Hacker News new | past | comments | ask | show | jobs | submit login
How I killed app sales by going freemium (medium.com/shuveb)
207 points by shuss on June 7, 2015 | hide | past | favorite | 128 comments

I think part of the problem with freemium is that it changes your perception of where the value lies. The user discounts the value of the free component to nothing and only measures the value of the additional functionality that is unlocked with buying the premium portion. A lot of the time, this looks like a ripoff.

The developer on the other hand tends to look at the value of the whole application and prices the premium portion in consideration of this.

I'm not sure that I have a solution, except to say that the developer might be more succesful to discount the value of the free portion to nothing as well and only price the premium portion at a rate that considers the actual value that it adds to the user experience. I think you have to consider the free portion as a marketing expense.

When I was working in retail ages ago this was said many times as advice for new companies starting out. "It's easier to lower the price when you don't sell enough than to raise the price on something that is selling well."

Today it seems to work this way: "We start with free stuff and then we try to find ways to make money later". I think this is part of why we have the customers we have today. They expect everything to be free and then give you 1 star reviews if the background color is off a shade.

Today, that could be because devs tend to confuse how to measure success and keep a foot in multiple camps.

If you are trying to create a viral app, then giving your software away for free increases your chances of hitting the jackpot. But it decreases your chances of achieving steady cash flow in the short to medium term. If you charge upfront, then you probably aren't going to go viral. But you might be able to pay your costs. Freemium is an each way bet.

Everyone dreams of virality, but this dream might be sabotaging your nice, niche, steady earning app. You need to choose between your ego playing the app lottery or making small steady gains with real money. Sometimes going after virality is the right decision, but you have to look at your app and its target market and choose your charging model carefully.

As for the customers having low expectations, if the developers don't think that their software is worth anything, then it's probably not surprising that the customers agree.

Today, that could be because devs tend to confuse how to measure success

Devs/VCs. Or, rather, a VC is looking for a big sale down the road, not a company that makes steady income with low priced transactions.

Not only that but 'customers' who don't pay can also be some of the more difficult people to deal with, they are the ones who leave a 1 start review instead of sending you a bug report.

Behavior Consistency. I recently learned about this phenomenon. People make decisions on an emotional level and then logically support them afterward. In fact, there are studies where people who have had their brain damaged and no longer process emotion have a really hard time, or simply can't, make decisions.

Someone has to emotionally decide to put X amount of hard earned money on something before hand. The rationalization of this affects the overall experience. When an item is free their commitment to the app was zero.

There was a fascinating study about this.

Festinger, L., & Carlsmith, J. M. (1959). Cognitive consequences of forced compliance. The Journal of Abnormal and Social Psychology, 58(2), 203.

Where participants were given a task to convince other people to perform a long and boring task. One group was given $1 to convince people while another group was given $20. They found the $1 dollar convincers actually believed what they were saying where as the $20 group shrugged off their actions to convince people to perform the boring task as, "I did it for the money." $1 was too low to rationalize away.

>Behavior Consistency. I recently learned about this phenomenon. People make decisions on an emotional level and then logically support them afterward.

This is something I wish more people understood. It explains so much of human behavior when you realize that all (most) rational reasoning is merely a justification of their already decided upon stance.

On a somewhat related note, you may enjoy this TED talk if you haven't seen it. It's on the paradox of incentives and motivation: http://www.ted.com/talks/dan_pink_on_motivation?language=en

I think this is a key thing that is usually overlooked. When someone gets the app for free, they haven't self-qualified by paying and as a result they have zero personal investment. They don't want to "get their money's worth" or "look forward to the next version." Instead they get annoyed, leave a comment, and quickly forget about your app.

I think you should always charge something - even $0.99 - just to filter most of the casual users and get a more qualified (and invested) customer.

You should not charge anything if your app is semi-viral. I've done the experiments and charging kills growth for high retention apps.

If everyone did this (for cases where it actually mattered. Some app types are supported fine by in app ads) I would love to see how it would change the entire market perception towards paying for software. Right now it's viewed as a risk. A "should I make my app a paid for app or not?". I'd love to live in the utopia of developers simply being able to look at their creation, say that this is the value for it, and have people respect that.

Charging money will definitely reduce the number of stupid comments, but that is not totally foolproof either. On Amazon, I've seen many times people leaving bad reviews (after paying for it) not because the product was bad, but because they understood the description wrong and hence the product did not meet their expectations. It is amazing how many people won't take 2 mins to actually read the description before buying.

I totally agree. I put an app on Google Play Store recently with a yearly subscription based payment for use. Even though it has a 10 day free trial I have people downloading it, leaving one star reviews without even trying the app and leaving reviews with the likes of 'Should be free'. Very disheartening..

This is representative of the Android user base in general. It's one reason you see so many iOS first apps despite the market share difference.

And meanwhile, I keep using Android because I enjoy and prefer several things about the platform versus iOS and sometimes have to deal with an app being unavailable. It really is the one thing that I dislike about the platform. The freemium model (when done properly) is something I vastly prefer as well.

I look at it like the old shareware titles. I get a free demo that's either time limited or where I need to pay to unlock the full version. That way I can test the software or play the game or whatever, then if it's something that meets my needs, I can pay $1-10, secure in the knowledge that I won't be disappointed or surprised after buying something that turns out to be unwanted.

