Q: Tiered commissions or one flat rate for my new program?
Start flat. Add tiers only once you have data on what a 'good' partner volume looks like. New programs that launch with five tiers usually set the thresholds wrong and end up overpaying mid-tier affiliates who'd have stayed at base.
When you do tier, build it on this:
— Tier on net revenue or validated sales, never on raw clicks or leads
— Set the first jump where your top 20% of partners already land, so it rewards real performers, not aspiration
— Make tiers reset monthly, not lifetime, so a partner has to keep producing
A cleaner alternative to tiers is a performance bonus: flat base rate plus a one-time kicker when a partner crosses a volume target. It's easier to budget and you can sunset it without renegotiating everyone's rate.
Caveat: every tier you add is a line item your finance team has to reconcile. Complexity has a cost on your side too.
Got a question? Send it in.
Start flat. Add tiers only once you have data on what a 'good' partner volume looks like. New programs that launch with five tiers usually set the thresholds wrong and end up overpaying mid-tier affiliates who'd have stayed at base.
When you do tier, build it on this:
— Tier on net revenue or validated sales, never on raw clicks or leads
— Set the first jump where your top 20% of partners already land, so it rewards real performers, not aspiration
— Make tiers reset monthly, not lifetime, so a partner has to keep producing
A cleaner alternative to tiers is a performance bonus: flat base rate plus a one-time kicker when a partner crosses a volume target. It's easier to budget and you can sunset it without renegotiating everyone's rate.
Caveat: every tier you add is a line item your finance team has to reconcile. Complexity has a cost on your side too.
Got a question? Send it in.
Q: How do I recruit affiliates who already promote my competitor?
Don't lead with a higher rate. Partners who switch for money switch again for money. Lead with what your competitor can't match.
What actually pulls them over:
— Better creative and landing pages (most programs starve affiliates here, and it directly lifts their conversion)
— A named manager who replies same-day, not a shared inbox
— Cleaner reporting and a longer cookie or attribution window than the incumbent
— Reliable, on-time payouts (ask around, this is where many programs quietly fail)
Find them by looking at who ranks for your competitor's brand-plus-review terms, who's in the comparison content, and who's active in the niche communities. Reach out with a specific reason you want them, not a blast.
Caveat: an affiliate who'll dump a competitor the day you call will dump you too. Offer a 60-90 day elevated rate to test the relationship before you commit to top terms.
Got a question? Send it in.
Don't lead with a higher rate. Partners who switch for money switch again for money. Lead with what your competitor can't match.
What actually pulls them over:
— Better creative and landing pages (most programs starve affiliates here, and it directly lifts their conversion)
— A named manager who replies same-day, not a shared inbox
— Cleaner reporting and a longer cookie or attribution window than the incumbent
— Reliable, on-time payouts (ask around, this is where many programs quietly fail)
Find them by looking at who ranks for your competitor's brand-plus-review terms, who's in the comparison content, and who's active in the niche communities. Reach out with a specific reason you want them, not a blast.
Caveat: an affiliate who'll dump a competitor the day you call will dump you too. Offer a 60-90 day elevated rate to test the relationship before you commit to top terms.
Got a question? Send it in.
Worth your feed
@NetworkVitals. Hard numbers on CPA networks: payout terms, EPC benchmarks, hold times and… We read it, you probably should too.
@NetworkVitals. Hard numbers on CPA networks: payout terms, EPC benchmarks, hold times and… We read it, you probably should too.
Q: Should I run my program on a network or build it in-house?
Depends on what stage you're at, not on which is 'better.' Use a network to start, move in-house when volume justifies the engineering.
A network buys you:
— Instant tracking, payouts, tax handling, and fraud screening you don't have to build
— Access to their existing affiliate base for distribution
— A neutral third party that affiliates already trust to pay them
In-house buys you:
— No override fee (networks take roughly 20-30% on top of your commission)
— Full data ownership and direct partner relationships
— Custom commission logic networks won't support
The honest math: if you're paying a network $X in overrides per month, compare that to a full-time program manager plus platform license. The crossover usually lands somewhere in low-to-mid five figures of monthly commission spend.
