Native placement optimization for lead offers
Native placement optimization for lead offers means passing each placement ID into a sub ID, judging every placement on sold leads and returns once it has had a fair amount of traffic, and then blocking the ones that do not sell. Keep the placements that sell on an allow list and give them more budget.
Key points
- Pass the network's site or placement ID macro into a sub ID, or you cannot tell good placements from bad ones.
- Set a judging threshold before launch, tied to your break-even cost per click, not to a feeling.
- Judge on sold leads and returns, and wait for returns before final calls on placements with leads.
- Block, keep, or move to an allow-list campaign; review on a fixed schedule and re-test blocked placements rarely.
Why placements decide native results
A native campaign runs across many websites at once, and often across several ad slots on each site. The same headline and thumbnail can produce sold leads on one placement and nothing on another, because the readers, the page layout and the share of invalid clicks differ from site to site. Averages across the whole campaign hide this.
You are paid rev share on sold leads, and unsold leads are not paid. So the question for every placement is not "does it get clicks?" but "does it send people whose leads sell and stay sold?". This guide covers that work in depth. For the full campaign setup, start with how to run native traffic to lead offers.
Site ID, placement ID and widget ID: what to pass
Native ad networks use different names for the same idea. Read your network's help pages for its list of URL macros (placeholders that the network fills in when someone clicks). Look for these values:
| Value | What it identifies | Why you need it |
|---|---|---|
| Site or publisher ID | The website where the ad showed | The main unit you block or keep |
| Placement, widget or section ID | The ad slot on that site (for example, below an article or in a feed) | Lets you keep a good slot on a site and block a weak one |
| Campaign ID | Your campaign | Separates offers, devices and flows |
| Ad or creative ID | The headline and thumbnail pair | Shows whether a placement fails for every ad or for one |
If the network passes only one of the first two, use it. If it passes both, put them in separate sub IDs. The sub ID naming template helps you plan which sub ID holds what.
Step 1: Build the click URL and test it
- Take your Summit tracking link for the offer (or your tracker's campaign URL, if your tracker sends to the Summit tracking link).
- Put the site or placement macro in one sub ID, the campaign macro in another, and the ad macro in a third.
- If you use a pre-lander, carry the same values from the pre-lander URL into the tracking link on its buttons.
- Preview the ad, click through, and check that each sub ID holds a real value, not the macro text in brackets.
- Submit nothing on the lander during the test. Confirm with your reports that clicks are arriving under the right sub IDs.
Leads are reported by tracking link with your sub IDs. Ask during onboarding how reporting is delivered to you and which other tracking options apply to your account. See tracking links and sub IDs explained.
Step 2: Set a judging threshold before launch
A placement with a handful of clicks and no leads has not failed yet. A placement with a lot of spend and no sold leads has. Decide in advance where the line is, so you do not block good placements too early or pay for bad ones too long.
- Work out your break-even cost per click from your sold rate and rev share using the rev share break-even calculator.
- Set a spend threshold per placement: the spend at which a placement should have produced at least one sold lead if it were close to break-even.
- Below the threshold, a placement is "too early to judge". At or above it, it gets a decision.
- Write the threshold in your campaign notes and use the same one at every review.
Example (made up, round numbers): a publisher's calculator shows a placement needs one sold lead for every 200 clicks to break even. They set the threshold at 300 clicks: any placement past 300 clicks with no sold lead is blocked; any placement past 300 clicks with sold leads is compared with break-even.
Step 3: Pull a placement report
On a fixed schedule (for example every few days in the first weeks, then less often), export two things and join them on the placement ID:
- From the native network: placement ID, impressions, clicks and spend.
- From your Summit reports: leads, sold leads and returned leads by the sub ID that holds the placement ID.
Add columns for cost per click, sold leads per click, earnings per click and the gap to break-even. Sort by spend, largest first. The placements at the top of the list are where your decisions save or earn the most.
