Home > Paid Traffic Sources > Native

---

Bids on Native Ad Networks? (6)

---

So I was just getting started with Revcontent and Taboola and have some questions as to the optimal bids & budgets. This is assuming someone starts with a fixed/manual bid instead of smartbids and targeting US as the geo:

1) What bid is recommended to start with at these 2 networks to get enough high quality traffic to get enough conversions and make further decisions?

2) Should an offer payout be considered when choosing a bid? If yes what's a good bid for leadgen ($5-$20 offer payouts) vs a CPS/VSL offer ($30 payout and up)?

3) Though I know that even search arbitrage is being run successfully still wanted to ask, should lower payout leadgen offers say $7 and below even be run on native ads?

4) Also what's a general overview of bids on these networks? What's the bid required to get premium placements like msn and others which generally result in high EPCs/offer conversion rates?

5) What approx. budget ranges should one consider to successfully test and get an offer profitable, which is already running successfully on a network, for a leadgen vs a CPS/VSL offer?

6) What does 'Avg. Position' really mean on Revcontent? What's a good avg. position for content/creatives to target?

7) Any other proven advice to make the most out of setting bids and budgets?

---

Hey @regjoe - I'll chime in and hopefully @platinum and some other folks can also share their thoughts

1. It depends on the geo and device and the expected ctr... On Revcontent you can start low and move up if you're worried about overpaying... on Taboola and Outbrain better to start a bit high (especially Outbrain). The highest I've ever paid is 1.40$ per click for a few Revcontent widgets on US Desktop Life Insurance, and I've heard of people paying $3.00 per click for some Taboola pubs on US Desktop Solar in really tapped out areas. The lowest I've paid for US is 30 cents on desktop or 20 cents on mobile for super high ctr offers. I'd say a starting bid in the US of 50 cents for most stuff is good, and you can move down or up from there. If a low-ctr vertical like life or solar though maybe closer to 80... the top teir 2 geo's are moderately less expensive... all the way down to a penny a click in India or Indonesia on Taboola.

2. No, they aren't connected, however, the expected ctr is related. So if the ad says "Insane Opportunity You Must See Now!" you could bid less because it will likely have a high ctr, whereas if it says "Life Insurance For Seniors Age 60-70 - Get Your Quote Now" you'd have to bid quite a bit more because the ctr will be really low.

3. It all just depends on the conversion rate... if the 7$ offer campaign has an overall conversion rate of 10% or more it would probably work well, since you could pay up to 70 cents a click on it. If its a 1% conversion rate it would be impossible to make work though unless its a really cheap geo.

4. Usually the best pubs will require the highest bids... but again, its all relative based on ctr, device, geo, etc.

5. This is a great question and really depends... I don't like to spend more than 200$ without a conversion... and I don't like to lose more than a couple thousand even on an offer I really believe in... with that said, if I'm at -10% I'll be more aggressive since profitability is within sight. Whereas if I'm at -90% I'm gonna give up way faster.

6. Its just a rough description of where you're bidding relative to other folks. So if you have a high ctr and a high bid you'd be in the 1 or 2 position, and vice versa. I like to stay around position 3-4, which is the same thing James Van Elswyck says in his nothing but natives course. I haven't done much on Revcontent lately though so take that with a grain of salt.

7. The more campaigns you start the better you'll get at it. So just start tons of campaigns, even if you turn a lot of them off quickly, as the more you start the more adept you'll get at bidding/budgeting, and the more clarity you'll have on what geo's/devices/verticals require what kind of bid.

---

Interesting Jack, thanks for your inputs.

One more thing that just popped up in my mind. Apart from offer conversion rate/EPC, the headlines, CTA, story/message of advertorial and also the creatives title, when we talk specifically about publishers that consistently provide high EPCs/conversions, what common characteristics have you observed among them?

As in say maybe having bigger ad widget sizes, their primary source of visitors (search or social), senior or more affluent demographic for CPS offers or a certain political or religious wing?

---

Great question... I don't know... I guess in general Christian sites, Conservative News sites, Weather sites, and senior-oriented sites tend to do well, since they usually have more affluent people reading them.

And then 'Clickbait' (content-arb) sites are hit or miss. They can be great because the people on them are people who are bored/etc, so they can convert well, but they usually have so many ads and elements on the page you get accidental clicks.

I remember on Revcontent, RoutineJournal used to have two widgets - one above the 'Back/Next' buttons and one below- and the one below as amazing, super high epc and clicks in 30-40 cent range, but the one above was absolutely horrible (cpc's at like 12 cents and you still couldn't profit!) because it got tons of fat-finger clicks and the lp ctr was always like 1/4th the lp ctr on the bottom widget lol.

And then in my experience streaming sites are usually horrible, since the people are trying to stream movies or whatever and have zero interest in the ads. So anything with 'Manga' in the url name I usually block. Same thing with 'App' type traffic. It can be okay sometimes, but usually it seems pretty low quality.

