Comment by zloof | Hacker News Reader

Comment by zloof | Hacker News Reader

Parent Full thread

Short answer - we're starting to experiment with trial options. This is a link to a 2-cup trial that converts to a sub: https://www.suddencoffee.com/?trial=true

Longer answer - sort of but not that simple in our current state. Even if you absolutely love a product, often times you don't buy again. Many reasons for this - it's a pain to checkout again, you don't have your credit card on you, you forgot about it, you're just too busy with other stuff.

If we had an Uber-like app - one click to order, saves your credit card, sends you push notifications to remind you - it would be a lot more feasible to send you a single cup of coffee and make it really really easy for you to buy again. Unfortunately, it's hard for us to do that right now and we end up losing a lot of people who truly did love the product.

Before LTV or any of that stuff - the big question is if we call up our subscribers, are they happy? We hear again and again "Yes, I love the product, I love that it shows up, it tastes great and saves me time". We would not be building a subscription service if that were not the case.

However, we are absolutely testing different trial options to get at what you're talking about.

This is a product where clearly some kind of free sample is necessary, because when you make claims like you are about your coffee, people want to try it before subscribing to it. If you're giving free samples, you need to collect credit card information to avoid fraud anyway, so you may as well add the customer to a subscription after 30 days unless they opt out. The rebill model is profitable for a reason.

I would go so far as to say you should create an affiliate program. I know the "rebill" affiliate marketing model was quite scummy (acai berry, teeth whitening back in 2008-2011), but it worked very well and doesn't have to be so scummy. The problem then was that the major products were hiding the terms of the rebill and made it impossible to cancel without multiple angry calls to callcenters. But it doesn't have to be this way. Lots of reputable products have rebill models, but they are clear about the terms when you enter your credit card info, and you can opt-out of auto-renew with a button click at any time in the next 30 days. If you can do this you can capture the advantages of the rebill model without high chargeback rates and reputation damage. The rebill model will increase customer LTV and enable you to pay competitive commissions to affiliates who drive signups.

Once you know the percentage of customers who convert from free trial, you can start measuring the average length of subscription to calculate LTV (if churn is negative the average LTV will increase over time). Then you can use that LTV to determine maximum customer acquisition cost (CAC). With that, you can launch an affiliate program (by registering as an advertiser on the few big affiliate networks) and pay a commission as some percentage of customer LTV (hopefully around $20 if you want to compete with other affiliate offers). As LTV increases you can raise commission for your best affiliates.

This is the kind of product that affiliates can sell very efficiently. You should let them do your marketing for you, and use the funding you're raising to bankroll your affiliate program.

I think there are ways you can get repeat purchases without the subscription model. Yes, having the CC details stored (actually, having some kind of token stored, ala Stipe) would make repeat purchases easier.

There are probably other things you could do to speed up the process. Maybe a sticker (or mug) with a QR code that leads to the purchase screen. Something that's always in front of my face that I can scan when I realise I'm down to my last cup.

Subscriptions are fine, and there's no way you should drop that model, but having both options would be way better.

Reply on news.ycombinator.com