Kevin Hillstrom wrote a recent article about measuring your marketing:
If you take away a marketing activity and sales don't change, you know that there was no power in that marketing effort.
By and large, most marketing campaigns cannibalize each other ... rendering each other meaningless. You're free to disagree with me.
Since the point of marketing should be to generate interest and revenue, if sales don't materialize or if you can remove the marketing without any impact on sales, then that marketing isn't working.
Marketing that isn't working should be thrown away.
Measuring the impact of marketing is difficult (like getting any kind of attribution for a Shopify app is a herculean project). It's easier to just throw money at marketing and hope it pays off. That leads to many Shopify stores getting stuck paying five or six-figures to Google every month while praying they don't get their ads shutoff.
Instead you can use testing to understand what works. Kevin mentions using holdout tests (A/B tests for mailing catalogs or not) but not every channel supports that (e.g. social media). The ones that do make it difficult to setup, e.g. built-in A/B tests in email marketing is more concerned with vanity metrics like open rate than sales.
Even though it's hard, figuring out what marketing doesn't work could have a significant payoff. What if half of your marketing was just a waste and you were able to cut out that half? That could let you rebudget towards the working channels or just cut acquisition costs in half.
Being able to know which channels work and how well they work is the point of Repeat Customer Insights' Acquisition Source tracking. Knowing that one sales channel produces much better customers than another is one of the best way to measure your overall marketing efficiency.
Add in holdout or A/B testing and you'll also able to measure individual campaigns and offers.
Eric Davis
Did last year's holiday customers come back?
Find out if last year's holiday customers stuck around with Repeat Customer Insights Cohort Report.