Home > Paid Traffic Sources > Native

---

The Ultimate Automation Rules For Native [TheOptimizer[ (13)

---

Hello ya'll.

Have you ever wanted to earn 8 figures daily without even touching a button? It's impossible, but at least we can try to minimize the time we're sitting daily, breaking our backs, heads, and something/someone else

I wanted to discuss the best automation optimization rules for Native (using TheOptimizer, and I'm not sponsored). And let me know what do you think is missing from mine for the ultimate automatic-optimization game

For search feeds - here are my rules (based on sites for Taboola or Sections on Outbrain):

Basic click & conversion rules:
1) 10/15 clicks & 0 conversions = Block. (10/15 depending on how low the average CVR for the particular vertical is)

Blocking Sites/Sections based on ROI:
2) 3x payout spent & ROI < -50% & 1x payout Spent = Block. (Last 30 days excl. today)
3) CVR < 5% & ROI < -40% = Block. (Last 30 days excl. today) - I don't know the point of this one, I just "stole" it from someone smart like @platinum

Unblocking Sites/Sections:
4) Unblock Sites with ROI > 0% (last 14 days)
5) Unblock Sites with ROI > 0% (last 30 days) - felt like it was important to give this rule both timeframes.

Change bid to 60% EPC:
6) ROI is between -50% and 0% & 3x payout spent = change bid to 60% of EPC (last 9 days, excl. today)
7) ROI is between -50% and 0% & 3x payout spent = change bid to 60% of EPC (last 5 days, excl. today) - Again, I felt like it was important to give it data from multiple timeframes.

The reason I have set is to 60%, is because Taboola can sometimes bid up on a widget like +150% as you know and then even 70% or 80% of EPC won't be profitable. Although sometimes this backfires and the resulting bid is too low.
I wish there was a way to deal with this, for example using API and switching automatically at that point of a bid change to a "fixed bid".

As you may have noticed:
I don't use any rules for campaign's stats (overall) and no rules for creatives.
I also don't have rules for bidding up on profitable widgets as I find it too complex to set up (for example, I would like to bid up 15% on a widget above 50% ROI but even if I were to set it to run once a day I can't really account for the resulting changes that quickly and it will end up jus continuing bidding up 15% until I'm not profitable).
You may also notice I absolutely don't have any rules for "publishers" on Outbrain (only for sections - which is just a copy of the rules for Taboola).

Also - for giggles, here's an interesting question, given there were absolutely no constraints of TheOptimizer and you could come up with literally every rule you had in mind - which rules will you have? (although realistic ones lol based on real data you get). And a dream for me personally, broadly speaking, would be to set different conditions inside a rule based on different timeframes, and to also have an option to exclude a rule from executing again on a certain placement if it already executed on it "x amount of time" back.

Geekout time!

---

@roiter123 - very cool man!

How is the search arb going for you on native? I never tried using TheOptimizer for it... perhaps I should have done that... that's very smart... especially the the blocking sections rule at 0 conversions after 10/15 clicks...

If I could "magically' come up with any rule for search arb + optimizer it would be to reduce bids and budgets based on epc fluctuation patterns (times of day, times of week, times of month, etc)...with that said though is difficult to predict what direction those fluctuations are going... I tried finding historical patterns but never had much luck...

I think if one could really figure out how to have a close pulse on the backend google search bidding that would be a huge advantage... but I'm not sure how to do that... but if you could see in real time how much people were bidding on your keywords it would help immeasurably on ensuring profitability... like if you had an agency that focused on both google search AND google search arb on native, and were an expert on both, man... you could do some damage...

---

LOL You really got me going there for a second! Thanks for the chuckle!

And thank you so much for sharing your optimizer rules! Makes me want to run native again just to give them a whirl. I'm sure they're help out some members!



Amy

---

Thanks a lot bro.

---

Thank you man!

It's going shit That's why I turned to STM for help
All jokes aside, it's a tough game as you know, but I was able to get small profits going on Native.

That's a good point about fluctuations, it's pretty hard to find the patterns for me so far. The best I'm doing in that regard is watching the same day temporary stats' RPC and try to make some decisions based on that.

What are your rules looking like for Native my man?

---

My rules are simple for lead gen and ecom:

If TS clicks 30+ and LP CTR < 5% = block widget (for Taboola revcontent and Outbrain, for MGID and other shit-tier Native 10+ TS clicks)

If widget cost = 1x payout and no conversions: block widget

That’s like 90% of my rules. Then some extra:

If widget costs = 2x payout and not > 2x revenue, block widget

If widget is profitable and has 2+ conversions, set EPC to 70-80% CPC (on rev content, doesn’t work so well on Taboola)

If widget EPC is < CPC over 7 days, drop bid by 50%.

If camp is profitable today, raise budget by x% AND reset budget every morning back to $100 a day or whatever

Then a bunch of dayparting because I am a pussy and don’t know how to run 24/7 on Native.

---

Oh that's some good rules man!

Seems a bit rough, like maybe - If ROI is less than 0% Last 7 days set bid to 50% of EPC would be better.

This is very interesting and I wonder whether you can do something similar for bidding.

