How to build a solo ad seller scorecard from your own data

A solo ad seller scorecard is a spreadsheet or dashboard where you grade every seller you buy from on the metrics that actually matter: delivered clicks, unique clicks, conversion rate, earnings per click, refund rate, and list responsiveness over time. You build it by tagging every solo ad campaign with a unique tracking link, pulling the data into one place, and scoring each seller against your own baseline instead of their sales page promises.

Here's why this matters more than any marketplace rating: seller quality changes. A vendor who crushed it for you in March can quietly swap in a lower-quality traffic source by June, and you won't notice unless you're comparing their numbers against their own history. A scorecard turns "this seller feels sketchy" into "seller #14 has dropped from 12% opt-in to 4% over three buys, stop buying." That's the difference between a hobby and a business.

This guide walks through the exact fields, formulas, and thresholds I use. You can build it in Google Sheets in about 30 minutes, or automate most of it with a tracker that already stores per-source performance.

What a solo ad seller scorecard actually needs to track

Most buyers track one number: clicks delivered. That's the least useful metric on the list, because solo ad sellers control it. What you want are metrics the seller can't fake — or can only fake at real cost to themselves.

Here's the minimum viable scorecard. Every row is one purchase from one seller.

Field	What it tells you
Seller name / marketplace ID	Who to buy from again (or blacklist)
Purchase date	Detects quality drift over time
Spend	Denominator for every cost metric
Clicks delivered (seller-reported)	What they claim
Clicks tracked (your tracker)	What actually arrived
Unique clicks	Filters out refresh spam
Delivery ratio	Tracked ÷ reported. Below 0.85 is a problem
Opt-ins / leads	Top-of-funnel response
Opt-in rate	Leads ÷ unique clicks
Sales / conversions	Bottom line
Revenue	Commission or product value
EPC	Revenue ÷ unique clicks
ROI	(Revenue − spend) ÷ spend
Refund / chargeback rate	Catches low-intent or incentivized traffic
Geo + device mix	Catches mismatched traffic (Tier 3 when you bought Tier 1)
Fraud flags	Datacenter IPs, bot UAs, IP velocity
Notes	Open rate, reply rate, complaints, delivery delays

That's 17 columns. You don't need all of them on day one. If you're just starting out, the four that separate good sellers from bad are delivery ratio, opt-in rate, EPC, and refund rate. Everything else is diagnostic — it tells you why a seller failed, not that they failed.

Step 1: Tag every solo ad purchase with a unique tracking link

You cannot build a scorecard without clean attribution. If two sellers send traffic to the same URL, you have no idea which one produced the sale.

The fix is one tracking link per seller per purchase. Not per seller — per purchase. If you buy from the same vendor three times, you want three separate links so you can see whether buy #3 performed like buy #1.

Use sub-IDs to encode the seller and the buy. Something like:

https://yourdomain.com/go/offer?seller=trafficjoe&buy=2026-03-14

In Adtraxo, you'd create a campaign per seller and use sub1 through sub5 to pass seller name, marketplace, purchase date, and list segment. The tracker stores all of it against every click, so when a conversion fires via postback, the revenue lands on the right row automatically. No manual reconciliation, no "I think this sale came from Mike's list."

If you're running on Udimi specifically, the setup is slightly different — Udimi appends its own click IDs to your link, and you need to make sure your tracker captures them. That process is covered in detail in Udimi tracking setup: how to track solo ad clicks properly.

Step 2: Measure delivered clicks, not promised clicks

Sellers report clicks. Your tracker counts clicks. These two numbers should be close. When they aren't, you've found either a technical problem or a lie.

Calculate delivery ratio:

Delivery ratio = tracked clicks ÷ seller-reported clicks

Benchmarks I use after a few hundred solo ad buys:

• 0.95–1.05: Normal. Small variance is expected from ad blockers, redirect failures, and people closing the tab mid-redirect.
• 0.85–0.95: Watch it. Could be a slow redirect chain or a tracking script loading late.
• Below 0.85: The seller is padding numbers. Ask for a refund or a make-good.
• Above 1.10: Suspicious in the other direction — usually means your tracker is double-counting, or the seller is running your link through a rotator that hits it multiple times.

Unique clicks matter more than raw clicks. If a seller delivers 500 clicks but only 210 uniques, you paid for 290 refreshes. That's a 58% waste rate. Any decent tracker will separate these for you; if you're counting raw clicks only, you're flying blind. The mechanics of how sellers fake this are worth understanding — see How to spot fake solo ad clicks before the seller gets paid.

Step 3: Score conversion quality, not just click volume

Clicks are the input. Opt-ins and sales are the output. A seller delivering 800 clicks at a 2% opt-in rate is worse than a seller delivering 300 clicks at 14%. The second seller costs more per click and delivers more leads per dollar.

Two ratios do most of the work:

Opt-in rate = leads ÷ unique clicks. For a typical squeeze page with a lead magnet, a healthy solo ad seller produces 8–20%. Below 5% means the list isn't engaged with your offer, or the traffic is junk. Above 25% often means incentivized traffic that will never buy anything.

EPC = revenue ÷ unique clicks. This is the number that decides whether you buy again. If you're paying $0.60 per click and your EPC is $0.40, you're losing money on every click. If your EPC is $1.10, you scale.

