The Ecomm Analyst

A friend who runs a beauty brand asked me last month whether she should be investing more in SMS. Her email list was performing well. Her SMS list was smaller and she wasn’t sure if it was worth the cost per send.

The right answer is almost always “it depends,” but in her case I had data to point at. I’d recently pulled the same per-subscriber economics across three of my client accounts in ThoughtMetric, and the patterns were instructive.

Brand A is a supplement company with a 90-day reorder cycle. Email revenue per subscriber, annualized, came in around $4.20. SMS came in around $9.80. SMS wins easily, mostly because the messages are short, urgent, and well-suited to “your reorder window is open” prompts.

Brand B sells home goods with a long consideration cycle and an average order value over $200. Email landed at $11.40 per subscriber annualized. SMS landed at $6.20. Email wins, and it isn’t close. The product needs explanation, photos, multiple touches before a purchase. SMS is too compressed a format to do that work.

Brand C is apparel, mid-AOV, frequent new drops. Email at $7.10. SMS at $7.40. Functionally a tie. Both channels work because the brand has new things to talk about constantly, and customers want to hear about them quickly.

The lesson, which my friend ended up applying, is that channel performance is mostly a function of what you sell and how customers buy it. If the product needs context, email wins. If the product needs urgency, SMS wins. If both, run both and don’t expect one to dominate.

A few caveats from doing this analysis a half-dozen times now.

Per-subscriber revenue is the right metric, not total revenue. SMS lists are almost always smaller, so they look worse on absolute numbers and better on per-subscriber numbers. Looking at totals will mislead you in both directions.

Send frequency matters more than people admit. SMS that goes from 4 sends a month to 8 sends a month does not double in revenue. It usually grows maybe 30% and triples in unsubscribe rate. The same dynamic exists on email but it’s slower and less visible.

Attribution windows matter even more. SMS gets credit for the click that closed, almost always, because it lands in the moment. If you’re looking at last-click numbers, SMS will look better than it really is and email will look worse. A blended attribution view tends to flatten the difference between them, which is closer to the truth.

For my friend, the answer ended up being to invest in growing the SMS list, but not to shift email budget to do it. The economics didn’t justify a tradeoff. They justified addition.