By profitable today do you mean like > 50% ROI or higher?
And how do you reset the budget inside Optimizer?

---

Go to rules, create new rule, check out the templates

For resetting budget, I just check every hour and have it fire at 3AM, pretty much the same as the template.

You can do the same with bids, but bids are usually by % and can get messy. I’ve had better luck just running two camps for day/night shifts at different bids. Not the most optimized, but seems to work for me.

---

Thanks man!

---

Sorry I missed this thread, had to take a small break from those huge hours behind my screens.

@roiter123 considering we have a direct integration with all major search arb players like System1, Tonic, Sedo and DomainActive I would do the following:

1. Integrate Search Feed providers directly to TheOptimizer Native (even if you're using a click tracker).
2. Add an Automatic Update schedule for the Last 3 days on each traffic source - this way Search Feed provider stats will be updated back again when revenue is confirmed.
3. Create an Est. ROI custom metric - this way you can later use it for ROI based rules that consider intraday revenue reporting


4. Add ROI based rules using both intraday and confirmed revenue. (note that intraday revenue is usually an average of the last 3 days on most providers)

A few example rules:

Example rule 1 - Increase & Reset daily budget when profitable


Example rule 2 - Block Low ROI OB Publishers


Note: Although many publishers have tens of sections, it is always good to have such a rule in place to get rid of those singe section or 3-4 section publishers that just waste ad spend. The propose here is to leverage publisher blocking limits in order to gain room on section blocking.

Example rule 3 - Use widget/publisher or section level payout to block




Explanation: Here I've used a custom calculated metric to dynamically calculate the payout on the widget/publisher, section, content level. This will help covering up the differences of the avg payout from one publisher to another (usually the publisher influences the payout). For a more generic approach, another variation of this rule would be to use a % of campaign payout.

Example rule 4 - Update Section Bids 2x Week for when Section has 20%+ ROI



Explanation: Here I'm touching bids only on sections that have spent 5x the avg. campaign payout have an avg lower cpc compared to that of the campaign level and have 20%+ ROI. The bids are updated 2x a week by 22pm CET (which is pretty much after the standard confirmation time of the previous day data) based on the last 7 days of data excluding only today.

Obviously there are many other potential variations of the above rules that can be implemented to address specific patterns. The important thing is to follow along with the traffic source algo instead of fighting it.


I'd be more than happy to post rule screenshots based of specific scenarios you might want to address if needed.

---

So glad you're taking the time to take care of yourself @platinum (Losid)! Really!

Is it accurately reporting directly to the API as much as using S2S postbacks to the tracker? Also because, I heard from your support (I think Erwin) that if we're passing S2S postbacks (&revenue) to the tracker and ALSO passing API to the Publisher in Optimizer, it would count the Total conversions and revenue twice in TheOptimizer (which is not desirable for stats in the Optimizer).
And then, if I don't report to the tracker, what if I want to break down certain data (for example: Revenue per device) in the tracker?

With DomainActive I noticed that this could work, as the estimated revenue is not that far off from finalized revenue there.
However, with TONIC: I noticed finalized revenue can jump 30% up or 50% down, or for LATAM countries they are reporting 0.01 RPCs until revenue is final (RPC is way higher in final)
How do you deal with this for big networks like TONIC?

Would you use rules for the sections in Outbrain as a copy of what I'm using for the sites in Taboola?

Sometimes certain widgets/publisher/sections will report a completely different payout, and for me personally, I remember recently with the support where trying such a condition in the rule gave me some trouble as rules were executing and adjusting bids for widgets that spent really low because their payouts were low (that was finalized revenue only). Have you tested such rule?

Personally, I'm running such a rule daily, what do you think about that?
And also, re: 20% ROI, this is a bit problematic with Taboola that adjusts the bid (sometimes up +100% as you know, and even higher on Outbrain), this is why I have it set to 65% of EPC (still wouldn't be perfect in all cases), is there some way to set the bid of that placement to fixed somehow?

Thank you so much for your efforts and sorry for so much questions

---

You are right, that was a detail I forgot about. But there is a workaround to it

To have revenue from both Voluum and DA, you can pass S2S postback revenue on custom conversion event (e.g. Confirmed Rev) and exclude this conversion from being reported in the main conversions. This way you can still have the revenue reported back to Voluum, at the same time from DA directly.

Regarding Tonic your observation is correct, I've seen such fluctuation in LATAM generally, not on other countries. In any case having stats from both the feed provider and tracker should give you more insights.


Yes! Just keep present the difference between Taboola SmartBid and Oubtrain section level bidding.


That's the point of calculating the payout on the widget/section level. Considering the huge fluctuation between one publisher and another, or sometimes even within same site different widget ids, it is more fair to calculate the payout on the widget level, rather than do it on the campaign level.

I like changing bids less frequently to be honest. It might be more beneficial to change the bids more frequently, but I feel like changing them too often will somehow go against the traffic source algo. Less frequent changes will be kinda smoother IMO.

---

is there a way to turn off all widgets, except the winning widgets, then increase those bids, then slowly open up widgets base off CTR?

---

Home > Paid Traffic Sources > Native