Home > Paid Traffic Sources > Native

---

Better to focus and delete BAD sites on NATIVE (5)

---

Hi,

Newbie native here

So recently I started native with a small budget. Getting some conversions but not profitable yet.

I am digging into data everyday to see what can be improved. I see a lot of people talking about "blocking bad sites" etc to get better results & improve ROI.


I am guessing this simply means that start blocking the sites that are not giving better conversions or may be CTR/CPC etc. Is that correct?

1) If yes, what exact things do you guys look at? Is it CPM or CTR or CPC?
2) How do you guys analyze "bad sites" with lower budgets? I apparently see one site giving the best conversions. So far I had 7 conversions out of which 5 are from one site only. I see a total list of 70-80 sites.
3) When is the right time to start blocking? Any benchmarks with spends or time?


Any other tips/tricks to improve ROI?

TIA.

---

Hey,

Since you're new to native I'd just stress that blocking bad sites isn't likely to move the meter much for ROI. The real big levers are good offers and ads, then landers. Easier too if you adjust global bids than micro-managing bids on a bunch of sites, often which are spending very little per day individually.

And if you're having to block a bunch of sites then the offer or ads more than likely aren't strong enough. It's easy to fall into a cut everything mindset when starting out, but you should have better results just testing more offers, geos, and ads honestly.

If sometimes not working on a larger more premium site then I'd drop the bids there so that your effective cost per click (eCPC) to your offer is close to your overall expected offer EPC from lander. Usually people recommend something like 2-3x CPA per site before blocking, but you can sometimes make a call using eCPC sooner.

Just be mindful that some sites will have lower cpcs than others and if you're going off your tracker's averaged stats you can be mislead at times. Tend to use traffic source's reporting with conversion tracking more often than my tracker for site performance.

Also try not to get too attached to making an offer work in the beginning. Ideally test 3-5 and focus on one which does best. I wouldn't recommend testing unproven stuff your affiliate manager sends your way until you have a good grasp of that vertical and/or know that advertiser is solid.

---

I am going to share a little secret with you, one I have found in the last month of running Native for the first time. Note that I am an extremely proficient FB marketer but this is my first time dabbling in native and I have learned alot of things.

So I think the secret to native is a triangle consisting of good CTR, low CPC and a good offer.

When I started with native, I asked my rep which offer was performing. I was very careful to select an offer which was performing excellently for other people ON NATIVE and ON THE AD NETWORKS I WAS ON. So I have 100000% proof that the offer works, if I can't get it to work, then it's just me.

So one piece of the puzzle is gone. Which leaves CPC and CTR.

CTR is 100% ad based. Again, I was able to more or less use the same *types* of ads my competitors were using. This particular ad network does not allow ripping, so I had to create my own ads but they were similar to my competitors. So #2 piece of the puzzle gone.

Which leaves us with CPC. Let's talk a bit about that.

#1 . Don't trust Adplexity for determining your site placements. I tracked my own campaigns on Adplexity and guess what....most of the placement data was WRONG. Just plain wrong. So don't trust it completely only use it as a guide.

So what's my big secret as to Native? My secret is that the guys who know how to run campaigns on native really are not sharing the vital information you need to succeed. Not James Van Elswick, not any of these guys. They don't talk much about what it takes to make a campaign profitable and I am convinced that CPC management is the main thing they are withholding.

When to kill widgets. When to keep them going. When to bid higher, when to bid lower. This is among the most protected info in Native because it is the very difference between profitability and losing money.

I am trying to develop my own rules for CPC management but as of yet I am still working the process.

To recap:

- Easy to get a good offer.
- Easy to do good ads that mirror what successful people are doing.
- hard to mange spend so you don't lose your ass.

Maybe others can chime in here.

---

The question of blocking widgets depends immensely upon which network you are running on.

For Taboola SmartBid campaigns I agree with TheDude totally - there should be virtually no blocking of sites apart from a few random outliers that have 'push' or 'propeller' in their name maybe, but for the most part the SmartBid will do the work and you just want to focus on finding really good ads, and of course a converting offer, and getting clicks cheaper than the EPC of the funnel.

On Revcontent though, if you are running a Topic campaign, you will definitely need to block some sites (at least in my opinion). They've cleaned up most of the 'bot widgets' so that's not the problem, but some sites just won't be profitable.

On Revc it seems to me the secret is finding those really great widgets and then re-using them all the time.

I've run lots of stuff on Revcontent the last 6 months and no matter what vertical it's in, there's 6-12 incredible sites I just use over and over again because they convert like crazy.

Then there's other sites that don't convert that well.

Example is Breitbart vs ConservativeInstitute.org. You would think they would have virtually identical traffic right? Similar focus, similar widget placement, similar audience, etc.

But Breitbart is not very good at all, and I haven't had it profitable in a whitelist campaign in almost a year... no matter what vertical I'm running...

ConservativeInstitute.org, on the other hand, kills it every time... does great in every vertical I try...

MGID is a network I don't currently run on, but my understanding is you have to block like crazy on it.

Outbrain is kind of in the middle between Revc and Taboola. If you are running normal (non algo-optimized) campaigns on Outbrain there's usually some blocking, and certainly some adjusting of coefficients, but most of your clicks will always come from good quality sites like MSN, CNN, etc, so if you're having to block a lot of sites its usually a sign the funnel just isn't working.

I agree with you that cpc's and ad ctr's are the 'magic sauce' to a great extent. I don't think there's neccesarily any huge secrets about them people aren't sharing though, more some people just have really good teams that create awesome ads and try tons of ads out and use multiple campaign iterations for testing ads vs scaling, and all sorts of stuff like that.

It seems like finding a profitable campaign is the hardest part, but actually I think maintaining the cpc's on profitable campaigns is just as challenging, especially on Outbrain (Taboola and Revc not as bad in that regard).

Anyway though to answer your question, most folks on Revcontent who block sites (either manually or with TheOptimizer) do so based on 2-3 rules, mostly blocking those with extremely low landing page ctr's (say under 5% after 50-100 clicks if the normal lp ctr is 40%), and blocking widgets with no conversions after 1X offer payout spend on them (or 2X if you're more conservative, or .5X if you're super aggressive, etc).

---

Awesome advice here. I have a lot more clarity (or reminders) on certain aspects of native. Thanks for everyone who commented. (And thanks @psychicboy for the question)

---

Home > Paid Traffic Sources > Native