I'd say that the vast majority of apps and games I've purchased worked that way. Free download, mess with the app or play the first 5 levels of the game and then when I run into the limit where I need to pay, it's not a big deal because I know what I'm getting. The plague that is pay-to-win or other microtransactions in games is awful but that's probably not too relevant to this thread and is just as big a problem on iOS.

When I see the comments on the Play store or the App store, full of people complaining that the dev has the gall to ask for $1 or $2 for some useful or entertaining bit of code, it is very disappointing. That's the sort of money people pay for a Coke or to have someone drive a pizza 10 minutes to your house. I feel like if I get more than a day or two's use out of a mobile app, it's easily worth at least a buck or two and I'd vastly prefer to pay once than to deal with popup ads or recurring subscriptions and microtransactions.

Yup. And the problem now is if its not free, your product is not going to get ANY users. I really hope, everyone starts rolling back on freemium, because its hurting everyone (except customers).

I would argue that it hurts customers too, at least for some apps. If a developer doesn't get constant revenue there'd be little incentive or resources to update and maintain the app.

The customers are then stuck with an app that might have been good but decays over time.

Until the next cool freemium app in that space comes out, then the customer just switches to that one.

The real problem is that we just don't need 80 gazillion apps. The "app" market is extremely competitive, that's really what people are complaining about, profits have been driven down to basically nothing. Normally you would expect some developers to drop out of the market, which would help the situation, but two things are stopping that from happening, in my opinion.

First, the platform companies (Apple and Google) have been incredibly successful at promoting app development as a gold mine. For every developer who gets sick of it and moves on to something else, two people become "app developers".

Second, there's a startup bubble and the wisdom out there says you need an app or your startup is nothing. This partly relates to the first point, the hype around "apps", even years after Apple launched the app store, is deafening.

App stores should try charging people $100/year after the first year an app is in the store.

Or the reverse, pay $100 to pull your app from the store after its first year.

I believe Apple does do this - it's $99/year to become an Apple Developer, and I think you need to be an Apple Developer to have an app in the app store.

It is $99/year to have apps in the App Store, but it is per developer account and not per app.

I almost think devs would be better off if it was per app. At least any devs planning to make a living off the store. It would cut down on a lot of the crappy spammy apps out there.

It would also kill the one-off hobby stuff with which many, especially teens, get their taste for (app) development, which would be a high price to pay. Without a secondary path to distribute apps I think it would be a bad idea.


In my opinion, freemium is a form of predatory pricing.


I do believe many companies are employing this practice with freemium apps. There is something bigger happening here with freemium. At first we all called it "race to the bottom" pricing. And if felt like that. If felt like the value of our apps were dwindling to stay competitive.

But let's just look at how things were for a second. If you sold an app for 4.99 you making a promise to the consumer that you would deliver value equal or greater than your price. And you if you didn't keep that promise it was up to the consumer to get a refund or suck it up. This was largely a pre-internet thing. The internet has made distribution a non-issue. Now you can get your products and ideas out there for little to no cost.

What we are seeing is an opportunity to show initial value to our users and offer higher value at a premium. This increases the reach of our work and helps us to find those 1000 true fans. Of course, we have to nurture those relationships. It's a lot of work.

I don't think there is anything wrong with premium pricing. People are use to it still. Why aren't we just focusing on the value we bring to people and whether it's premium or freemium none of it would be an issue.

It can be done responsibly. But unfortunately, I agree it's mostly predatory.

Freemium can mean several things. I've purchased several apps (and generally paid more for them) when there was a free download that offers some functionality and the option to pay for the full version. Sometimes these are games where the first chapter is free and then you pay $5 for the full game after you've played a bit. It lowers the "friction" for the customer because you get to try something out at no cost but if you enjoy the first hour of play, you're more confident that you'll enjoy the other 5 hours.

Other times these are utilities. The Philips Hue lamps have an official app that's passable but the third party "Hue Pro" adds functionality. The free version does basically what the Phiilps app does or you can pay $2 and unlock all of the "fun" addons like music sync and lava lamp mode. Once I know that the app works and isn't buggy, I'm glad to pay. Same for launchers like Nova. Free version is a nice launcher that I vastly prefer to the stock launcher but I can pay a few bucks and get several more options. I probably wouldn't have paid $3 times 4 or 5 launchers just to find the one I liked but once I tried all of the free versions, I paid for the full version of the one I liked best.

It's like shareware. With the glut of options, you don't want to pay every time you try several solutions to find the best one so you try a bunch then buy the best. If I had to pay $1-10 to try every app on my phone, I'd just install less apps (or, to be honest, I'd probably try cracked copies to find the best option, then buy the best one).

As long as the quality of mobile apps is so varied, it really is hard to convince potential customers to pay any amount of money sight unseen. I've installed so many iOS and Android apps that lasted a total of 10 minutes on my devices it's not even funny. If I paid even $1 for each of them to learn those lessons, it would've turned me off to buying mobile apps at all.

There are a few apps that are making a lot of money from the freemium model. As long as that continues other developers will chase that dream.

Hmm.. which all are? Its hardly even a choice. As a consumer myself, anything not free doesn't interest me anymore. Companies are spoiling consumers (including me) silly.

