the reps to give us.

I'd be curious to get your opinion too @platinum - as its always something I've wondered about, i.e. whether these big native outfits are getting much of their volume from retargeting, or doing everything on the front end.

---

Yeah, I built it to play along, but it basically got no traffic probably wasn't worth the time spent on it.

Right now I'm focusing on the front end, ads and lander. I'm also probably going to implement an offer wall soon to show people after they click on the main offer. Hopefully pick up a few conversions that way.

---

Didn't like the "Try retargeting" suggestion from the rep neither, and honestly hate it when they give such suggestions. This tells me right away that they either don't know what they are talking about (using internals scripts), or even worse they didn't bother taking a look at the account in order to understand if there's enough data to back up their suggestion.

I haven't thoroughly tested retargeting, as all the times I did run some tests it ended up getting little to no traffic at all. Retargeting non-native traffic on native (end vice-versa) might end up yielding better results, however the volume needs to be significant, otherwise, there would be not much to retarget. I'd think it would be worth retargeting even landing page views or some other custom events. This way it is possible to clean out fat finger clicks and focus on those visitors who really expressed some interest on the lander.

Native networks aim at a good avg. ad ctr on high volume audiences. Limiting targeting on a small audience (a few thousands) caps the retargeting campaign pretty hard, giving it no chance to deliver enough traffic to satisfy the advertiser, publisher and the network at the same time.

I've done some little tests with lookalikes too and still not convinced about them. There are people who have seen great success with them, but I personally haven't. Most probably it has to do with ad spend levels too.

If I'm getting this right, are you planning to send visitors to another lander after they click on the main offer? Main offer opens on a new tab, then original lander redirects to another lander?

---

Correct, that is my plan. The volume and LP CTR needs to be high enough to make it worth implementing, but yes.

---

I believe it would be best to show the offer wall under the same landing page rather then redirecting the visitor to another one. At least this way you will avoid having compliance issues with the traffic source.

It can be a little tricky to be achieved (requiring some technical skills), but at least won't get you into trouble with compliance.

---

Absolutely love the FA man, can't wait to see where ya end up in a few months!

---

Thank you. You are too kind. My FA is no where near as detailed or updated as frequently as yours or @jaybot 's.

---

Awww.

But get this:

If you never started a Follow Along, I would have never started Native

Your FA has more value to me than mine does.

---

Strange way to ask for an apology for getting you into Native...

Thank you though. I'm glad you found it helpful.

---

Been a while, but yesterday was a good day.



I am also getting a lot of "extra" clicks from Taboola. I've asked both Taboola and Voluum, so far no ideas as to why. And when I say a lot, I mean multiples of what I'm actually paying for. At this point I'm really just having to look at Taboola to optimize.

---

w00000t!

Congrats on getting back to green!

I swear one of us is going to make solar work someday

Taboola's inflated clicks are annoying, but basically as you said, just have to use the extremely delayed stats on their dashboard to try and manage it while updating costs in your tracker.

Taboola's volume and quality is worth it tho.

---

Home > Native >