Step 4: Decide for each placement
| What the placement shows | Decision | Note |
|---|---|---|
| Below the spend threshold | Keep running | Too early; review next time |
| Past the threshold, no leads | Block | The readers there are not responding to the offer |
| Past the threshold, leads but none sold | Block | Check whether one headline caused it before blocking a high-volume site |
| Sold leads, earnings per click below cost per click | Lower the bid on that placement, if your network allows it; block if it stays below | A bid change can bring it to break-even |
| Sold leads at or above break-even | Add to the allow list | Candidate for more budget |
| Sold leads, then many returns | Block or lower the bid; check for invalid traffic | Returns are deducted under the terms agreed during onboarding |
Returns can arrive after the lead is sold. For placements that produced leads, wait until those leads have had time to be returned before you add the placement to the allow list for good.
Running native traffic with placement-level tracking? Apply as a publisher with your sources and states.
Step 5: Run an allow-list campaign
Once your allow list holds enough placements to spend a useful budget, set up a second campaign that targets only those placements:
- Copy the original campaign, then limit it to the allow-list placements.
- Use the same offer, states, device and creative, so the only change is the placement list.
- Give it its own campaign value in the sub IDs so its leads are reported apart from the original.
- Raise bids on the allow-list campaign in small steps, and watch sold leads per click as you do.
- Keep the original campaign running at a lower budget to find new placements, and keep adding its blocks.
Scaling past that point is covered in how to scale a winning lead gen campaign.
Watch for invalid traffic at placement level
Some placements send clicks that do not come from interested people. Those leads are rejected or returned, and they can put your account at risk with the network and with buyers. Check each high-spend placement for:
- Click spikes at odd hours from one placement.
- Very short time on the pre-lander or lander from one placement.
- Many leads with contact details that do not work, from one placement.
- A high click rate with no sold leads after the threshold.
- Placement IDs that look generated or change constantly.
Block those placements early and tell the network. More detail is in bot and fraud traffic: how networks detect it.
When to re-test a blocked placement
Most blocks should stay. Re-test only when something about your campaign has changed enough that the old result may not apply: a new offer, a new headline angle, a new pre-lander, or a different device. When you re-test, do it in a separate small campaign with the same threshold, so a returning loser does not spend inside a working campaign.
Common mistakes
- No placement ID in sub IDs. Without it, there is nothing to optimize except the whole campaign.
- Blocking on clicks or click rate alone. A placement with a low click rate can still send readers whose leads sell.
- Blocking on one day of data. Use the threshold. A placement with little spend has not had a fair test.
- Allow-listing before returns come in. A placement can look strong until its leads are returned.
- Mixing devices. A placement can sell on desktop and fail on mobile. Keep devices in separate campaigns.
- Changing headlines and placements in the same review. Change one thing at a time so the result is readable.
Frequently asked questions
How many clicks should a native placement get before I block it?
Enough that, at your break-even sold rate, it should have produced at least one sold lead. Work that out from your break-even cost per click and set the threshold before launch.
Should I block a whole site or a single placement?
If your network passes both values, judge them separately. Block the slot that fails and keep the one that sells. If only the site ID is passed, judge the site.
How often should I review native placements?
On a fixed schedule: more often in the first weeks, then less often once your block and allow lists are settled. Daily changes on small data lead to wrong blocks.
Can I see sold leads by placement in my Summit reports?
Leads are reported by tracking link with your sub IDs. If a sub ID holds the placement ID, you can split sold and returned leads by placement. Ask during onboarding how reporting is delivered to you.
Related guides
- Native traffic for publishers: How native campaigns work with Summit landers.
- Native ad headlines for lead offers: Write and test headlines that qualify the reader.
- When to cut a losing lead gen campaign: Stop rules for the whole campaign.
- Reading your publisher reports: Sold, unsold and returned leads by sub ID.
- Why leads get returned: What returns say about a placement.
Send your best native placements to Summit landers
Every publisher is reviewed before going live. Email us your native setup, verticals, states and expected volume.
Or write to team@summitleads.ai. We reply by email.
What happens next
- Step 1: You email us.
- Step 2: We reply by email.
- Step 3: Every publisher is reviewed, and return and payment terms are agreed during onboarding, before you go live.
Already generating leads on your own forms? Sell your leads instead.