Home > Paid Traffic Sources > Native

---

Taboola: How to Scale? (5)

---

Hi guys,

I have a low budget ($100/day) 'Smart Bid' campaign that seems promising and I'd like to try to scale it, possibly without messing it up


The current plan is pretty simple and I wanted to check it out with you first: increasing the budget by about 50% every day or so (not to be too aggressive on Taboola's optimizing algo) and see how far I can push it with the current bid (and if ROI allow it slowly increase the base bid as well to squeeze some more volume).


At the same I thought I should probably keep adding ads to the campaign, however, I'm not sure about a few things here (currently the campaign has 6 ads, different combinations of image/copy):
- how many ads makes sense to run/test at the same time (and so how many ads to add daily/weekly)
- negative ROI ads of course we're stopping them, but what about the ads that do positive ROI but under-performs compared to the top ones? Should we pause them or let them run?


Lastly, in general, is it a good idea to touch manually the Sites of a 'smart bid' campaign (either by increasing or lowering their bids or completely pausing them)? I'm asking this as that's something the smart bid algo should already do on its own, but I thought that perhaps it might still be necessary (or optimal) to intervene manually from time to time.

Thanks a lot in advance!

---

Hey,
Keep in mind that taboola's algo is not that sensitive as fb when increasing budget. If you are doing well, you can increase the budget as much as you like.

negative ROI ads of course we're stopping them, but what about the ads that do positive ROI but under-performs compared to the top ones? Should we pause them or let them run? -keep. pausing too much will kill your campaign, you have to keep ads to rotate.

Lastly, in general, is it a good idea to touch manually the Sites of a 'smart bid' campaign (either by increasing or lowering their bids or completely pausing them)? i usually recommend doing fixed bid if you have a smaller wl and adjust the cpcs accordingly.
Hope this helps (ii'm doing native ads for 6 years now

---

I usually pause ads if 1) they aren't profitable, or 2) they aren't getting spend. I'll let there be up to 8 profitable ones at once although that is extremely rare, usually 1-4. I like to upload tons of new ads as I think that's the biggest "secret sauce" for success, however other smart folks say to never add new ads to a profitable campaign and only test them via cloning.

Re: sites I don't like to touch them on SmartBid but if one goes 0/100$ and the campaign isn't profitable I will sometimes... all my most profitable campaigns are ones where I never even look at the site performance though. I just treat it like facebook and assume the SmartBid will either work or not work and there's nothing I can do about it as far as site-optimization goes.

For scaling you can either a) increase budget, b) target new devices and/or geo's, or c) clone replicas. I generally do a mix of all three. Especially once you get into $xxxx+ per day spend I find it much better to use a number of clones. Typically reduces performance volatility, and you can test new ads with less risk that one spends a lot with zero conversions. Also individual campaigns will get finicky all the time so better to have a bunch of them and continually pause the bad ones and start new ones then just have one great big one with everything riding on it.

As always though there's a million ways to profit on Taboola and other folks do things 180 degrees differently!

---

As @nativeq said, in general native ad networks are not that sensitive to daily budget increases. You can easily double the daily budget without putting your campaign in "learn mode" like facebook does.

If you are running on SmartBid, Taboola will take care itself on bidding up or down whenever needed. However, since you can still increase (or decrease) site level bids when running on SmartBid, you still have room for improvements if specific sites are giving you excellent results. It's like telling its system that you're willing to pay even a bit more for that site's traffic. Just keep present the limits of bid increase/decrease when running on SmartBid vs. Fixed.



When it comes to adding new creatives, you need to make sure that you are not rotating more than one landing page until the new creatives get approved. Not sure why, but Taboola doesn't seem to like rotating landing pages when reviewing ads. So if you are rotating two or more landers, keep present that the approval time for the new creatives might be a bit longer compared to when you submit a new campaign, so you may need to pause the landing page split testing until new ads get approved.

Regarding scaling you have quite a few options to consider:
- Increase your profitable campaigns daily budget.
- Create a clone of the same campaign targeting new devices (e.g. started with Desktop, then duplicate the campaign on mobile).
- Copy the campaign to other ad networks. You can maybe copy your current Taboola campaign to RevContent using some slightly more aggressive creatives.
- Launch a new campaign with similar offer using your previous campaign data (blocked sites, bids, creatives, landing pages).

---

Thanks everyone for sharing your thoughts and expertise!

Interesting, I'm seeing across many campaigns that most of the Sites just doesn't work (for instance, super low LP CTR where others are steadily at 30%+) and Taboola keeps spending happily on those (of course I'm aware they doesn't have the LP CTR stat so they need more data in order to make decisions). This is making me think that the optimal thing to do would be just block these Sites (or lower their bid heavily) as soon as I realize they'd not make money, and to rely on the Smart Bid just to actually manage the bid of the Sites that I don't block. Not sure though

---

Home > Paid Traffic Sources > Native