- Affiliate Programs
- Affiliate Networks
- Ad Networks
- Articles
- Tools
- Social Media Accounts
- Events
Affiliate marketing is a great source of income, especially when you know how to test effectively and scale strategically.
In this case study, conducted together with our partners at RollerAds, a global ad network, we explore which bidding model delivers the best results during early campaign tests: the traditional CPC or the more advanced SmartCPC.
Let’s briefly review both models before diving into the case study findings.
Let’s make sure we’re on the same foot here and recap the bidding models. CPC (Cost-per-Click) is a model where you set a bid, let’s say, $1, and you pay exactly that amount for each click. A fixed bid is good when you have data on a particular source, but if you explore the zone, you enter the guess-game.
This is where SmartCPC comes into play, saving you time and money by automatically adjusting your bids according to the current prices. That’s right, with SmartCPC (and Smart CPM), you might bid at $0.5, $0.7, $0.9, saving on costs. At the same time, the funds saved can be allocated to more expensive sources. Basically, SmartCPC is an automatic optimization for your budget spending.
After running tests with this model and collecting whitelists of zones and feeds, you can confidently start a more precise campaign with CPC and maximize your earnings. Make sure to read the RollerAds article about Smart models, and don’t forget to register on their platform if you’d like to try out the model. If you’re still not sure, read on.
As we mentioned earlier, in this case study, the RollerAds team conducted two initial tests and compared how CPC and SmartCPC performed in real-life scenarios. The offer was provided by their partners, CpaRoll; find all the key details below.
For starters, RollerAds media buyers began with CPC, and chose the Direct Click ad format. Direct Click is their innovative ad format, involving strategically placed links that are open upon clicking. Such places include “Exit” or “Enter” buttons, and some other high-traffic areas. Unlike Pop ads, Direct Click is CPC based.
Then, there was the usual stuff: campaign name, URL, specified in the offer GEO, and proxy traffic was disabled. For bids, the limit was set at $0.05. Considering the relatively humble payout, the daily budget limit was $20. As for the total budget, it was uncapped because no one knew how long it would take to gather enough data for the performance evaluation.
When OS targets were selected based on the offer’s KPIs (Key Performance Indicators), the campaign’s setup was completed, and it was sent for moderation.
After gathering some data, the media buyers began optimizing their CPC campaign. On the first day, they excluded red zones with low ROI.
The same process went on during the second and third days.
After three days of testing, despite all the effort, there were no promising zones, and the ROI ended up being -66.64%. It was time to move on, so the RollerAds team accepted the $37 loss and switched off the campaign.
It was time to try out SmartCPC as an alternative approach to see if it fits the testing stage indeed.
For the experiment's accuracy, besides switching to SmartCPC, the other campaign details, like offer, GEO, ad format, and so on, stayed the same.
When working with this bidding model, remember that there is the bid control slider. It helps you to control the number of impressions by setting the bid higher or lower than the average auction price. For starters, it’s better to leave the slider as is, i.e., in the middle.
Another distinctive feature of SmartCPC is that you set a maximum amount you’re willing to pay for a click, rather than a fixed bid, so all the traffic you get will cost you no more than the specified amount per user. For the case study, the RollerAds team set the average auction price without capping the bid to avoid limiting the prospect of more expensive, high-quality zones.
Another thing that remained unchanged from CPC testing is the daily budget limit—$20. The total budget was also uncapped to allow flexibility and time to collect all the necessary data.
After the preparations, the campaign was submitted for moderation.
After the first day, the zones that didn’t generate any conversions were excluded.
For the second and third days, it was rinsed and repeated.
And zones with very low ROI were also excluded.
Finally, on the fourth day, the results of the SmartCPC campaign were summarized.
Despite the still negative ROI of -16.70%, the SmartCPC campaign performed significantly better than the CPC one (-66.64%). What’s important here is that the RollerAds team was finally able to create a whitelist of converting zones and feeds. With these results, the team could continue the research with more precise fixed bids, which wasn’t an option with the CPC campaign.
Creating a whitelist like this is easy with any tracker; you just need to sort everything based on the number of conversions. To ensure the success of the future campaign, zones with low ROI were excluded.
The same goes for feeds.
And another list was created for the OS versions that delivered positive results.
After gathering the necessary data on performing zones, feeds, bids, and OSes, RollerAds launched the more precise campaign on CPC.
Keep in mind that you absolutely don’t have to start a CPC campaign after the first test is successful. If you wish, you can continue optimizing the original SmartCPC campaign by refining its elements and, if necessary, adjusting the bid slider to increase competitiveness for traffic. However, it is better to switch to the CPC bidding model if, by the end of the tests, your whitelist is small, and you want to increase its potential.
Once again, because RollerAds was promoting the same offer, the details from the CPC and SmartCPC campaigns also remained the same. The only new element in this campaign were the custom bids from the SmartCPC test. It was added to help build optimal bids for some zones and feeds.
Speaking of whitelists, RollerAds added one based on the results from the SmartCPC campaign.
And who can forget to add the most performing OS versions?
As per usual, after the campaign was prepared, it was submitted for moderation.
Every campaign needs daily optimization, so RollerAds refined and updated custom bids for zones that generated conversions but weren’t yet delivering a positive ROI.
The custom bid for well-performing zones was increased to get more traffic.
And the RollerAds team removed zones and feeds from the whitelist that weren’t showing positive results daily.
After all the optimization, the campaign started generating profit on the fourth day.
The experiment started with the CPC test, which resulted in a negative ROI of -66.64%, $37 loss, and no hint of promising zones and feeds. This was a dead-end and a sign to stop the campaign for good.
On the other hand, the SmartCPC campaign resulted in a still negative but promising -16.70% ROI.
Even though the team lost $6.40, they obtained something much more valuable in the long run — the whitelist of zones and feeds, which helped them launch a campaign with a +11% ROI.
This case study suggests that SmartCPC is a superior choice over CPC for testing for two reasons: fewer losses and more chances to find promising sources. True, neither campaign brought profit immediately, but thanks to the sources from SmartCPC, things were straightened up with the following CPC campaign.
In the end, we want to say that CPC and SmartCPC aren’t rivals. The key to success is to treat them as complementary tools: work smart when testing new offers and traffic, and then, if you want, go precise with the traditional method.
Another key to success is working with RollerAds, the high-performance advertising network. More than 50k advertisers already joined the family and significantly boosted their revenue.
And if you don’t know whether you should be a part of the team or not, RollerAds created a promo: use the code Partnerkin50 before December 31, 2025, and get an additional $50 on your first deposit. Register on their platform and tell the code when chatting with their support team or your manager assigned.
