Attribution is a billion dollar marketing problem. When done right, you'll know where customers are coming from and be able to avoid wasting marketing money.
The problem is that attribution is nowhere near perfect and it likely never will be. The more you spend, the better attribution you can get but it might not be worth the extra cost.
Shooting for good-enough attribution is almost always more profitable.
Good-enough attribution involves two metric sources:
- where are people coming from?
- where are people first ordering from?
The first is easy online. Just look for the Landing Page in something like Google Analytics. If you can connect that to order amounts, great. If you can't and only have traffic and bounce rates, that's good-enough.
For non-online visitors, you'll have to resort to asking the customer. It's harder to get a lot of data but the quality will be higher. Try to classify them into standard responses (e.g. friend referral, online ad, paper mailer).
The second metric source involves looking at your orders per sales channel in Shopify but only the first order for each customer. A customer who orders from your POS first and then your online store should give credit to the POS (your physical store) for attracting the customer, not the online store.
(In Repeat Customer Insights this first order attribution is called the Acquisition Source and is a filter in a lot of reports and graphs)
Once you have those two metric sources collecting data, use the sales channel (#2) for your main attribution with the 1st metric source as a way to refine the sales channel. This will tell you something like e.g. your Shopify website is handling many of the orders and from that, a lot of people are coming from Pinterest.
To go beyond good-enough, go through your sales channels and look at the different metrics. Compare AOV, Customer LTV, and other measurements to see which sales channels are sending you your best customers. When I added this to my app I was surprised at the differences between sales channels.