Caveat: in-house means you also own fraud, payouts, 1099s, and disputes. Networks earn their cut partly by absorbing that headache. Don't go in-house to save money and then under-resource it.
Got a question? Send it in.
Depends on what stage you're at, not on which is 'better.' Use a network to start, move in-house when volume justifies the engineering.
A network buys you:
— Instant tracking, payouts, tax handling, and fraud screening you don't have to build
— Access to their existing affiliate base for distribution
— A neutral third party that affiliates already trust to pay them
In-house buys you:
— No override fee (networks take roughly 20-30% on top of your commission)
— Full data ownership and direct partner relationships
— Custom commission logic networks won't support
The honest math: if you're paying a network $X in overrides per month, compare that to a full-time program manager plus platform license. The crossover usually lands somewhere in low-to-mid five figures of monthly commission spend.
Caveat: in-house means you also own fraud, payouts, 1099s, and disputes. Networks earn their cut partly by absorbing that headache. Don't go in-house to save money and then under-resource it.
Got a question? Send it in.
Q: What's the cheapest way to catch affiliate fraud without enterprise tools?
You don't need a fraud vendor on day one. Most early fraud shows up in patterns you can pull from your own reports.
Watch these signals:
— Conversion rate way above the partner's traffic source norm (50% conversion on cold display is fiction)
— Time-to-conversion clustered at seconds, which means cookie stuffing or pre-filled data
— Geo mismatch: clicks from one country, conversions from another
— A partner whose refund/chargeback rate runs 3-4x your program average
— Identical user-agents or IP ranges across 'different' conversions
Set a manual validation hold on any new partner's first 30 days, and review their cohort before you release payment.
Caveat: don't auto-ban on a single flag. Real partners have weird weeks. Pause the payout, ask the partner to explain the traffic source, and judge from there. False accusations cost you good affiliates and your reputation in a small industry.
Got a question? Send it in.
You don't need a fraud vendor on day one. Most early fraud shows up in patterns you can pull from your own reports.
Watch these signals:
— Conversion rate way above the partner's traffic source norm (50% conversion on cold display is fiction)
— Time-to-conversion clustered at seconds, which means cookie stuffing or pre-filled data
— Geo mismatch: clicks from one country, conversions from another
— A partner whose refund/chargeback rate runs 3-4x your program average
— Identical user-agents or IP ranges across 'different' conversions
Set a manual validation hold on any new partner's first 30 days, and review their cohort before you release payment.
Caveat: don't auto-ban on a single flag. Real partners have weird weeks. Pause the payout, ask the partner to explain the traffic source, and judge from there. False accusations cost you good affiliates and your reputation in a small industry.
Got a question? Send it in.
Q: Should I auto-approve affiliate applications or review each one?
Manual-review at launch, then graduate to rules-based auto-approval once you know what good looks like. Auto-approving from day one fills your program with bonus-hunters and brand-bidders you'll spend months cleaning out.
A workable application gate asks for:
— Their actual promotion method and primary traffic source
— A live URL or channel you can open and verify
— Whether they'll bid on your brand terms (this one question filters a lot)
Reject or hold anyone who won't name a traffic source, lists a parked domain, or whose 'site' is a thin coupon aggregator you didn't invite.
Once you've reviewed a few hundred, codify the patterns: auto-approve known-good sources, auto-hold the rest for human eyes.
Caveat: a slow approval queue kills momentum. If you make people wait a week, the motivated ones join a competitor instead. Aim for a 48-hour decision even while reviewing manually.
Got a question? Send it in.
Manual-review at launch, then graduate to rules-based auto-approval once you know what good looks like. Auto-approving from day one fills your program with bonus-hunters and brand-bidders you'll spend months cleaning out.
A workable application gate asks for:
— Their actual promotion method and primary traffic source
— A live URL or channel you can open and verify
— Whether they'll bid on your brand terms (this one question filters a lot)
Reject or hold anyone who won't name a traffic source, lists a parked domain, or whose 'site' is a thin coupon aggregator you didn't invite.
Once you've reviewed a few hundred, codify the patterns: auto-approve known-good sources, auto-hold the rest for human eyes.
