One of the things I get to do at Nexcess is influence when & how much we discount our products. There are some obvious dates like Black Friday and the New Year. But there are also less obvious dates like the end of summer, Halloween, and WordPress' birthday.

Should we put our products on sale? And if so, by how much?

Now there's a lot that going to a promotional pricing strategy that we're not going to get into. But there is one very important concept that we are going to look into: subsidy costs.

What are Subsidy Costs

Your business incurs Subsidy costs when you spending promotion dollars to create a transaction, when the customer would have bought anyway.

So let me give you an example. If you own coffeeshop and you try to get new business in the door by offering a 50% off coupon your daily customers will get the 50% off and you incur a subsidy cost.

Good & Bad Subsidy Costs

Subsidy costs aren't necessarily bad. Going back to the coffee shop if you offer 50% off coupon and you get 4X the usual customers you still double your revenue. I'd take that deal every day.

But if you don't get enough sales to offset your subsidy costs you can actually lose money. That's why with a lot of ecommerce stores they'll limit coupons to new users, order minimums, or specific items.

Subsidy Costs and Hosting

So how does all of this relate to hosting? Well, I'm trying to measure subsidy costs. And unfortunately, I can't share internal data. But I can share some of my own.

One of my favorite coffee shops in Denver is Bardo. They have good coffee and good treats so I used to go there pretty regularly… until one day I saw an offer.

I received this offer from Bardo after not visiting them in ~3 months. And if you haven't seen one of these campaigns they're usually called winback campaigns or if you have multiple tiers it might be a discount ladder.

Once I saw this offer I went back to Bardo and used the code and got the fantastic offer. But then I knew the offer was out there. So tested waiting another 3 months. And guess what? Another offer.

So now I know that if I wait for 3 months I can get really cheap delicious coffee. And I was curious how that affected my attendance. And since I'm an absolute data nerd I “check-in” wherever I go with Swarm (formerly Foursquare).

I copied all of these check-ins into Google Sheets and grouped them by year.

The data is a little confusing because I moved in 2016 to a different part of town and wasn't quite as close to Bardo as I was before. I received my first winback email in 2018. And those codes last ~1 week. And if I was in town that week I'd go to Bardo to take advantage of the offer.

But if you notice: I've never been to Bardo more than 3 times a year since receiving this offer. And even in 2017 when I moved away I still visited to the store (without any promotion) and I visited once in 2018 without any promotion.

So let me ask you this is this winback campaign useful? Or are they giving away too much in subsidy costs?

Let's Do the Math

I'm assuming I'd visit Bardo 1-2 times a year without any sort of promotion. But to be fair to Bardo I'm going to round it down to 1 time a year.

So Bardo is giving away 50% and they're doubling my attendance. The campaign is a wash at best. And at worst they're limiting my attendance to once every 3 months.

And we're not doing a truly fair comparison because we should be looking at net profit instead of revenue. Basically, if I spend $10 per order. Probably $2 of that is the cost of goods. The coffee beans, the croissant, and the napkins. So the net profit of the order is $8.00.

With the discount, I pay $5.00 (revenue) and after taking into account the cost of goods the store owner makes $3.00. That's less than they'd make if I just walked into the shop 1-2 times a year. The subsidy costs outweigh the additional revenue.

Calculating Subsidy Costs is Hard

It's incredibly hard to calculate subsidy costs because you have to know which of the people ordered with a promotional code would have purchased anyway. And we're doing some cool stuff at Nexcess to figure that out.

What I can say is that lots of platforms say their winback campaigns generated revenue. And I'm sure Bardo has some reports that say these campaigns are amazing and they're generating hundreds of dollars. But when you dig beneath the surface you might find you're giving away more money than you need to.