A product launch is the easiest time to waste a marketing budget without noticing. Everything happens at once: email, paid social, a few creators, a homepage takeover. Sales go up. Everyone is pleased. Then someone asks which of those things actually drove the sales, and the meeting turns into a debate about opinions.
The fix isn't a fancier attribution model. It's deciding, before launch, what you're going to measure and what would count as success.
Five decisions to make first
1
What is this launch for?
New customers, revenue from existing customers, or awareness for a new category? They're all reasonable goals, but they lead to different measurements, and a launch that tries to do all three will be judged on whichever one looked best afterwards.
2
Which one number decides success?
Pick a single primary metric (for example, first-time orders that include the new product) and write down a target. You can watch other numbers too, but this is the one you'll report.
3
What's the comparison?
"Sales went up" needs a "compared with what". Use the same weeks last year, the four weeks before launch, or a previous launch. Choose before you see the results, so the comparison can't be picked to flatter them.
4
How will each channel be tagged?
Every link in every channel gets a tag that follows one agreed convention. This is the only way to separate email from paid social from creators once the orders arrive.
5
When will you read the results?
Set the review date in advance, after enough time for the slower channels to show their effect. Launch week alone usually flatters email and underrates everything else.
A worked example
Here's what this looks like written down for a hypothetical launch: a skincare brand adding a new cleanser.
Goal
Bring in first-time customers
Metric
New-customer orders containing the cleanser
Baseline
Same four weeks of the previous launch
Review
Six weeks after launch day
That's one page, agreed by everyone involved before a single post goes out. It doesn't need to be more sophisticated than that.
Don't let the review date drift
The most common way a launch review goes wrong is reading it early because the first week looked great, or late because it didn't. Either way, the result becomes whatever is convenient. Put the review in the calendar when you write the plan.
Why bother?
Because the next launch is coming, and the most valuable thing this one can give you is a clear answer about what to repeat and what to drop. Without the plan, you get a feeling. With it, you get a decision, and usually a budget that works harder the second time.