Caveat: a slow approval queue kills momentum. If you make people wait a week, the motivated ones join a competitor instead. Aim for a 48-hour decision even while reviewing manually.
Got a question? Send it in.
Q: Should I let affiliates bid on my brand name in paid search?
Default to no, with narrow exceptions. When affiliates bid on your brand terms, they're charging you commission for a customer who was already searching for you. You pay twice: ad cost competition plus commission.
Your policy should state:
— No bidding on the brand name, common misspellings, or brand-plus-modifier ('[brand] coupon', '[brand] login')
— No brand in the ad copy or display URL
— Direct-linking to your site from a brand ad is an instant violation
The narrow exception: you can authorize specific trusted partners to defend your brand SERP against competitors, on terms you set, usually at a reduced commission.
Enforce it by monitoring your brand SERP on a schedule and from multiple geos, since violators often geo-target away from your office location.
Caveat: write the rule in plain language with examples. 'No trademark bidding' is ambiguous to a new affiliate. '[brand] discount = violation' is not.
Got a question? Send it in.
Default to no, with narrow exceptions. When affiliates bid on your brand terms, they're charging you commission for a customer who was already searching for you. You pay twice: ad cost competition plus commission.
Your policy should state:
— No bidding on the brand name, common misspellings, or brand-plus-modifier ('[brand] coupon', '[brand] login')
— No brand in the ad copy or display URL
— Direct-linking to your site from a brand ad is an instant violation
The narrow exception: you can authorize specific trusted partners to defend your brand SERP against competitors, on terms you set, usually at a reduced commission.
Enforce it by monitoring your brand SERP on a schedule and from multiple geos, since violators often geo-target away from your office location.
Caveat: write the rule in plain language with examples. 'No trademark bidding' is ambiguous to a new affiliate. '[brand] discount = violation' is not.
Got a question? Send it in.
Q: What should affiliate onboarding cover in the first 30 days?
Your goal in month one is a partner's first conversion, because partners who convert early stay, and ones who don't go dormant forever. Engineer toward that one event.
A tight onboarding sequence:
— Day 0: welcome with their links working, top 3 converting creatives, and the program terms in plain language
— Day 1-3: a short call or message offering your best-performing landing page and offer for their specific traffic type
— Day 7: check they've placed links; if not, ask what's blocking them
— Day 14: share a benchmark ('partners like you see X'), which sets expectation and nudges effort
— Day 30: review their first results together and recommend the next offer
Give new partners your converting assets, not a generic banner pack. The fastest path to their first sale is your existing proof.
Caveat: don't automate all of it. One human touch in the first week measurably lifts activation. Reserve your time for partners who've actually placed a link.
Got a question? Send it in.
Your goal in month one is a partner's first conversion, because partners who convert early stay, and ones who don't go dormant forever. Engineer toward that one event.
A tight onboarding sequence:
— Day 0: welcome with their links working, top 3 converting creatives, and the program terms in plain language
— Day 1-3: a short call or message offering your best-performing landing page and offer for their specific traffic type
— Day 7: check they've placed links; if not, ask what's blocking them
— Day 14: share a benchmark ('partners like you see X'), which sets expectation and nudges effort
— Day 30: review their first results together and recommend the next offer
Give new partners your converting assets, not a generic banner pack. The fastest path to their first sale is your existing proof.
Caveat: don't automate all of it. One human touch in the first week measurably lifts activation. Reserve your time for partners who've actually placed a link.
Got a question? Send it in.
Q: What KPIs should I hold my affiliate manager to?
Not 'number of affiliates recruited.' That metric rewards a bloated, dead roster. Hold them to activation and revenue quality instead.
The KPIs that drive a healthy program:
— Active partner count (placed a link and drove a sale in the period), not total signups
— Activation rate of new recruits within 60 days
— Revenue from partners outside your top 5, which measures whether the program has depth or one whale
— Net revenue after clawbacks and refunds, so they don't chase junk volume
— Partner retention / reactivation of dormant accounts
A good manager grows the middle of the roster and reduces single-partner dependency. That's the durable program.