Oh, and then local news sites are usually really good in my experience too.

---

Yeah that has been my guess too regarding arb. sites. The reason for those high CTR/low CPC visits and high bounce rate visits from those widgets above the 'Back/Next' buttons as you mentioned is fat-finger clicks on mobile and also on desktop when the page loads the ad widgets load last after all the content has loaded, the visitor often just wants to speedily skim through the slides to the one presented in the ad (they usually find it placed last or even sometimes nowhere in the whole slideshow lol), so in their hurry they often click on the ads above them which coincidentally manage to load just as they are about to click on the 'Next' button most of the time.

I too would assume that the more established/reputed news sites generally would have higher EPCs than sites visited by freebies/fun seekers.

---

I think @jack_l has already provided detailed answers to all of your questions. I'm adding a few personal observations (including a few weird tests I've been doing recently).

Both networks are very different from each other, both in terms of compliance rules as well as bidding or optimization approaches. Taboola tends to have more reputable sites and definitely the quality is there. Meanwhile, with RevContent the more you spend, the larger becomes the inventory you have access to.

For Taboola, I'll definitely suggest to start with SmartBid and if possible Even creative traffic allocation. This last one directly influences the testing ad spend, however it gives you more control on choosing the best performing ads for yourself. I've seen many cases where Taboola's algo penalized pretty great ads just because its algorithm thought the one it picked was the best one for the campaign.

Meanwhile, regarding RevContent I think that currently the account spend history influences the bid levels. If you're on a brand new account with no account managers assigned to it, you can even start with 0.45 in US, but as you start blocking pubs you're somehow be forced to slowly increase the bid. They sometimes will throttle traffic on some ads, giving priority to some others, but until the campaign has collected enough data, it's worth mentioning that other than blocking publishers or tweaking widget bids, it's best not to touch any other setting of the campaign, otherwise you'll barely get any click with the bids you started.

Again, I want to emphasize that the above regarding RC is what I experienced with two different new accounts I tested out of my personal curiosity and compare them to an aged account with significant spend history. Even the bid levels tend to be a lot different when comparing new vs. old RC accounts. I'm really curious to see what others have experienced if they took the trouble to run a similar test.

It depends a lot from the GEO, targeting details and conversion flow, but just as a reference 0.55 should a good point to start on either one of them.

Generally not, but in regards to the offer payout, its conversion rate is pretty important.

For an offer paying less than $10, I'd say you need ad least 15% CR with a good LP CTR in order to back up the spend. It has become a lot harder to get traffic on a ~ $0.25 Avg. CPC on native lately, so unless the LP CTR and Offer CR is high enough to back up the high CPC, sometimes it might not even be worth the time, money and effort to test low payout offers.

Always consider factors like: Conversion goal friction (click, personal information, payment details, etc...), Lander CTR (the copy and ease on navigation here makes wonders), and Conversion Rate (either click-to-conversion or lander-to-conversion)


Bid rates for premium placements are subject of the targeting details, ad image & headline quality, as well as a few other factors that have to do with the type of the offer one is running.

Premium placements have the tendency/ability to penalize spammy looking headlines and images, or even ban a low reputation offer from their ad placements. They care a lot about their reputation, and when you think about it, that is what has made them premium placements. Sometimes the image, headline and the lander/offer might be influencing factors to the bid levels and positions your ad is placed in their site.

On the other hand, with a low bid, you can be in a situation where you still get traffic from these premium placements, but the quality is not there at all. That can be easily noticed by the LP CTR. If you're getting an awful LP CTR from these placements, then it can be a good indicator that your bid is too low to get to their performing spots. Someone else, served on a better spot with a similar offer to yours might have already burned out the curiosity of the visitor.

It is pretty hard to give any suggestions here, because sometimes you push a $5 offer with a high CR and you can easily spend 20x to 40x its payout a day to quickly collect data and optimize faster. On the other hand if you promote a $100 payout offer you might be spending from 5x a day to 10x a day to test it. Generally the budget should be something that makes sense for the avg. bid, payout and conversion rate of the offer.

I haven't paid any particular attention to this, so cannot add any comments about.

Most native networks have switched to a Conversion-focused approach to determine bids. So to get a better idea of the optimal bid levels for your actual ads/lander/offer is the Avg. CPC you can notice on the publisher level, for when the network algorithm plays with the bids up and down to deliver conversions.

For example, when running with SmartBid on Taboola you will notice some placements that have the tendency to go way above your campaign level bid, and some others slightly lower.

Besides that, for networks that alter bids to deliver as much conversions as possible, sometimes its better to slowly lower bids on low ROI high converting placements, rather than blocking them straight away. This has to do with what their algo learned during the learning phase, and by totally blocking a high converting placement instead of lowering its bid can negatively impact the campaign overall.


There are literally a lot of factors to take in consideration, and all of the suggestions given in such discussions should be taken with a grain of salt. Your campaign data will be the most reliable source of decisions.

---

Home > Paid Traffic Sources > Native