Hey everyone,

Welcome back for another bite to chew on.

We have spent a few issues making the case for affiliate.

Pay on the sale, not the impression. Run a roster, not a sign-up funnel. Build the program before the Q4 auction starts.

All of that still holds.

But there is a version of a growing affiliate program that looks great on the dashboard and loses money on the P&L.

Most operators cannot tell which version they are running.

The reason is that affiliate is graded on revenue.

A commission is a percentage of an order, so the channel reports orders. It does not report whether the order was going to happen anyway, whether the customer ever came back, or whether the box came back instead.

Those three questions decide whether a 15% commission was a bargain or a 15% tax on money you already had.

Obvi runs a heavy subscription program, so we judge every acquisition channel on the second order, and affiliate gets no exemption.

This issue is about the margin side of the channel.

Not whether to run it, but how to pay it so a record quarter does not arrive with thinner margins than the one before.

Let's get into it.

One brand hit their 4x ROAS target all Q4. What they didn't realize? 22% of those "wins" were losing money on every order.

Every holiday season, brands hit record sales and still come out with thinner margins, because acquisition costs climb, discounts creep, and fulfillment gets messier the moment volume spikes.

Here's the number that should worry you:

Stockouts cost brands 140% more than baseline on Black Friday, and 248% more on Cyber Monday.

That's not a traffic problem. That's a margin problem hiding inside a revenue win.

Levanta pulled together the real levers 9 e-commerce leaders are using right now to protect margin before Q4 peaks, covering pricing, paid ads, fulfillment, returns, and acquisition.

No predictions, no "do more with less" nonsense. Just the plays that are actually working.

Inside, you'll find out how to:

Click here to get the full playbook for FREE.

Every affiliate dashboard shows the same four numbers: clicks, orders, revenue, commission.

Divide revenue by commission and you get a ROAS that looks excellent next to paid social.

That number is real.

It is also the only number the channel will volunteer, and it stops at the first order.

Picture two partners on the same 15% rate, each driving $20,000 in a quarter.

The dashboard ranks them as equals.

Partner A's customers reorder at 35% inside 90 days and half of them start a subscription.

Partner B's customers bought the discounted bundle once and never came back.

Twelve months out, A's cohort is worth more than double B's on the same commission spend.

Those numbers are illustrative, but the shape shows up in almost every program we have looked at.

The spread between the best and worst partner on repeat rate is usually wider than the spread on conversion rate. You just cannot see it from inside the affiliate platform.

The first is new versus returning.

Some share of the orders coming through a creator's link are your existing customers who went looking for a code before checking out.

A 15% new-customer commission paid on a returning customer is a 15% coupon you never meant to issue.

Most programs find this number is not zero, and for a few partners it is most of the volume.

The second is 90-day repeat rate, or subscription take rate if you sell one. This is the number that separates the two partners above.

The third is discount depth per order.

Some content converts full-price buyers. Some only converts at 25% off, and every order it drives arrives with that margin already gone.

None of this comes out of the affiliate dashboard.

It comes from your order data, joined to the code or link that drove the order.

That is a monthly spreadsheet job, not a data project. It takes an afternoon the first time and an hour every month after.

Once you can see cohort quality per partner, the payout can follow it.

Pay a higher rate on first-time customers and a reduced rate, or nothing, on returning ones.

Add a flat bonus for every subscription start.

Give review and comparison content a longer attribution window, because it converts slower and returns better customers than flash-deal content does.

For a subscription product, a partner who starts subscriptions is worth more per order than one who moves one-time bundles, and the payout should say so out loud.

Creators respond to the rate card. Tell them you pay more for subscribers and the content shifts toward the routine, the refill, and the reason to stay.

The goal is not to pay creators less. It is to point the same commission budget at the partners whose customers are still buying in month six.

The most expensive affiliate order is the one that did not need an affiliate.

It happens at checkout.

A shopper has the cart built and the card out.