Caveat: don't over-index on new recruitment in any single quarter. A manager can hit a recruiting target by lowering the application bar, which dumps cleanup work on you next quarter. Pair every growth metric with a quality metric.
Got a question? Send it in.
Not 'number of affiliates recruited.' That metric rewards a bloated, dead roster. Hold them to activation and revenue quality instead.
The KPIs that drive a healthy program:
— Active partner count (placed a link and drove a sale in the period), not total signups
— Activation rate of new recruits within 60 days
— Revenue from partners outside your top 5, which measures whether the program has depth or one whale
— Net revenue after clawbacks and refunds, so they don't chase junk volume
— Partner retention / reactivation of dormant accounts
A good manager grows the middle of the roster and reduces single-partner dependency. That's the durable program.
Caveat: don't over-index on new recruitment in any single quarter. A manager can hit a recruiting target by lowering the application bar, which dumps cleanup work on you next quarter. Pair every growth metric with a quality metric.
Got a question? Send it in.
Q: Are coupon and deal-site affiliates worth paying full commission?
Usually not full rate, but don't ban them outright. The problem is attribution: coupon sites often catch a customer at the last click after your own marketing did the work, so you pay commission on a sale you'd have made anyway.
How to handle them:
— Put coupon/loyalty/cashback partners in a separate commission group at a reduced rate
— Use last-click-excluding rules or a 'no-voucher' commission tier so they don't override an affiliate who actually introduced the customer
— Only issue exclusive codes you can track, never let them surface codes you didn't authorize (that trains customers to abandon checkout and hunt for one)
They do have real value for cart-abandonment recovery and price-sensitive closers. Just price that value correctly.
Caveat: audit your checkout for a 'have a code?' field that leaks margin. If it sends buyers off-site to find a code, you're funding coupon affiliates to solve a problem you created.
Got a question? Send it in.
Usually not full rate, but don't ban them outright. The problem is attribution: coupon sites often catch a customer at the last click after your own marketing did the work, so you pay commission on a sale you'd have made anyway.
How to handle them:
— Put coupon/loyalty/cashback partners in a separate commission group at a reduced rate
— Use last-click-excluding rules or a 'no-voucher' commission tier so they don't override an affiliate who actually introduced the customer
— Only issue exclusive codes you can track, never let them surface codes you didn't authorize (that trains customers to abandon checkout and hunt for one)
They do have real value for cart-abandonment recovery and price-sensitive closers. Just price that value correctly.
Caveat: audit your checkout for a 'have a code?' field that leaks margin. If it sends buyers off-site to find a code, you're funding coupon affiliates to solve a problem you created.
Got a question? Send it in.
Q: My roster is full of dead affiliates. Should I deactivate them?
Clean them out, but reactivate before you remove. A dormant partner once chose your program. That's a warmer lead than a cold stranger.
Run this sequence:
— Segment partners with zero clicks in 90+ days
— Send a targeted reactivation: a new top offer, a temporary rate bump, or fresh creative aimed at their original traffic type
— Give it two touches over a few weeks
— Deactivate only the truly silent, and keep their data so you can re-invite later
A bloated roster isn't just cosmetic. It distorts your active-rate metrics, slows your reporting, and hides which partners actually matter.
Caveat: don't deactivate a partner who drives one big sale a year (seasonal or event-driven traffic is real). Check lifetime value, not just last-90-days activity, before you pull anyone with a meaningful history.
Got a question? Send it in.
Clean them out, but reactivate before you remove. A dormant partner once chose your program. That's a warmer lead than a cold stranger.
Run this sequence:
— Segment partners with zero clicks in 90+ days
— Send a targeted reactivation: a new top offer, a temporary rate bump, or fresh creative aimed at their original traffic type
— Give it two touches over a few weeks
— Deactivate only the truly silent, and keep their data so you can re-invite later
A bloated roster isn't just cosmetic. It distorts your active-rate metrics, slows your reporting, and hides which partners actually matter.
Caveat: don't deactivate a partner who drives one big sale a year (seasonal or event-driven traffic is real). Check lifetime value, not just last-90-days activity, before you pull anyone with a meaningful history.
Got a question? Send it in.