Google and Facebook receive the most attention, advertising budget, and press. They’re also the most competitive. For anyone looking to build a lead gen funnel at scale, especially in high-intent verticals, they’re often the most direct route, and the most expensive if you’re not careful. If you’re costing out keywords for finance, insurance, utilities, or the holy grail of sweepstakes, you’ll quickly see how “other” starts to look incredibly appealing.
Popunder, redirect, and push notification traffic is seriously not for everyone. It’s the most dramatic funnel departure you can design but it’s not just as simple as waving your bids and hoping for the best. A CPL model that’s designed to get you real volume on one of these networks stands out pretty clearly against the tidy unit economics of a $4.22 average CPC on Google Ads.
It’s not a volume versus quality situation, and the efficiency comes pretty easily if you really get how your alternative traffic is going to behave once they hit your landers.
Understanding What You’re Working With
Popunder traffic works like this: A new browser window is generated in the background of the current session of a user. The ad isn’t in the user’s face; it’s there for the taking when they close or minimize their current tab. The ad is then displayed to the user. This is how viewability rates can be high even when the initial ad is not viewed immediately.
Push notification ads work like this: A user has opted in to receive notifications from a site. An ad is then pushed to the user’s screen without being inside the browsing session. It’s short and sweet with a headline, short description, and an icon or image. The click is the decision the user makes.
What is true about both is that the user wasn’t out there actively seeking your product. They weren’t halfway down the funnel searching for insurance or a utility app. This isn’t search traffic, and that will become apparent in a minute when we talk about how to design a funnel around the lack of search intent.
Choosing Traffic Sources and Understanding Inventory Quality
All pop or push inventory is not of the same quality. Large ad networks aggregate traffic from thousands of publishers, and the difference in performance between a high and low-quality zone within any network can be 10x or more. Some zones are clean human traffic from real publisher sites. Some resold inventory has been passed through 7 intermediaries, while some non-human traffic will never convert.
This is why source selection is so important, long before optimization even comes into the picture. Choosing a pop ads network that obtains traffic directly from publishers, rather than reselling from other networks, gives you access to lots of cleaner inventory right away. There are fewer middlemen, less chances for fraud, and the volumes are always more stable.
Once you’re in a network, the key is to start campaigns with intentionally broad targeting and then study the zone-level data that arises. The purpose of this first phase is not to “make it work”, it’s to discover which publisher zones are actually converting at your target CPL.
The Funnel Architecture: Why You Can’t Go Direct
You can’t send pop or push traffic directly to an offer page; it almost never performs well. The conversion rates are too low, you can’t make the economics work, and you’re just spending blindly. A pre-lander, sometimes called a bridge page, is the answer and for alternative traffic, you don’t even get to debate that, because a pre-lander is non-negotiable.
A pre-lander serves three roles. It explains to the user why this offer being shown to them isn’t out of left field. It engenders enough curiosity or perceived relevance that the user is willing to take the desired action. And as a gate, it blocks many of the low-intent users who would bounce immediately from the offer page and poison your conversion rates.
The best-performing pre-landers for alternative traffic are simple ones: a short quiz (“Find out what coverage you qualify for”), a sweepstakes entry frame (“Spin to win a $500 gift card”), or a single-page benefit teaser. These all have a single obvious call to action. The shorter, the better.
The quiz one is particularly powerful for many reasons. The single page though image with a single-field submission and a benefit heading is the simplest possible pre-lander and often what I reach for first when I launch a new angle.
It’s very good at creating micro-commitments. Each answer the user gives is one more step in the direction of seeing the information or result they want.
Making the user click your direct pop or push offer pass-through link seems small, but getting a user to click even a single time from any pop or push source isn’t easy. This step is mostly here just to stop the people who would bounce anyway.
Offers That Actually Convert on Alternative Traffic
Not all types of offers will be successful. Offers for expensive items that require a long decision-making process are not suitable for pop and push traffic since users have low intent and make fast decisions. What can work are offers that appeal to a wide audience, have low barriers to entry, and can be completed quickly.
Sweepstakes tend to perform consistently well because they only require an email address (sometimes a zip code as well). Financial verticals (personal loans, debt settlement, insurance quotes) perform well in most cases as they are high intent and the need exists in every demographic, regardless of what the user was searching for when they came in. Utility apps and VPN offers work in Tier 2 and 3 geos because the user does not need to make a payment to convert. Local services (solar quotes, roofing, home warranty) work in Tier 1 geos because the payouts are high enough to support the necessary CPLs.
The filter I would apply is that if someone needs twenty minutes of education before they would even consider converting, it’s probably not right for this traffic type. If you can show someone the offer, they understand the value, and they can convert in under three minutes, you should test it.
Technical Setup: Tracking is the Foundation
Running high amounts of alternative traffic when you don’t have a proper tracking setup and the right technical requirements is the easiest way to lose a lot of money and not know why. The technical requirements are not a nice little option. You need this setup.
