Hacker News Reader

Comment by jakobegger | Hacker News Reader

Sorry if this is a bit off topic, but I'm getting so tired of all those posts about how to grab the maximum amount of money from your customers. When I started selling my software, I read a lot of these articles, and was constantly stressed that I am leaving money on the table.

But at some point I realised that it doesn't matter. You don't need to make the maximum possible amount of revenue.

If you sell your product for a lower price, you allow more people to use it.

Sure, make your product cheap and you might leave money on the table with big enterprise customers. But if it's cheap, maybe a school can afford it.

That must be worth something too, right?

I'm not saying that you should give away your work. But once you make enough profit, why not try to maximise the positive impact your software has on the world?

View on HN

We're a B2C selling to students, and for us it's not just about maximizing profit or not leaving money on the table.

We definitely want our service to be available to lots of students and to help them learn. But pricing it too low can not only leave money on the table, but actually make our product less valuable. Why? When you pay for something, you want to make the most of it. So our students have an even bigger incentive to keep learning. If it's too cheap, it loses its value, and in a way becomes less useful.

So we try to tier prices and offer a much lower price in developing countries for example. But we still want to find a "sweet-spot" for our main customer base, so people feel that they get value, but also that it's not too cheap so they stop caring about it.

Pricing experiments aren't always evil and the motivation isn't always to grab as much money as possible. Plus, don't forget that you can maximize profits by lowering prices in lots of cases. But you need to be able to test it to find out...

Have you done any pricing experiments? I've been tossing around some ideas about how to optimize pricing through machine learning, but I haven't been able to validate the idea at all.

It would look something like:

1. Use historical price data to determine the optimal price for each customer given what we know about them. 2. For a small number of customers (<10%), randomly adjust prices slightly to find additional data.

Once we have the data, we could find the price elasticity for customers (elasticity = % price increase / % change in purchases; so if elasticity is > 1, it makes sense to increase prices, if elasticity is < 1, you should decrease it), and use that to increase prices.

We did very limited experiments so far, mainly due to the fact that we feel it's unfair to offer different prices at the same time on the website. (And even if we didn't, some users might figure it out and/or complain).

We did a few experiments via email, where we offered a different discount to different cohorts. For example, some people will get an offer for 20% discount, and others for 40%. We tried to extrapolate pricing elasticity from that, but it's still hard... We have a few ideas for other experiments, but didn't get round to it yet. It's one of the areas we need to be more careful.

I'd be happy to talk to you and try to understand what you have in mind. (I didn't quite). Hope it's ok to contact via the contact details on your profile? (Or feel free to reach out to me.).

I'll shoot you an email.

If I was a solo SaaS founder, I'd rather have 100 customers paying me $100/month apiece than 1,000 customers paying me $10/month.

Your time is worth something, and customer support is a very real burden. In the case of a $10/month user, answering one customer support email more than wipes out all of the revenue you've earned from them that month.

Is there a sweet spot for customers?

1 customer is a high risk.

10 are better.

But 1000 are Bad again?

It's not the number of customers that really matters but the value per customer. It just happens that if you fix the total revenue for comparison, they're inversely proportional.

higher customers = less volatile risk of churn, but the volatility in churn rate for 100 customers is probably about the same as the churn rate for 1000 customers, but 10x the support burden.

Hm, then how is it that companies like Google have millions of users and virtually no support?

If you are overwhelmed by support requests, one way to fix it is to charge more. Another way is to solve the underlying problem.

If only one in 100 customers sends a support request, you can spend an hour to answer it, and you still made a lot of money even if each user only pays $10.

> If you are overwhelmed by support requests, one way to fix it is to charge more. Another way is to solve the underlying problem.

These are not mutually exclusive choices.

If you haven't already, give the article linked below a read. The section that I highlighted seems to fit pretty squarely with your point.

"As Douglas Rushkoff says, we need a new operating system for startups. The current one will keep producing the same extractive and monopolistic empires we’ve gotten so far. No, what we need is a new crop of companies that are institutionally comfortable with leaving money on the table. Leaving growth on the table. Leaving some conveniences and some progress on the board, in order to lead the world into a better direction."

https://m.signalvnoise.com/exponential-growth-devours-and-co...

In the end a startup is still a business, and if you don't make more money, you answer to investors, or you start losing money and then shut down, which makes things worse.

Having said that, I do think that trying to min-max pricing is quite a waste of time and resources, assuming you're already at a good place for pricing, and like you mention, are already gaining a decent profit. You'd probably be better off spending developer time and resources into satisfying your customers' needs, performance improvements, instead of extracting the most profit out of them. It's easier to tweak pricing once you build out more value for your customers, instead of paying a bunch of people silicon valley salaries to figure out if 25$ per user vs 26$ per user is better.

Reply on news.ycombinator.com