Look at the apps on the top grossing chart. Supercell, for example, makes over a billion a year.

I'm not sure freemium games with in-app purchases are a very good model for apps in general (I'm not saying you're suggesting anything other than that they make a lot of money, just pointing it out in the context of this discussion).

It's hard for me to see how they're anything but a form of unregulated (yet) gambling/exploitation of people vulnerable to gambling.

Who in their right minds think that tens, hundreds of dollars is a good price to pay for virtual gemstones?

At any rate, for others that where unfamiliar with Supercell, I'm assuming parent means the makers of:


A somewhat related article:


(And there's of course been a number of articles on "predatory" game design, based around compulsive/addictive behaviour)

Interesting. How does this fit with the common advice (from patio11, etc.) that developers generally underprice their software?

It’s worth remembering that when people give that advice on HN, it’s usually in a business-to-business context. Business-to-consumer sales are a different thing entirely, with different motivations and often a different purchasing process. Consumer purchasing behaviour can be highly price-sensitive (and affected by numerous other factors as well as price) so for software or a service with low marginal costs, offering lower prices with greater numbers of sales can easily work out better than higher prices but fewer sales.

Patio11 is working in a different domain than mass-market products. He deals with niches; by raising your price in a niche, you find a different niche that is willing to pay more, which by definition will make you more profit.

This basically never works with the mass market, because the mass market, by definition, contains all those people who can't afford to pay more. To make a mass-market product work, you basically need a vision as to how everyone on earth should be doing things differently than they're currently doing them, and the sales and product skills to convince users of that vision.

I'm no expert either, but I don't think this only applies in b2b scenarios. There's some old anecdotes about new restaurants pricing themselves too low, failing to gain customers (due to being perceived as too cheap/must be low quality). After raising prices, serving the same food, they suddenly got more customers. Similar stories for retail (yes, anecdotes are not data -- but AFAIK this kind of thing is marketing 101).

I've not seen any studies on this wrt app stores, but my impression is that you can be completely free (and it helps with a brand name, like Autocad or Adobe) -- or you can be a little pricey with no free trial -- and still be perceived as being "high quality".

Note that big companies offering great apps for free is most likely a case of predatory pricing -- they might not have any ambition of making money from the apps ... ever. But if it funnels a certain percentage into their for-pay other apps/services that might be an acceptable price to pay for marketing.

Add a free trial, freemium, or low price point -- and your app is likely to be judged more harshly.

The price points/perceptions are likely a little different in the various app stores too, and vary a bit from country to country (eg: median salaries (and median prices for a cup of coffee) are likely to affect what people think are reasonable value for money).

I do think many developers dream of getting a big number of users, and tend to overestimate how quickly success should come. I would guess that getting a hundred paying users should be an important milestone, and depending on the app, one might need to keep those around for quite a while (a year?) before one might really start to see an effect of word-of-mouth in terms of sales. This isn't nearly as sexy as getting to one billion users in a year, but probably more realistic?

I'm not an expert like patio11 or others, but I think that it could be that developers aren't always as interested in economics as they should be and this can lead to not understanding the behaviour of markets.

Technically, developers have to underprice their software to be successful. This is true for any product or service that you want to sell. Particularly for productivity focused software, the end user has to derive signficantly more in value than they pay in direct costs to the developer. In general, in my experience involved with b2b software/services, people want to being saving 20-30% of their time or money to strongly consider changing their existing habits.

The reason for this is that there are also indirect costs that the end user incurs that developers don't always take into account in their cost models. End users incur costs in learning new software products, there are ongoing cognitive load costs in remembering how to use the product and there are opportunity costs in choosing your products over something else, not even necessarily in the same market. There is a cost that the user incurs in taking a risk that your software will solve their problem before they really understand its full capability. Even taking up an icon space on their phone/tablet/desktop has a cost to the user as it's a place they can't put a different app. Your app takes up harddrive space that the user has paid for and your app uses up some of their bandwidth. Finally, you aren't just competing against other software vendors in your market, you are competing against every other thing in any market the end user might want to spend their money on. End users have a limited amount of money and time and there are a huge amount of things they can spend their money on.

You can't accurately measure any of these indirect costs, but you know they are there and you can estimate them. When you take into account the indirect costs suddenly there aren't as many dollars left on the table that developers aren't charging for.

Another point is that the closer you price your software to its marginal utility to the end user, the more likely the end user is going to shop around looking for alternatives. You open yourself up to competition which could increase your marketing expenses. You could also suffer more churn, which is expensive to you.

The point is, you want your users feeling like they are getting amazing value from your product. You don't want them feeling like your product costs just a bit less than the value they derive (which may seem like the most efficient price to you) because if that's the case your business is actually teetering over a precipice and any sudden shock could push you over the edge with many customers leaving all at once. Loyal, happy customers paying you recurring revenue buffing up your cash flow is worth more than anything.

> The user discounts the value of the free component to nothing and only measures the value of the additional functionality that is unlocked with buying the premium portion.

I think you could solve this by having two distinct websites, apps or even brands. It's not going to feel like a ripoff if they are perceived very differently from each other.

Sorry, I have to disagree.

If you did this, why would you even have the free version? The whole point of the free portion is to act as feeder for the paid version. If the branding is completely separate then how are the users going to know they can have the same level of confidence in the paid version as what they do in the free version? May as well not have the free version at all.

