Ask HN: Which business model - monthly recurring billing vs. pay-per-sale?

[fezzl]

Hi. We sell a hosted daily deal solution for online retailers, and we manage everything from marketing to the actual sales transactions of the deals. I was wondering which business model would make more sense for both our customers (the retailers) and ourselves.

We currently use a recurring billing model (with 30-day free trial), but we haven't been seeing good sales with that model. Should we switch? Make the monthly fee cheaper, or shorter/longer trial period? Or should we move to a pay-per-sale a.k.a. commission model? Or a combination of both, i.e. monthly billing + small cut for each sale?

No doubt, we can analyse which business model would be more profitable if we had a lot of data, but we don't. I was wondering if you guys have a preference based on gut feel, experience, or some reasoning.

Pay-per-sale seems to align our interests with those of our customers, thus making it a softer sell, but if we don't make much sales for our customers, our revenue will be pretty much stuck. Monthly recurring billing ensures predictability of revenue (allowing for churn), but would seem to leave a lot of money on the table if our customers get a substantial amount of sales out of our tool.

What do you think?

[ActVen]

In addition to discussing this with customers and a/b testing, I think you should consider the capital your firm has available. If you have sufficient capital to use a commission model, that lines up all of the incentives correctly. However, you should also pay attention to the cost to onboard the new customers. If this cost and/or time commitment is significant, you may have additional pressure on your bottom line during a time of high growth. You should also think about how many customers only conduct one of these deals and then drop off the map. If your cost to onboard is high then these customers will pose problems with your commission model. You might consider adding a commitment to a minimum number of events(deals) or price points if they want the commission model. That should help address the possibility of one time users becoming a liability.

[mgkimsal]

I hate to say "a/b test" but... "a/b test"? Or perhaps just test CTA pages to see which ones people click more? Or survey existing users to ask for their input?

I've wondered this on some projects, and the "a/b test" answer usually didn't feel right because in the beginning you're dealing with an insanely small amount of traffic (hard to gather meaningful results on tests from < 10 users, for example).

If you have some existing customers, send them a questionnaire and get answers directly from them, perhaps with a 'free ipod' drawing to sweeten the deal.

[fezzl → mgkimsal]

Thanks for your suggestion. We have sent out surveys, but still there is little discussion about this on the internet AFAIK. About the free iPod: lie or really give one out?

[mgkimsal → fezzl]

An ipod shuffle is $50 - with shipping and tax about $65? That expense shouldn't be that much.

[luis_ca → fezzl]

I hope you are not seriously considering lying about giving out the iPod.

[fezzl → luis_ca]

No. But the original comment said iPod "drawing"...

[kayhi]

Being in the ecommerce space and a possible end user of the product I would charge based on the size of the deal without a monthly fee. The idea being to reduce friction even more than the 30 day free trial.

For example: widget costs $20 dollars if 50 people buy and take a percentage of the total so 1% (I would a/b test the percent and possibly ad a flat rate fee per an offer) would be 10 dollars.

I think this would align your interests more with the end user. The more they use your product, the more they pay you.

[allwein]

> but if we don't make much sales for our customers, our revenue will be pretty much stuck

If you don't make much sales for your customers, they'll leave anyway.

The answer to your problem seems pretty clear to me: pay-per-sale. It minimizes the downside for your customers: If you make no sales, they pay no money. And it maximizes your upside: Sell 1000 units, make 1000 x fee.

[davidw]

Interesting question that I'm wondering about myself, for LiberWriter ( http://www.liberwriter.com ). It seems pretty natural to just sell it once, but it'd be nice to get money that kept rolling in somehow.

[petervandijck]

1. Actually talk to your customers.

2. a/b test

[pitdesi]

I'm assuming this is a high-touch sale. In that case, you are talking to your customers all the time and should be able to see why they opt out of your service. If indeed it's the recurring model, you should offer them the commission model and see how they bite. It also might have something to do with how it's presented - what is the site?

The other thing to consider is how much money you have to burn... commission model typically takes longer to make serious bucks.

[fezzl → pitdesi]

Actually, we're talking about low-touch sales -- as low-touch as low-touch can be. We charge USD9.95/mth, and we're at http://www.fezzl.com (redirects).