4 May 2021
Over the past couple months, I’ve spent several hundred bucks on sponsoring newsletters. It’s an interesting format. Instead of targeting via algorithm (as in FB/Google ads), you target by picking newsletters with relevant topics. I kind of like it.
The downside is performance: the ads were quite hit and miss for me. In a couple cases I had a cost-per-acquisition (i.e. Sample subscriptions) below $2, which I consider good. But in most cases, CPA was more like $20. (In fairness, Google and FB ads performed even worse).
There’s also some weirdness because publishers typically charge a flat rate. They give you the subscriber count, open rate, and click-through rate, but usually not the click-through rate for previous ads. I don’t blame them; it’s a simple way to charge, and maybe they don’t even have metrics for individual links set up. But it does mean that, after dropping $200 and getting mostly nada, one can’t help but think was my ad the problem, or do all their ads perform that way?
it's rare to find a newsletter that charges per click, because underneath all the technology and engagement stats, the truth is email newsletter ads don't get clicked much.
—from the Fomo blog. (It has an infuriating lack of capital letters, but good content otherwise).
Ouch indeed. That reminded me of this Paul Graham quote, from the non-descriptively titled 6,631,372 essay:
In 1998, advertisers were overpaying enormously for ads on web sites. In 1998, if advertisers paid the maximum that traffic was worth to them, Yahoo's revenues would have decreased.
I think about that often.
This is also a relevant topic for me because I’d like to make a living by selling newsletter ads, preferably without customers overpaying being a requirement for profitability. At this stage, The Sample’s ad system is there mainly so I can measure ad click-through rate (it’s about 0.5%) and so it’ll be in place when we have enough subscribers to make ads worth thinking about. At first I did the usual pricing strategy of charging a flat rate (since that’s what I saw everyone else doing), but after reading the blog post above I’ve changed it to $0.75 per click. Nice and transparent, and prepaid for simplicity.
So for 15 bucks we give you 20 clicks. Given the ad CTR of 0.5% and the way the algorithm works*, that means we’d need… 8,000 subscribers before we can service a single ad campaign. We’re at about 250 subscribers right now, so I, uh, suppose we’ve got a ways to go. On the bright side, we’ll probably be much more efficient by the time we get to that size. In the mean time, if you’ve had a hankering for small, slow, cheap marketing, do give us a try.
*In order to keep ads somewhat targeted, we don’t send any single ad to more than 50% of the subscriber base.
Side note: my brother has officially graduated and joined me as a partner, so now when I say “us,” I mean it in the literal sense, not in the sense of “us, as in the organization I belong to, of which I am the only member.”
Zooming out a bit, it would be cool if paid advertising was efficient enough that any product could be bootstrapped with it—as long as the product is good. “If you build it, they will come,” you know. That’s part of what I’d like to achieve.