> how are the users going to know they can have the same level of confidence in the paid version as what they do in the free version?

You are going to tell them. Kind of like this: "Need more space? Try our paidapp."

I don't think this is a fair comparison. You changed the business model from v1 to v2, by moving from a one time payment for unlimited articles to a pay-per-article approach.

A more suitable change would be to offer the app for free, with 10 free articles and 1 per day, and offer a one time purchase of 1.99 to lift the limits entirely, turning it into the equivalent of v1.

The way you did it is you changed it so that the act of sending an article cost money, instead of paying a small one time fee and sending as many articles as you wanted.

As a user I wouldn't understand why each article costs money, therefore I would feel ripped off.

Based on the article, I understand exactly why it costs money - he has to pay for the server that does the hard work.

That said, as a developer I don't understand why the server is involved at all. Here's the relevant part of the article:

> The server would then fetch the article, format it, package it into a .mobi file the Kindle can deal with and send it to Amazon for delivery to the device.

I understand all this, I really do. What I don't understand is why the computer that I'm holding in my hand when I ask for the article to be converted, which is probably as powerful as the $20/mo Linode instance, can't do this conversion itself.

Surely it can fetch the article, convert it, and pop up an Email sheet with the attachment ready to go and the "to" address filled in. All it would require is for me to tap the Send button and it would get emailed to my Kindle without having the developer's server involved. This way if the developer decides to discontinue the app and take down the server everyone who purchased it isn't left out in the cold.

I agree with <tomnipotent> about being able to update the server independently of the app. That's a good reason right there (the app requires network connectivity for email, after all).

Another reason is specific to this use: Amazon distributes a Linux command line tool called KindleGen that converts HTML to .mobi formats. You could install KindleGen on a Linode box in seconds and have a proof-of-concept converter running in hours.

I suppose you could reimplement KindleGen yourself for iOS if you had unlimited free time, but it likely wasn't worth it for this particular project. Just look at the release notes for a single version of KindleGen:

-) Enabled support for JP vertical rendering and multiple page writing modes (L to R, R to L)

-) Enabled support for facing pages (left or right)

-) Enabled support for double page spreads

-) Enabled properties to support spine for fixed format content

-) Added KF8 and M7 file size stats that will be displayed after conversion

-) Added Mobi7 support for HTML tags with multiple classes (e.g., class="class1 class2 class3")

-) Data URI support for images and embedded fonts so that they can be referred directly in HTML and CSS files

-) Multiple bug fixes and enhancements.

And remember the author of the linked article wrote that the purpose of the app was to "experiment with various revenue models on the App Store..."

A server easily handling hundreds of thousands of requests per day would cost only $10 per month. And if you want to link the app price to server costs, a monthly subscription would make more sense.

True, though having the conversion logic built into the app would require no server (costing $0 per month) and scale up to billions of installs without any effort on the developer's part.

You can make server-side changes without requiring the clients to update. That alone is more than enough reason.

Exactly this. It just seems the wrong type of app to use a freemium model. No surprises here.

Agreed, I was quite surprised when I realized there was a pay-per-share model being attempted. I mean I'm sure it's cheap but as a consumer I HATE games/apps that do this. I'll pay to unlock features or level packs but paying for something that costs the developer nothing. The costs of the server were tiny and getting a couple hundred dollars a month from V1 sales dwarfs the server cost.

Their best bet would have probably been to just add the new features to the existing paid version behind a $1-2 IAP (One time).

Yeah that's what I was thinking. Paying per article is absurd IMO. when I started reading the article I assumed the change would be exactly as you described, x free articles and then $1.99 for full version.

Unfortunately, Apple offers a very short menu of business models to the (iOS) app developer, most of which are absurd for the developer. The very sensible approach you suggest is called a "demo" or "free trial" version by Apple and explicitly forbidden. As you should know, clarky07, since I've learned quite a few interesting things about the life of an Apple dev from your much appreciated blog posts.

Ah but the only restriction is the app can't become a brick. He was still offering 1 a day, and could probably restrict that further and be perfectly fine under that rule. In my apps I limit the things you can save. You are welcome to delete them and save more, hence the app doesn't brick, but most people like to keep their data.

They have this rule, which they claim is one of their most common reasons for rejecting and app (from Apple.com):


If your app doesn’t offer much functionality or content, or only applies to a small niche market, it may not be approved.

This certainly suggests that the more you cut back, the more likely you are to end up rejected entirely at some point when you submit some unrelated bug-fix upgrade for approval.

It also suggests that iOS developers can't create useful apps that are just for their own family and friends. If Apple doesn't see the benefit to their other customers, your family and friends won't be allowed to have your app.

Your app is too niche for a freemium model. You Said yourself that the app solves a very specific problem and it's the only app in the store that solves that problem. I would suggest you switch back to up front payment but jack up the price to the 5-10 dollar range.

I've seen my revenue triple when switching from 3.99 to 9.99 for a very niche app that was almost without competition. And that was on Android where people are even less likely to pay.

This was my analysis as well, although we both have the benefit of hindsight, and everything is obvious once you know the answer. But it looks to me like the following models are likely optimal:

1) App that solves a trivial problem, no viral potential, lots of competition: free (or don't do it at all). Apps that fall into this category are things like distance measurement, to-do lists, etc.