A coupon browser extension lights up with a code it found, applies it, and in the same moment drops its own affiliate attribution on the order.

Your brand pays the discount plus a commission on a sale that was closing at full price thirty seconds earlier.

Several of the largest coupon extensions have been publicly accused of exactly this.

The second version does not need an extension.

Creator codes get scraped onto coupon aggregator sites within days of going live.

From then on, every shopper who searches your brand name plus "discount code" finds a creator's 20% off and uses it.

The creator gets paid for a sale they never touched.

The shopper gets a discount they never would have asked for.

You pay for both.

Take a $60 order that was closing at full price.

A leaked 20% creator code brings revenue to $48.

A 15% commission on $48 is $7.20.

The order that was worth $60 a moment ago is now worth $40.80, and the $19.20 gap bought nothing.

If contribution before marketing was $22 on that SKU, it is now under $3.

The dashboard records another affiliate conversion.

The creator whose code leaked climbs the leaderboard.

Their conversion rate and revenue per partner look elite, so the program manager raises their rate and sends them more product.

The leak corrupts the per-partner data you use to decide who gets scaled, which is the quiet part of the damage.

Start with the codes themselves.

Give every creator a unique code and rotate it.

Search your brand name plus "promo code" once a month and kill any code that shows up on an aggregator.

Make codes first-order only where your platform allows it, and do not let them stack with sitewide sales.

Then fix the terms.

Exclude coupon, toolbar, and extension partners from the program outright, or put them on a separate rate low enough that the math still works.

Set attribution so a creator's code beats a last-second cookie, and so content that introduced the product wins over a toolbar that appeared at checkout.

Most serious affiliate platforms let you set both rules. Almost nobody turns them on.

The same thing happens on Amazon with social promo codes that hit deal sites.

Treat them the same way: unique per partner, short-lived, and pulled the moment they leak.

Most programs pay commission on a 15 or 30 day cycle.

Most return windows are 30 days, and in Q4 many brands extend them to the end of January.

That means holiday commissions clear before the returns do.

In January, the gifting orders start coming back and the commission on them has already gone out the door.

There are two ways to fix this, and both are fine.

Hold commissions until the return window closes, then pay on net orders.

Or pay fast and net any returns and cancellations against the next payout.

The second keeps creators happy, because they still see money quickly, and keeps your margin honest.

Pick one, write it into the terms, and apply it to everyone the same way.

What creators hate is not a validation period.

It is a payout schedule they cannot predict.

A fixed rule they can plan around beats a fast payout that gets quietly clawed back with no explanation.

When one partner's orders return at twice the program average, the payout is the last thing to fix.

The content is over-promising, the sizing guidance is missing, or the audience bought for the discount and never wanted the product.

Deal-driven buyers return more.

That is not a moral failing, it is a pattern, and it shows up in the partner's return rate before it shows up anywhere else.

The fix lives in the brief and the seeding.

Give creators the product long enough to actually use it.

Send sizing guidance.

Tell them who it is not for.

Content that says "skip this if you want X" converts fewer people and returns far fewer boxes, and the buyers it does convert stay.

One more move worth stealing from the returns side:

Route affiliate-driven returns toward exchanges instead of refunds, and keep the commission intact on the exchange.

The creator still gets paid, the customer stays a customer, and the order does not unwind.

The monthly table is simple.

For each partner: gross revenue, discounts, returns, net revenue, commission, fulfillment and payment fees, contribution.

Rank by contribution per order, not by revenue.

The ranking will not match the affiliate dashboard, and the gap between the two lists is where the margin has been going.

The decisions fall out of the table.

Partners with strong contribution can afford a rate increase, and they should get one before a competitor offers it.

Deal-driven partners move to a first-order-only rate.

Coupon partners get cut.

A partner with high revenue and negative contribution is not a top performer.

They are a discount you are paying twice for.

Run this table for October and November before anyone sets a Cyber Week commission boost.

The boost should go to the partners the table says can carry it, not to the ones the dashboard says are winning.