Ask HN: Is $2.95/month too cheap for a B2C SaaS?

[egmike]

I'm looking at pricing for a new B2C-based SaaS, and I'm wondering if going with $2.95 or $3 per month would be considered "too cheap." Does anyone have any experience with this pricing that they'd care to share?

The reason why I'm asking is because the first time I saw a $2.95/month price plan a few years ago, I thought to myself "that's too cheap, they must be skimping on something." It was for JungleDisk (an online backup service).

I now know companies like JungleDisk (who use the Cloud) can afford to do this, so I don't think it bothers me that much anymore. But in your opinion/experience, when it comes to potentially subscribing to a new service, do you feel that some prices are "too cheap" and shy away from signing up? Along the same lines, do you think customers who don't know about the Cloud would avoid signing up if the pricing looks "too cheap?"

[ig1]

Make it an annual sub if you think $3/month sounds too cheap. But in general pricing is one of those things you only learn by trial-and-error, so stick up an A/B test and see what converts better.

edit: Also offer multiple levels of pricing, offer a "premium" product with a few extra bells and whistles for people who want to pay more.

[byoung2]

Definitely follow the above advice and A/B test to see which one converts. I personally don't think $3/mo is too cheap for B2C (e.g. Skype has a $2.95/mo plan). Even B2B can be pretty cheap (when I was just using Amazon S3 I had bills of just a few pennies).

[egmike]

Thanks you two, that's good advice... and I'll definitely be offering multiple packages for premium usage. :)

[bradly]

I'm not sure whether $3 is too cheap, but I've noticed a few sites use ~$25 annual subscriptions.

http://flickr.com/ http://backpackinglight.com/

I sure there are many others.

[egmike]

Annual subscriptions are a good idea, thank you, and thanks for the references.

[dpapathanasiou]

It's hard to start low and then raise the price, w/o a substantial increase in the number of features.

Psychologically, a better approach is to start high (i.e., at the highest per month you think people will pay), and only drop the price over time if you're not getting enough subscribers.

If you do stay in the $2, $3 per month range, though, then definitely make it one charge of $24 to $36 per year.

[egmike]

Thanks for your response. Hmm... see, I used to think starting with the higher price was better, but now I think it definitely depends on the feature set at the beginning.

Right now I'm planning on launching it mainly as a Minimum Viable Product, so I'm thinking of starting at the lower price and raising it as features are added. I think that waiting to see if people buy at the higher price may take too long, since I'll be trying to validate whether people are willing to pay for the product at all.

For the yearly price, do you mean remove the monthly option entirely and replace it with the yearly option? Are people more likely to spend $24 or $36 upfront than $3/month? I was planning on offering a yearly option at a discount...

[dpapathanasiou]

I used to think starting with the higher price was better, but now I think it definitely depends on the feature set at the beginning.

It's more subtle than that: you need to discover what the current features are worth to people.

You might be surprised to find that value is higher than what you would think, regardless of the number of features.

It's possible that even one (great) feature is something you can charge a lot for.

As for the psychology of it, people are ok with paying less for something that used to cost more, but the opposite situation can create resentment.

Are people more likely to spend $24 or $36 upfront than $3/month?

There's some research I've read about the psychology behind the failure of web-based micropayment systems you should read (I can't find the reference now, but I think Shirky wrote it).

Anyway, the gist of it was: at too low a price (e.g., $1, $2) people just don't bother getting out their credit card to pay for something.

So you're much better off bundling it in a larger package, or in your case, for a longer time period.

[egmike]

It's more subtle than that: you need to discover what the current features are worth to people. You might be surprised to find that value is higher than what you would think, regardless of the number of features... As for the psychology of it, people are ok with paying less for something that used to cost more, but the opposite situation can create resentment.

Very, very good points, thank you again! Wow, it looks like I've got some thinking to do.

Anyway, the gist of it was: at too low a price (e.g., $1, $2) people just don't bother getting out their credit card to pay for something.

Hmm... interesting. Thanks, I'll look into this.

[sebg]

I would think it depends not so much on whether your users can afford it, but whether you can sustain development on $2.95/month. The price difference between 0 and 0.01 is huge, but between $2.95 to $8.95 is minimal. If you are going to charge for it, charge for it so you can keep supporting it.

[egmike]

Also a very good point, thank you for that! :)

[singer]

I wouldn't worry too much about people thinking the price is too cheap. I'd wonder how many of those people paying $2.95/month would also be willing to pay $9.95/month, $19.95/month, etc.

[egmike]

Very good point, thank you. Yes, I'll definitely be testing the pricing, but since it's a new service I'm trying to avoid people not subscribing for silly reasons like "it's too cheap" or something.

[TomK32]

Offer it for $36/year, as a customer I'd go nuts about receiving a monthly bill about a lump sum. And offer a premium version with a little extra and quicker support for $36/MONTH :)

[egmike]

Thanks for that, they'll definitely be additional plans! :)

[sirwitti]

i think that depends on your product. what is your target group? is there a pattern on how wealthy this group is?

is the service very specialized?

when a/b testing, interpreting the data correctly is key.

take your time and try stuff...

[egmike]

Thanks for your response. I actually don't know how wealthy the group is/will be because I haven't actually figured out the target just yet. I have an idea of who the product will benefit, so I'll be trying to reach them and see how they like it. :)

I'll certainly be doing a/b testing, but could you elaborate a bit on interpreting the data correctly please? I know I shouldn't just look at the numbers, but I get the impression that you may have some specifics on interpreting the data that I'd love to hear.