2) App that has viral potential but lots of competition: freemium (games fall into this category)

3) Niche app that solves a hard problem for a small number of users: paid, with the price as high as possible.

There is a fourth category, but it's illusory: app that solves a hard prolbem for many users and has viral potential. If you're an individual dev, the odds are in those areas you'll be facing competition from players with much deeper pockets who are often giving apps away free to drive their core business.

This person's experience tends to validate this general analysis.

If you want to experiment with letting people try the app for free, you might want to consider just using a single permanent IAP purchase of some higher-than-$1.99 value (say, $4.99 or $9.99) that unlocks unlimited articles. Without this purchase, you could limit them to something like 20 or 50 articles. Enough for them to get a feel for what the app is doing and decide that it's worthwhile. The idea being that anyone that does send enough articles to hit the limit would likely have decided it's useful enough to be willing to pay the higher cost (which they may not have been willing to pay upfront if you simply charged that much).

Or you could just go back to your original price, since that was working out. It's hard to say which would do better.

Exactly what I was thinking. "Credits" feel like a huge ripoff.

The credits makes people think 'Is this article worth sending to my kindle? It costs money to do so. Maybe I should just read it?', that adds loads of friction and poor UX.

But the OP was experimenting, so I guess that's all good :)

Previously the user paid 1.99$ for unlimited use, forever. Now you nickle and dime users, forcing them to think about credits, what the value of the service really is, etc.

This got me thinking about something.

What's to stop an unscrupulous developer selling an app for, say, $1.99 with the (implied) promise that the app will work forever and then, once people have bought it, turning off the backend so it no longer works?

Is this illegal? Would Apple shut down the developer's account?

It's not illegal, and Apple wouldn't shut down the account.

Many companies do this with games that have a strong online component.

I think it would depend on the jurisdiction, and the "degree" of "implied promise". Eg. in Norway there's pretty strong protection for (private) buyers. On the other hand, proving fraud in the case of the company behind the app simply declaring bankruptcy/shutting down would probably be pretty difficult.

I suppose a relevant study would be Microsoft Zune/music store, and/or Yahoo(?) Music that shut down and killed the DRM servers, effectively removing music people had "bought"?

I'm not sure why Zune keeps coming up as an example in these threads. It got rebranded to Xbox Music, but all the servers are still up and anyone who subscribed still has all their music.

An earlier Microsoft DRM that was used by a bunch of services (PlaysForSure) was indeed shut down, but it's not the same thing as Zune.

Oh, I was thinking of the earlier one. I suppose it's often conflated with Zune, because of the spectacular flop of the Zune devices/rebranding thing.

Seemingly, a compromise is required, since "$1.99 for unlimited use, forever" is unsustainable if there are on-going costs.

I think there's a middle ground to be had though. For example, if the average user sends 1 article per day, offer 400 articles for $1.99, and in theory you have a recurring billing model that also doesn't require too much worrying from the user's point of view.

The other model might be $1.99 for unlimited use, and then release V2 of the app with new UI/features, and do an annual release model. It works for many apps, eg developer tools like IntelliJ. Would be interested to know if it works on app stores without getting grumpy user feedback though.

This is a really good point.

However, there are some apps that do have a model of $xx for unlimited use, even where there are ongoing costs.

This is sustainable if the app continues to gain new users, hopefully at an accelerating rate. It works like a pyramid scheme where the new users coming onboard subsidise the cost of the existing users.

Usually they release V3 and discontinue support for version 2. Maybe version 2 is somewhat useful still, but it probably doesn't have support for what you need anymore.

maybe the ploy should be to get to V3 as quickly as possible as this pyramid / growth rate collapses or cut costs in V2 (cheaper servers, more efficient code - if possible). This is an interesting thought to me nonetheless.

Absolutely. If you force the user to choose to pay on an item-by-item basis, they're definitely going to use it less.

For me, it's a bit like mobile phone data charges. When I used to pay per mb, I hesitated to use cellular data at all, but now I have a (high) monthly allowance, I now use my phone a lot more, even though the cost is higher.

Perhaps the developer should try switching to a pay-monthly kind of subscription, at least then the users aren't having to make that 'is it worth it?' decision each time they want to use the app.

I don't view freemium as a business model, I view it as a distribution model. The distribution model changed from paid to free. This usually results in a user base that grows faster but pays less. By making the app free you've set an expectation of not having to pay. This attracts a certain type of customer, which are usually much harder to convert. If you're trying to build traction for investors this might not matter. If you are trying to monetize then this does matter.

The business model changed from a one time purchase model to a micro payment model. My recommendation would be to try a monthly payment model for v3. Let the users pay to access the service (setting an expectation to pay for value) but allow them to pay less than what you perceive to be the cost barrier (the amount you believe is the maximum amount they would pay to give it a try). I'm going to agree with others that you are priced too low. One of the best lessons I've learned is that it's not what you think it's worth, it's what your customer thinks it's worth.

I noticed a key metric was missing from your article. What was the average number of transfers per day per user with the original version? The explanation for the drop in sales could be as simple as most of the existing users were doing 1-2 per day...

