Was browsing a post for a paid search engine just before - the pricing model was USD $ /mth.<p>I didn't bother to investigate further. There is zero chance I will ever pay for search via subscription. The amount is not the problem - it's the fact of being a recurring subscription.<p>Why don't more services offer a "pay as you go” pricing model?<p>As a customer, I like the "pre-paid" or “top-up” model; where you use a credit card to charge your account with an amount, then the product is charged in a “micro-transaction” pricing structure, like 5c per search or something (the amount isn’t important to this post). But I also don't mind the "post-paid" model where I get a bill at the end of the month that varies with usage.<p>What are the downsides of this from the business perspective?<p>Obviously, usage tracking/billing is more complex and you need to deal with the "excess usage" problem, but I'm wondering about non-obvious stuff like:<p><pre><code> • the business has varying/unpredictable liability to the customer base
• most customers don't actually like it - they want the certainty of a monthly bill - especially fearful of "blowout" bills
• customers don't like it because of the perception that the business is incentivized toward inefficiency - e.g. use dark patterns to require the user to do two searches instead of one
</code></pre>
I have experience as a customer of pay as you go pricing models - and I personally like it for many different reasons.<p>If I wanted to use a “pay as you go” pricing model for my own service - what problems should I be prepared to deal with?
I work for a B2B business that offers a pay as you go model.<p>It hasn't been a complete disaster... but I would now strongly recommend against it.<p>- Most customers want a fixed bill for budgeting purposes. They would often rather overpay a little bit than having to set aside a special budget that they may or may not need any given year.<p>- It will make forecasting <i>your</i> revenue a real SoB. For a bigger company you may be able to annualize the payments and deal with liquidity crunches. But imagine being a smaller company and going through months at a time when you aren't getting any money in because of the cyclical nature of demand.<p>- Customers feel punished for using the product more rather than "getting more value" out of it.<p>- It's extremely punishing for your biggest customers. 10% of your customers will use 90% of your customer resources. Making them basically foot the entire bill for your company will make them not happy. You will have to find a way to cut them their own pricing or they will go to your competitor (or worse, stand up their own competitor instead of paying you).
> What are the downsides of this from the business perspective?<p>I can’t answer for the specific service you’re thinking of, but I can answer for the general SaaS case. Even assuming the service has a meter-able metric, there’s two big problems:<p>1. Often, the cost to provide the service is not linear with usage. One moderately complex support ticket might cost the vendor more than the variable cost to handle a small customer’s usage for a whole year.<p>Imagine a new customer needs a support ticket or has a pre-sales product question, then only uses $5 of services in the next year. Of course, there are ways to address this - charge for support, charge a monthly minimum - but these have other tradeoffs.<p>2. The value received by the customer is almost never linear with usage. If a customer needs to send one e-fax a month, it’s worth more than $0.50 to them, even though a customer who sends 200 faxes per month might only pay $0.50 [1].<p>This also extends to competition. A customer doing a very low volume of something is comparing the price to a different alternative than someone doing a larger volume of it. At very low volume, the competition might be driving to a print/copy store (and wasting $5 or even $20 worth of time). At 200 faxes, the competition might be a $200 fax machine.<p>A vendor could address this with per-unit pricing discounts, but the scale can look offensive: $5 for your first fax, $2.50 each for your next 5, $1 each for the next 10, $0.50 for the rest.<p>In a sense, a small SaaS plan includes enough units that the effective per-unit price is at least reasonably comparable with larger plans, and not offensive. In the example above, a plan with 5 faxes for $10/month doesn’t mean you’d be able to buy qty 1 for $2.<p>[1]: All prices are imaginary. I haven’t sent a fax in ages.
I've actually been hearing more recently about SaaS companies moving to usage based instead of subscription based models. Things like AWS are already like this, and it does make sense for a lot of businesses that you can quantify the usage. Search is a good example. Sorry I don't have a link, there's an article I read that lays out the case, from the business' perspective of why it makes sense to price it that way.<p>Edit: <a href="https://techcrunch.com/2021/01/29/subscription-based-pricing-is-dead-smart-saas-companies-are-shifting-to-usage-based-models/" rel="nofollow">https://techcrunch.com/2021/01/29/subscription-based-pricing...</a>
IMO it's not offered more because it's not user friendly to most customers; they have to do more math to figure out what their costs are (not as transparent, takes a little more effort to know what monthly cost will be, what if usage is not known, etc.)<p>If a company offers metered pricing like this and a competitor offers a similar functionality at a flat monthly rate, the latter is usually perceived as the more user friendly experience.<p>I think in some instances it makes sense but not always. There was an essay someone linked on HN a few weeks or months ago that was written a while back which touched on this exact issue... on the wrong computer and can't find it rn
It makes sense to align your costs methodology with your income methodology.<p>In other words most businesses have figured out that their big costs (salaries, rent etc) are "subscriptions" and so by making their income into subscriptions there are few "surprises." Income becomes predictible, costs are predictable, and so on.<p>This ultimately makes business a lot less stressful. For everyone.<p>Annecdotally I play golf. I pay an annual membership at a club which includes "unlimited golf". In other words I pay a nominal $2 per round to play, but a big lump sum annually to be a member.<p>The actual marginal cost of my round to the club approaches 0. Their costs are fixed, so it makes sense for their income to be (mostly) fixed.<p>Incidentally paying once upfront means I actually play more. Even if weather is not ideal I can play because the marginal cost is so low.<p>Aside: this really helped the club during covid, no one was allowed to play, but the bulk of the income was assured. Other sources like bar, visitors etc went to zero, but the base-line meant the club survived better than some other clubs. No jobs were lost.<p>Every business is different though, and while I don't recommend pay-as-you-go only, perhaps offer both options. Then keep track of which group is more popular, and which is working best for you.
> If I wanted to use a “pay as you go” pricing model for my own service - what problems should I be prepared to deal with?<p>The big one is convincing people that the HTTP request they're about to make is going to be worth the price. That's friction that will lead to hesitation, and that hesitation may lead to users never using your product more than a few times.<p>In the specific case of search, users don't and can't know whether pressing submit will be worth whatever number of pennies it costs. If that's the case with your SaaS I think pay as you go might be a bad choice.
The calculus of setting up a billing/pricing system against the potential incremental revenue (people who will pay per usage but not subscription) probably moves the endeavour to 'we'll do it, but not today' pile<p>Having tiered features not tied to usage skews the calculus even more
Maybe people simply have different spending habits.
Personally I try to stay away from "pay as you go" businesses. Moreover, I try to pay for the year upfront if I can help it.<p>Purely psycologicaly I think it's an attempt to take control of your future, you know? Reducing uncertainty is more valuable then many things...
"The best way to predict the future is to make it" and all that.<p>Like, if I'm not sure whether I'll need it and how much of it I'll need, there is something wrong and I'm not gonna buy it at all.