The Affiliate Program Settings Most Merchants Never Change

93.1% of Shopify affiliate programs never change the attribution window they were given.

They keep the 30-day default, and in almost every case nobody sat down and decided on 30 days.

Meanwhile the commission rate gets debated for a week.

That is backwards, and this article is about why.

The attribution window decides which clicks turn into commissions at all.

The rate only decides what a commission is worth once one exists.

One of those is a bigger lever than the other, and it is not the one you have been thinking about.

Below are the six settings that actually shape a program, ranked by how much they matter, with what the distribution across 1,178 Shopify affiliate programs shows for each.

Every one of them is free to decide this week and expensive to change later, because once affiliates have published links under one set of rules, changing the rules means renegotiating with every person who trusted you.

1. The attribution window

The decision: how long after someone clicks an affiliate link does that affiliate still get credit if the person buys.

What the distribution shows:

Shopify affiliate programs that change the default attribution window

Program setting	Share of programs
Kept the 30-day default	93.1%
Set a different window	6.9%

Almost nobody touches it.

That is not evidence that 30 days is wrong, and this article is not going to tell you to change it.

It is evidence that the most consequential setting in the program is arriving by default rather than by decision, which is a different problem and a worse one.

What to pick:

Start from how long your customers actually take to buy.

On the store side of our data, the median time from an affiliate click to the resulting order is about nine and a half minutes, and the distribution is bimodal: a large group buys almost immediately, and a much smaller group takes days or weeks.

If you sell something people buy on impulse, the window is doing very little work and 30 days is a comfortable default that costs you nothing.

If you sell something expensive that people research, compare and come back to, the window is doing a great deal of work and 30 days may be cutting off exactly the affiliates who are doing the hardest part of the job.

Our guide to cookie duration covers the mechanics of how the window is actually enforced once you have set it.

Why now: shortening a window later takes money out of the pockets of people who are already promoting you, under rules they agreed to.

Lengthening it is easy.

Shortening it is a conversation you will lose.

2. The commission rate

The decision: what percentage of the order the affiliate earns, and what that percentage applies to.

What the distribution shows: 10% is both the most common rate and the median rate actually realised across programs in our data.

That is a genuinely useful number, because it tells you what your affiliates are comparing you against when they decide whether to bother.

Here is the part that should change how much time you spend on this.

Across 471 programs with at least three affiliates, the correlation between commission rate and program performance is ρ = -0.032, with a 95% confidence interval from -0.122 to 0.059.

That interval straddles zero, which means we cannot distinguish the relationship from nothing.

State the limit honestly, because it matters: that is a correlation across different programs, not an experiment on any one of them.

It does not prove that raising your own rate would do nothing, and nothing in this article depends on that claim.

What it does support is narrower and still useful: the rate is the setting merchants over-think relative to its leverage, and the programs paying most are not visibly the programs performing best.

What to pick: 10% unless you have a specific reason, and spend the time you saved on the other five settings.

If your margins or your order value genuinely argue for something else, how to set your affiliate commission rate works through it properly.

The more important question is what the percentage applies to.

Decide now:

• Whether you pay on the list price or on the discounted amount
• Whether shipping and tax are included

A 12% rate on list and a 12% rate after a 20% code are different offers, and your affiliates will eventually work out which one they were given.

Why now: you can raise a rate whenever you like, and people will be pleased.

Lowering one, or quietly changing what it is calculated on, is how programs lose their best affiliates.

3. Links, codes, or both

The decision: how the affiliate gets credited in the first place.

A tracking link attributes the visit.

A discount code attributes the order.

Most merchants pick one without realising they have picked, usually because the app they installed defaulted to one.

What to pick: both, with the code unique to each affiliate.

Links catch the people who click and buy in the same session, which our timing data says is most of them. Codes catch the people who saw a recommendation on a podcast, a video or a story and typed the brand name into Google afterwards, which no link can ever capture.

The failure mode is code alone, and it has a predictable ending.

A code that works for anyone spreads.

It gets posted to a coupon site, and from then on you are paying commission and giving a discount to customers who were going to buy at full price anyway, which is the same margin twice.

Per-affiliate codes are what contain that, because a leaked code identifies itself. We covered the rest of the fix in affiliate coupon code leaks.

Why now: the affiliate who has published a shared code in ten videos cannot unpublish it.

Issuing individual codes at the start costs nothing.

Retrofitting them means asking every affiliate you have to change every piece of content they have made.