After averaging $5.00/day for a year with my iOS app using a freemium model ($1.99 in-app purchase), I decided to switch to a paid upfront model just to test. It resulted in a sustained average of $25/day. That's a big deal to me and so counter to everything I read about freemium being the only way to go in today's mobile market.

Out of curiosity, what is your app? I'm interested in seeing what kind of app that would happen to.

How I killed app sales by switching to metered pricing

I agree with the sentiment, but everything about the service, apart from the app is metered as well (i.e. keeping the server)

Selling an app is fine, but never sell the promise of something (a service) forever.

If the service is continuous make sure there's a recurrent charge. This is not about nickel-and-diming, this is about keeping your service profitable (or at least covering its costs)

How I killed all goodwill by trying to milk customers.

Which is more-or-less exactly the same as saying "freemium", 99% of the time.

Not really. Many apps have a one-off "premium features" IAP. Certainly more than 1%.

FWIW a while ago I decided to publish a free + IAP version of my iOS photo/video effects app, which was previously a paid-only app. At first the results were a bit disappointing but over time it's grown to generate about three times the revenue of the paid app, and about 20 times the overall user base. I'm sure it's true that freemium is not a panacea but it's worked for me.


Hey, this is off-topic, but I'm working on an app that's in the same market as yours. (MIDI-based DAW-ish thing with an emphasis on ease-of-use. I don't think my app would be competing with any of yours.) If you're willing to share your thoughts, how easy was it get the number of users for, say, Sythe as it has now? Did you do any marketing? Was the relatively high price point ever a problem? Are you considering freemium for that one, too?

I was hoping that I could mostly wing it and just make a nice video and send out some review copies to people who might be interested. The music app market seems small enough where this might be possible. But this is just me guessing, and it would be great to hear from someone who's actually had some success in the space.

Anyway, loving your design aesthetic! Will give Sythe a download as soon as I charge up my phone.

(And, obviously, if this is information you're not willing to share, then no hard feelings! Just thought I'd ask.)

To be honest, Scythe was a huge amount of work and brings in about half of the revenue of my photo effects app, which took about 1/10th of the time to write. Knowing what I know now I would never write another iPad app again, and I'm not sure I'd write another serious music app. If I do decide to write any more music software I'll target the VST plugin market, where people are still willing to pay $99+ for a good instrument. It's a shame because I think the iPad has tremendous potential as a platform for music making but it seems like the volume just isn't there to make up for the low price points users expect for iOS apps.

One thing I would do differently if I started from scratch on Scythe today though is make the base app free and make up the difference with IAP preset packs and additional features.

Thank you for your insight! It's too bad Scythe isn't as lucrative as you hoped... seems like it's hard to get serious software noticed in the App Store over free, bite-sized entertainment apps. Incidentally, my app is targeting the iPad as well, but I don't really mind the business reprecussions: I'm writing it for my own personal use above all else, and I wanted to give my trusty old iPad 3 something useful to do instead of just serving as a very expensive web browser. I was thinking of going Lite/Pro, but perhaps freemium might be better after all; I just really don't philosophically like the idea of charging for individual features, especailly when they're part of the MIDI standard already. (But that's just business, I guess.)

There's been a lot of talk recently about the iPad languishing, and it seems that Apple is taking notice. Would not surprise me if they made more of a marketing push towards promoting iPad apps in the coming years, especially with the possible arrival of the iPad Pro.[1][2]

Crazy that the VST market can be so lucrative! It makes me happy that some people are still willing to pay a fair price for their "workhorse" software.

[1]: http://www.macrumors.com/2015/02/08/apple-ipad-grammys/

[2]: http://www.macrumors.com/2014/12/31/apple-expands-start-some...

The VST market isn't really lucrative. I have it on good authority that most plugins sell at most 200 copies. Music software is also very challenging to write because it combines multi-threading with soft-realtime requirements. So any programmer good enough to write non-trivial music software can easily make more money with fewer headaches writing almost any other kind of app. So I'd say the only good reason to write music software really is just for the enjoyment of doing it.

As for myself, as much as I enjoy writing music software I enjoy making music more and it seems like I only really have the mental energy for one or the other.

As someone who would like this sort of app there's no way in a month of Sundays I'd be willing to pay for each article after the first

This sort of thing is a one off purchase (if my device does the work) or fixed monthly fee (if your servers do the work)

This seems more fair. But it is weird, because thinking about it, isn't buying N units of work the most fair pricing scheme? I didn't have the same gut reaction to Amazon's Lambda service.

Perhaps it is that we know the article limit isn't related to any real-world costs that make the difference.

It's fair but the worry comes in with what happens if I run out of units of work and send a new article. Does it reject the article and I miss out? Does it charge as and when I send an article[1]? Do I get charged extra for unexpected usage?

If I have to ponder any one of these questions I'd just switch to a different app, there's a fair few that provide this functionality

[1] In this case the main worry is what if I accidentally somehow send every tab I have open to this service through a bug or something? Do I pay or does the app maker take one for the team?

As someone who would also like this sort of app, I can do this completely free with Pocket on my Kobo reader. I don't know why anyone would pay for this service.

Great article and I love the experiment, but as the author himself points out this isn't apples to apples.

What you really want to see is a freemium version where you get:

1. 3 total articles for free 2. A 1.99 in-app purchase that gives you infinitely more articles.