What you need for this type of traffic is an ad tracker (Voluum, Binom – again – are the most used) that connects your actual campaign through server-to-server (S2S) postback tracking.
S2S postback passes the conversion data directly between the systems at the server level. It doesn’t rely on cookies which makes it essential for pop traffic since cookies often don’t fire reliably. When a conversion is made the affiliate network fires a postback to your tracker which then attributes that conversion back to the exact traffic source, zone ID, campaign, creative, and device that drove it.
The custom tokens are where zone-level optimization comes from. You want to (at least) capture: zone ID (which is the specific publisher source within the network), OS + device type (so you can separate the results and bid accordingly), country, or even the city for mobile pops, and obviously the campaign ID.
These automatically populate inside your tracker when you set your campaign URL properly and these are what you will use to identify which sources are burning your budget and which sources are converting. Voila.
You don’t have this? You’re flying blind.
You have it? Every $ you spend generates data which you can use to make actual decisions.
Zone Optimization: The Blacklist/Whitelist Method
Large scale alternative traffic campaigns are operated by a few basic tenets: most zones won’t work, a few zones will, and your job is to find the few as quickly as possible.
This involves heavy testing to gain the statistical confidence necessary to distinguish between random and non-random results.
Because each geo and site defies easy comparison to any other, and two identical looking banners that are running on the same site but purchased as RON from different brokers will have non-overlapping audiences and may as well be two completely different campaigns, the only way to do this testing fairly is one zone at a time. Which means one zone per day, if the test is taking 48 hours.
2-3x your target CPL is an easy rule of thumb for when to cut a zone. If you’ve bought 10,000 impressions that worked out to a pre-token ECPM of $2 with no conversion, block it and move on. You are never coming back, no matter how good the later converting stats of that zone you’ve been watching get. 3x your target is not going to become profitable when combined with the data you already have.
Frequency capping is part of this same logic. Set impressions per user to one per 24 hours, sometimes lower in Tier 1 markets. The goal is maximum unique reach, not repeated exposure to the same users who already didn’t convert. Dayparting, restricting your campaign to the hours where conversion rates are highest based on your data, further extends budget efficiency once you have enough data to act on it.
Push Notification Creative Testing
The success of push notification ads depends entirely on the creativity of the message. You have very little space: a headline, a short line, an icon, and possibly a large image. You can’t afford to be verbose. Trigger curiosity or urgency immediately, if you don’t, the user won’t click.
Test different combinations in a structured, scientific way. Create three or four headlines that play on unique emotions: scarcity, benefit, curiosity, question. Design two variations of each headline with different icons. Give them an even split of the budget and let the CTR determine the winner after 48-72 hours.
Never assume a particular creative is going to do well. Use the data to decide. Emojis, numbers, and specifics outperform general claims every time. “Save $400 on your monthly bill” will outperform “Save money on your bills.” “Only 3 spots left in your area” will outperform “Limited availability”.
In some cases, ‘badge’ icons can be added to the corner of the main icon. They can improve CTR for verticals where urgency or legitimacy is a key buying trigger. A flame emoji signals urgency. A checkmark signals legitimacy. Test the badge versions against the regular icon.
Scaling What’s Working
When you find a campaign that hits a CPL that’s consistently profitable over several days with stable volume, you should theoretically be able to scale indefinitely on alternative traffic sources. In practice, this doesn’t happen often because the approach most humans take is to just increase the budget slightly each day and hope that this will increase volume. Of course, as volume increases, the number of profitable impressions you receive stagnates as you’re not increasing the bids or altering anything but the budget.
At some point in the very near future, you realize ROI has dropped, so you end the campaign, assuming it’s exhausted and/or others have picked up on it. The correct approach works briefly when trying to extend the use of a good campaign, but you can’t rely on it as your only scaling method.
When you find something good and profitable, you should immediately increase the budget to the max daily spend for that offer/campaign on the current traffic source. Then you immediately take the exact campaign, pre-lander, and creative and launch it on the next alternative traffic source, assuming the current one worked similarly. Pick the next closest geo to what you’re already targeting but using a different segment of traffic. As an example if you’ve been buying desktop, go to mobile, if you’ve been doing the US go to the UK.
Landing the System
Constructing a high-volume lead generation funnel on non-mainstream traffic is no more difficult than doing it on Google or Meta, but it demands a different kind of discipline altogether. The traffic is cheaper and scales further, but it won’t forgive slack thinking. Nail the pre-lander. Put in place tracking that allows you to access zone-level information right from the start. Execute your test campaign systematically, blacklist liberally, and don’t even think about scaling until you’ve proven the numbers. If you achieve your target ROI, the upper limit on pop and push traffic is higher than most search campaigns will ever get you.