Here's the trap: opt-in rate and EPC don't always move together. Some lists opt in hard and never buy. Some lists opt in at 6% and convert at 4% on the backend. Track both, and weight EPC higher when you're making the buy/no-buy call.

What counts as "good" varies wildly by niche and offer type. I've broken the ranges down by vertical in What is a good solo ad conversion rate in 2026? — worth reading before you set your scorecard thresholds.

Step 4: Layer in fraud and traffic-quality signals

A seller can deliver real clicks from real browsers on real residential IPs and still send you garbage — if those browsers are in the wrong country, on the wrong device, or running headless Chrome on a VPS.

This is where per-campaign breakdowns earn their keep. For every solo ad buy, check:

1. Geo distribution. If you bought a US list and 60% of clicks come from India, Vietnam, and Nigeria, you were sold a mixed-traffic list. Pull the geo report and compare it to what the seller promised.
2. Device mix. Desktop-only traffic on a mobile-optimized offer converts badly. If a seller claims to run a mobile-heavy list and you see 90% desktop, the numbers don't add up.
3. Datacenter IPs. Real solo ad traffic comes from residential and mobile connections. Any meaningful percentage of datacenter IPs is a bot signal.
4. IP velocity. Hundreds of clicks from a handful of IPs means one machine is refreshing your link.
5. User-agent anomalies. Old browser versions, empty UAs, or headless markers.
6. Referer patterns. If clicks arrive with no referer or a referer that doesn't match the seller's domain, something's off.

Adtraxo's fraud detection flags all six of these automatically and attaches them to the campaign, so when you're filling in your scorecard you can see "buy #4 from seller X: 31% datacenter IPs, 22% IP velocity flag" without running any manual checks. That's a column you'd otherwise have to build yourself.

If you're not using a tracker with built-in fraud detection, you can approximate it with a spreadsheet and a few manual spot-checks, but you'll miss the fast-moving signals. The red flags worth memorizing are catalogued in 9 solo ad red flags that mean you're buying bot traffic.

Step 5: Build the scoring formula and set your thresholds

A scorecard without a score is just a table. Here's a simple weighted model that works. Score each seller 0–100.

Metric	Weight	Score 100	Score 50	Score 0
Delivery ratio	15%	≥ 0.95	0.90	≤ 0.80
Unique click %	10%	≥ 85%	70%	≤ 50%
Opt-in rate	20%	≥ 15%	8%	≤ 3%
EPC vs. your target	30%	≥ 1.2× target	1.0× target	≤ 0.5× target
Refund rate	10%	≤ 2%	5%	≥ 12%
Fraud flags	15%	None	1 minor	Any major

Multiply each score by its weight, sum, and you get a 0–100 seller rating. In practice:

• 80–100: Tier 1. Rebuy, increase order size, negotiate volume pricing.
• 60–79: Tier 2. Rebuy at same size, test a different offer or landing page before scaling.
• 40–59: Tier 3. One more shot with a fresh link and a different offer. If it doesn't clear 60, drop them.
• Below 40: Blacklist. Request a refund and don't buy again.

Two rules that keep the scorecard honest:

1. Score per purchase, not per seller. A seller's average hides the fact that buy #1 was great and buy #4 was trash. Look at the trend line, not the mean.
2. Require a minimum sample. Don't score a seller on 40 clicks. Wait until you have at least 200 unique clicks and 5 conversions before the EPC column means anything.

How to use the scorecard to negotiate and reallocate budget

The scorecard pays for itself the moment you walk into a negotiation with data. Sellers who claim "my list always converts at 15%" get quiet fast when you show them a 4.2% opt-in rate across three buys with your tracker's screenshots attached.

Three plays that work:

Volume discounts on proven sellers. Once a seller clears 80 on three consecutive buys, ask for 15–25% off on a 1,000-click order. You have the receipts. Specific tactics are in How to negotiate with solo ad sellers using your own tracking data.

Make-good requests on failed buys. If delivery ratio came in at 0.72, you didn't get what you paid for. Ask for a re-send or a partial refund. Most sellers will comply once they know you're tracking.

Budget reallocation. Once you have 10–15 sellers scored, you'll usually find that 2–3 of them produce 80% of your profit. Shift budget there and cut the bottom half. This is the single highest-leverage thing a solo ad buyer can do.

If you're just starting and don't have seller history yet, the vetting process before your first buy matters more than the scorecard. That's covered in How to vet a solo ad seller before you spend a dollar and The solo ad buyer's playbook: vetting sellers and tracking every click.

Scaling the scorecard: from spreadsheet to tracker

A Google Sheet works fine for the first 20–30 buys. Past that, manual entry becomes the bottleneck and you start skipping rows. That's when you move the scorecard into your tracker.

Adtraxo's per-campaign analytics give you most of the columns automatically — clicks, uniques, conversions, revenue, EPC, ROI, geo, device, and fraud flags — grouped by the sub-ID you assigned to each seller. You export the report weekly, paste it into a scoring tab, and the formula does the rest. The AI Campaign Optimizer agent will also flag campaigns that are underperforming against your baseline, which is effectively the scorecard doing its own triage.

Either way, the principle is the same: the seller's reputation is a hypothesis, your data is the verdict. Build the scorecard once, and you'll never overpay for a mediocre list again.

Frequently asked questions