The current free limits are high enough to support the majority of casual readers.

I have never read up on app business models but intuitively I'd say in app purchases create unneeded extra friction. Additionally if you pay up front you have a sense of ownership.

My guess is that I'd only use freemium when network effects matter. The app described in the article doesn't rely on network effects at all but it's nice to see my gut instinct corroborated experimentally.

I'd be interested in seeing this repeated for Android because I'd suspect iOS users are generally more likely to buy stuff.

This comes up in advertising all the time - you can increase how many impressions an ad gets easily by just showing it more but the conversion rate depends entirely on the kind of user, match between user & ad than anything else.

In this case, the people who paid for the app self-selected themselves by the act of paying, but the barrier to install the free version is much lower than that.

I agree, it's a really complex interaction.

You have to consider: 1. The number of users that would never pay for your app, but are happy to download it for free. 2. The number of users that would never pay for your app without trying the free part first. 3. The number of users that would pay for your app without trying it for free.

You then have to consider the amount of revenue you lose from Type 3 to support Types 1 and 2. Hopefully you come out ahead, but you may not.

Are there any good examples of successful Freemium apps that sell additional functionality (rather than games or consumables like the article)? The only major examples that come to mind to me are photo apps that sell filter packs. Are there any apps that lock down their major functionality as demoware or upgrades? (like say, "add iCloud support for $2")

Maybe an abberration but Cyanogenmod's ClockworkMod comes to mind. They provide it for free but the auto updates are paywalled.

Diptic sells additional functionality, some of which I've happily paid for once I realized its value to me. It's been so long that I don't remember if the initial app was free or paid though.

Stronglifts 5x5 app does, Wine Secretary does, several of my children's apps do. These are just a few I can think of that I bought the last week or so.

The major fitness apps all do this. Things like runkeeper, map my run, Strava, etc

Google drive/dropbox?

I can only guess it's because he was faaar too generous with the credits. One article per day? I send maybe one-two per week. Maybe a better solution would be to give each user 5 credits total, and then they would have to pay to keep sending articles(I would do it as an unlock thing, rather than selling them more credits though).

yeah... definitely. Cut back on the free offering, and make it "pay to use" for anyone who ends up using it anything resembling daily.

Also probably worth making it a subscription ($1/month?) instead of credit-based - subscriptions can give dependable revenue, and users wouldn't feel nickle'd and dime'd.

Anyhow, the simple lesson here is that freemium isn't a magic bullet - probably takes some time to figure out the right place to put the paywall to get growth / usage via free users, and revenue from not giving everything away for free.

I'm launching an app in a couple of weeks and I'm really worried about the revenue model. It's an app that acts to an client to an other service and they'll probably start charging for the API in the near future.

I felt like I had a few choices seeing as I'm foolishly doing this full time and need some kind of revenue:

1) Free + Ads

2) Free + Freemium for "pro-features"

3) One-time fee

4) Subscription fee

I have a distaste for ads so I really didn't want to go down that road.

I don't like freemium because looking at "pro features" and valuing them based on them alone it seldom seems worth it. Also keeping useful features away from non-paying customers seems lose-lose.

I could go with a one time fee but then I won't be able to support it over time and the API fee's will eat away any revenue, no matter how small, over time.

So I've decided to launch with a subscription yearly fee (through a in-app purchase). From all examples I've seen of changing revenue model to subscription model after you've launched it seems almost impossible without the mob getting out their pitchforks. It also seems more honest to have the subscription carry over than launching a separate new app that you have to buy for future OS versions.

Still, mentally users aren't really used to paying subscriptions for an app so it remains to be seen if its successful. Also since most people don't pay for this service itself already there's another mental barrier.

Actually if the app wasn't dependent on a service I'd rather see you'd own the app and subscribe to updates, it seems like a much better model since you never lose "ownership" even if you don't pay, but unfortunately app stores don't support that model.

I think my app is better than the competitors but will people choose a competitor because of the subscription model? It will be interesting to see.

Mob with pitchforks is not something to worry about. No mob. That's what you need to worry about.

Funnily enough, I am just on the opposite side.

I've a free API that allows people to build clients for mobile apps (some are free, some are one time fee, etc) and the costs of running this free API are starting to worry me.

I ran a free system and while it is small it easy but as soon as you get tens of thousands of users you are on the hook for all of the background costs while your userbase grows exponentially because they are so happy. Better to have 10 customers then 10,000 free users. I was never able to get free people to pay anything. I bragged about my free users and I felt self-important while taking away my ability to use my time more wisely.

I've noticed online services which probably cost a few cents per user to run seem to either charge $0 or something like $10/month which works out quite a lot. I wonder if there is a gap for charging say 3x actual costs which might be say $5/year or $10 for 5 years or some such?

1.0 is rewarding use and 2.0 is addig friction to use. For this app to demonstrate value you should be able to freely use. The "buy credits" to do this niche action feels like is not worth.

I will risk to say that there is a Cognitive Load inbalance. Fremium induces people to track the economics behind the usage of the freemium app. If the problem it solves is less painful than the Cognitive Load of tracking its fremiumnality, then why bother?

It would be interesting to do tests about models without "risking the 1.0s"

Would love some insight and helpful advice on sales.

