I have a better idea for you, which has worked for several others, e.g. Dropbox.

Let your early adopters earn free service by referring others.

Instead of alienating your biggest fans and supporters and power users, you'll be turning them into promoters and marketers.

It's a good idea, and one that I've run during the closed beta. A handful of the users have a referral credit for referring other beta users, and once we leave the closed beta, their credit will go into effect. I plan on extending the referral program after the launch as well.

GP has already gotten a lot of value out of them, so if keeping them isn't costing much, I'd say grandfather them in for free.

Why for free? Give, like, a 50% discount. Their "free stuff" would be "50% for free".

Or even 90%. Maybe 75% as a compromise. Heck, being grandfathered into a very low price point will make them much less likely to quit even if they no longer use the product, thanks to FOMO. So you may see long-tail income from those early adopters with little expense.

Or a 5%-15% forever discount per referral that's paid for a year. After 6-20 referrals, you're on us - it's cool.

I bought ifttt pro subscription when they were offering "pay any price". I went for the lowest possible - and whilst I don't use the tool nearly enough to justify the pro at full price, I can't see myself ever cancelling the subscription.

As the other person said, if you offer a grandfather rate - you might have people who are signed up for life!

I don’t think you’re “bumping against the same “free” forces somehow”, because you seem to have given away the whole pizza for free for 6 months.

I think you need to study companies that transitioned from a free to a paid product rather than compare yourself against businesses that increase their regular sales through free samples.

To add some additional context, the closed beta users knew from the beginning that the product would be paid after the closed beta. Some even asked about the price, and I gave them a ballpark figure, and all of the responses to the figure were positive. However, once I made the official price announcement, a few people were pretty angry (ironically, some were people who asked about the price and were ok with it at the time).

So I'm not necessarily moving from a free product to a paid product. It's from a paid product that was subsidized, to a paid product that is not subsidized. But to your point, I think the closed beta ran too long and shifted their psychology about the cost of the product.

What if during the sign up stage, you actually show the current price of the free plan. But then give a 100% discount for a certain period. Afterwards, you would then reduce the discount X percentage until it's fully transitioned into a paid model.

If it's all transparent, maybe that would make it easier for people to know that it's not free for free sake and plan accordingly.

That necessitates actually having a price point nailed down prior to accepting any sign ups. Some people (including possibly OP) get the ball rolling without even having a reasonably narrow ballpark guess settled on.

Grandfather early supporers, not at free but a healthy discount.

Not a bad idea, but I'm just going to ask that they pay the normal price in a few weeks. If my product really is game-changing for them, they'll pay it. If they won't, then a discount is only prolonging the inevitable. My current price is really the most competitive price that I can offer, so it's take it or leave it.

The goodwill of your testers is likely worth far more than the marginal revenue you’d gain from having them pay full price.

I was in a similar position many years back and learned the hard way that people _absolutely loathe_ when you take things away from them.

I ran into this as well. I had an attempted launch a month ago with a different pricing model, and people were pissed. They loved the product so much that they were angry at the prospect of having to choose to not use it (because they didn't like the price). I've since adjusted the model to fit all of the feedback that I gathered from that experience, and I've sampled the users to get their feedback on the new pricing model, and it seems positive, so hopefully there aren't complaints this time around.

Many of the testers do have free service (from referrals) that they'll get. But after that, they have to pay. I'm feeling pretty strict about it this time around.

Yeah I can't fathom why you are doing this.

If the marginal cost is zero (i.e. this is software) they are worth more to you than any kind of incremental fee (i.e. assuming a few dollars a month, not thousands).

It's just one of the most fundamental rules of business that increasing prices is one of the most disastrous things you can do where there are expectations.

It's why there are granfather contracts all over the place.

Make your new customers delighted with your product at 'some price' and your 'alpha/beta' people feel empowered that they get it for free. You can sell them addons later.

The marginal cost isn't zero unfortunately. Each user is actually expensive, and their cost is unbounded (I send hundreds to thousands of SMS messages per user, per month, on their behalf). They've only gotten more expensive as they've used the service, because they're sending more messages to more people. There's theoretically no limit to how much they could cost me. So I just can't do the grandfather idea.

I'm not willing to spend the effort to build a custom billing system and pricing model just to grandfather these users in.

How much does 1000 SMS messages cost? If it's actually significant I would grandfather them in at a lower rate per X number of messages and offer further discounts for referrals.

Most services I've seen in Europe are around €0.09 so "thousands" would mean upwards of €100. I've seen US providers go as low as $0.005 though.

I think the mistake is to have unlimited messages if messages are such a major cost factor. Most services with "unlimited X" have a soft and hard limit on X if you dig deeper. If you don't want to charge per usage (or for overages), you should do a mixed calculation based on typical and min/max usage per customer and how usage ebbs and flows over time.

For the grandfather plan I'd aim lower than that even if it means losing a little money on high use customers because their usage might decline in the future (at which point they're providing profit while being less likely to cancel because of FOMO) and because the lower price point and legacy will make them good multipliers, not just for direct referrals but also as references or partners for whitepapers.

Basically you can get more out of them than just money so treating them the same as later customers means missing out on that.

"How to build SaaS that doesn't punish the user for using it"

Can you be more clear what you mean by this? Are you saying I'm punishing my users for requiring that they pay for the service?

Not clear what you are arguing. Early supporters should get a price something like cost+3%, new customers might get cost+30% or more.

I wasn’t, and don’t think other’s were advocating losing money on the originals.

Making them angry when they were on your side and helped you off the ground is a silly policy. The negative word of mouth factor alone could be incalculable and do unbounded harm, all for ~$100 bucks a month?

Treat them right and word of mouth will make much more than that in the long run.

Oh, in the case where there are material underlying costs, it should be a lot easier to charge. People 'get' that. You can given them a discount and that's that.

Is there just 1 tier? If the only 2 reference points are "free" and $X then $X is going to seem expensive by comparison, but if you add 2 tiers above it, suddenly it's "very affordable".

It's a pay as you go model. There's a small base fee, plus costs that scale as you use it. Originally, I had a 2 tier model, for basic and premium. But the feedback that I got from the beta users was that it wasn't granular enough. People who used it less wanted to pay less than the tiers could offer. So I switched to the pay as you go model.

I have no experience with setting up pricing for stuff, but I feel like it could have helped if you already had pricing information in place before your free beta, even if it was just an estimate. Make it clear it won't stay free forever, you know?

That's one thing I could have done better. They did know that it would eventually be paid, but I only gave the ballpark figure to those who asked, not everyone.

Make your product freemium. You could even gradually move features into paid plans.

It won't work for my product. The costs per user are too high.