4. Which one wins when a link and a code both touch the same order

The decision: affiliate A's link brought the visitor.

Affiliate B's code got typed into the discount box at checkout.

One order, two claims.

Who gets paid?

This question gets asked in Shopify's own community every few months and it does not get a satisfying answer, because merchants assume Shopify decides it.

Shopify does not decide it, and it is worth understanding why.

Shopify records the discount code on the order, because a discount code is a Shopify object with a real effect on the total.

Shopify has no concept of an affiliate at all, so the click attribution does not live there.

It lives in whatever affiliate app you installed, which watched the visit arrive and tagged it.

So at the moment the order is placed, two independent systems each hold half a claim, and the only thing that can reconcile them is a precedence rule that somebody chose.

If you have not chosen one, your app chose for you.

What to pick: find out what your app does by default, then write your rule into your program terms in one sentence so it is answerable before it is disputed.

The common options are:

• Code wins
• Link wins
• Last touch wins

Any of them is defensible.

None of them is defensible after the fact, to an affiliate who has just watched a sale they generated get credited to somebody else.

There is a cleaner way out of this, which is to stop treating the link and the code as two things.

Affilitrak issues auto-apply coupon links, where the affiliate shares one link, the discount applies automatically when the visitor clicks it, and the click and the code are the same event.

That removes the collision rather than arbitrating it, and it also removes the "did I have to type anything" friction at checkout.

If you want the broader picture of what can be tracked and how, tracking affiliate sales on Shopify covers the methods.

Why now: the first time this comes up will be a real order with a real affiliate asking a fair question.

Deciding then looks like a decision made to suit you.

5. Who gets approved, and on what

The decision: does anyone who applies get in, and if not, what are you actually looking at when you say no.

Most advice on this is abstract, so here is the concrete version, from the approvals screen we built.

The field that does the most work is the order history, and it is the argument for approving from inside your store rather than from a form.

Someone who has already bought from you three times is a different prospect from someone who found your program page, and no self-reported application field tells you which one you are looking at.

What to pick: approve quickly, and screen on two things rather than five:

• Does this person's audience plausibly want what you sell?
• Have they ever bought it themselves?

Long application forms filter out people who could not be bothered filling them in, which is not the same as filtering out people who would not have sold anything.

Why now: an auto-assigned rate at approval means you are handing out the rate you decided in section 2 rather than improvising per applicant.

Improvised rates are how programs end up with four affiliates on four different deals and no way to explain any of them.

6. The payout threshold and schedule

The decision:

• The minimum balance before you pay
• How often you pay
• How the money actually moves

This is the cheapest section here and the one most merchants meet for the first time when an affiliate asks to be paid.

We do not have first-party data on thresholds, so this is reasoning rather than measurement, and I would rather say so than dress it up.

What to pick: a threshold low enough that your smallest genuine affiliate can reach it.

Run the arithmetic before you set it: a $100 threshold on a program paying $4 a sale means twenty-five sales before anyone sees a penny, which is above the lifetime total of most affiliates you will recruit.

A hold period between payout and order is reasonable, because you are waiting out returns, and your returns window is the number to use rather than a round one you picked.

For the mechanics, Affilitrak pays automatically through PayPal, and where PayPal is not an option it lets you record payments you made outside the app and collect payment details from affiliates through a form, so the ledger stays accurate either way. The payouts setup guide walks through both.

Why now: this is the only setting on the list where getting it wrong produces a public complaint rather than a quiet loss.

Conclusion

Decide these six things before you recruit anyone, in this order.

1. The attribution window: start from how long your customers actually take to buy, not from the default. Lengthening it later is easy, shortening it is not.
2. The commission rate: 10% unless you have a reason, and decide what the percentage is calculated on in the same sitting.
3. Links and per-affiliate codes, both: never a shared code, ever.
4. The precedence rule: for when a link and a code touch one order, written into your terms in one sentence.
5. The approval bar: set at audience fit and prior purchase, with the rate assigned automatically so you are not negotiating per applicant.
6. The payout threshold: set low enough that a small affiliate can reach it, on a hold period matched to your returns window.

The reason to do this now rather than later is the same for all six.

Every one of them is a free choice today and a renegotiation with real people once affiliates have published links under the old rules.

If you have not set the program up yet, starting an affiliate program on Shopify covers the ground before these decisions.

You can install Affilitrak free and set all six of these before you send the first invitation.