We have a product that is ready for paying customers and we have a landing page (only) that is to collect sign ups. We get 100 sign ups a week & up to 15% of those express interest in being customers, but once we get to talking ... after a few emails these talks taper off and even a follow up email leads nowhere.

We have had talks with small parties no one has heard to huge Fortune 50 to 500 companies (this week a VP from Lowes started talks, previous Black & Decker, Motorola, Samsung, Beats, etc).

It seems most people want to use our unique WEB technology for free, though once we put it on the web these potential customers may just show their true motives of wanting to know how we accomplished X and copy it.

Thus, I've been thinking of doing a freemium model myself where users can use part of the product that is not our secret sauce yet have to pay X amount of money to add our secret sauce into the product for their specified use.

What do you think is a better sales channel then we have it now and or does my next sales channel idea sound solid?

Freemium is almost always a scam, remarkably for games where it degenerates into "pay-to-win". As a parent of two young tablet users, I'll happily pay a decent price for any game they want, but I won't allow them to spend even a single cent in in-game crap.

It's kind of scary the number of games aimed at children that have IAPs of £69, and parents sometimes don't know enough to prevent their children spending hundreds of pounds on IAPs.


I wish we'd crack down on this kind of abuse, but it's an arms race, and at some point hard to distinguish objectively. There's no solution without some parental responsibility... many parents are just too lazy, the hand a device to their kids and hope for the best. Parenting takes some effort; my children's devices have parental control apps (I use and recommend ScreenTime), as for purchases of any kind they have to ask me to put my password every single time because I don't check the "remember" checkbox, it's not terribly hard.

Freemium allows you to have more users and more chances to sell.

With a paid app, the same users would have to make a decision to buy based on just a picture, description and reviews.

Many freemium apps work really well by selling in-app currency. The most effective ones, sadly, were games that made it increasingly difficult to progress without spending said currency, making people addicted and choosing to spend money instead of time, which they valued more after a certain point.

Angry Birds found out two years ago just how much freemium beats the pants off them.

Look up pricing tricks online. One is when people see a too-low price that looks out of place next to the others. They think they caught a good opportunity.

Freemium could potentially work here with better calibration. I suspect some of the things that might be important for you users:

1) Unlimited usage or unlimited usage per month -> Managing credits is an annoying experience for when I just want to share an article: I'll have faster ways if the actual transaction is a hassle. Paying once and forgetting about it adds a lot of value. This could work with your original paid model or with a freemium model that includes an unlimited tier.

2) Better analytics to calibrate the use cases: How many articles were people using. Did you have a way to detect how many of the articles were actually being read? I suspect that 1/day might be a very high free tier if people are switching to a mindset of: is this worth being my 1 article a day? Or converting the cost of a paid article and asking "do I want to pay X for sharing this to my kindle?"

Other have made the point that a key aspect of freemium is viral growth. Do you have any social sharing features that people would want to use -> Could you share an article to Kindle and/or Facebook + Twitter? Could you get shares of your app in exchange for additional article shares to kindle?

If freemium was the right model it might end up at something like:

10 Initial Free shares 1 share/week free 3 shares/week free to device if also shared to a social network A Paid credits/usage tier A Paid Unlimited/time tier (monthly or weekly unlimited use pricing?) And Possibly a paid outright unlimited tier.

You'd want to measure and track analytics on how you get users. Something like web bounce links through your own analytics package on the way to the app store.

Here you actually have an interesting tradeoff: You get less money per user which you hope to balance out by driving down CAC through large amounts of organic growth.

The purpose of freemium is seldom a try before you buy value proposition that makes sense for the company, it's the ability to generate viral growth even out of users who never spend money. If the app won't get substantial viral growth then freemium is almost always wrong.

I think the biggest issue is you switched to using credits. That and you've only let this play out for 20 days. People can take quite a while to evaluate an app and decide to purchase. You have more experimenting to do. A one-time upgrade fee for unlimited or a yearly subscription seem like a better fit for how a user would expect to pay for this type of app than credits. At the very least perhaps add an "expensive" unlimited use option and I bet you get some purchases of that.

Products with ongoing pricing turn out very expensive in the end. A product that is $10/month will often workout more than one that it is $100 upfront. I prefer iOS apps that you either pay or upfront or, even better, a one-time fee to upgrade. I don't have to worry about ongoing costs then. I'm reminded of ongoing costs everytime I get a statement or receipt, and then I look for alternatives or look to cancel.

Fremium only works with high retention apps. The goal with a high retention app is growth, not revenue. You want people to use it a lot, tell their friends and they use it a lot. If you can hit 4%+ week-over-week growth then this is a good candidate for fermium.

Anything that has low retention should be charging up front.

This bit really hits the nail on the head, "One thing freemium compensates for is the lack of a free trial system in the App Store."

There is no reason the app store should not allow developers to offer free trials. This is the one thing that Microsoft got right with the failed Windows phone.

IAP does pretty much the same thing. I understand we wanted trials 6 years ago, but I personally feel like this is a solved problem.

If people have a specific problem, and your app solves that problem, you don't need a trial. Just tell people in the description that you solve their problem, and they will buy it.

Free trials are only necessary when you want to sell solutions to problems that people don't know they have.

It'd be interesting to see what the retention numbers were like for those initial v1 users, and how many people use the free version on an ongoing basis. It could be that people stop usin the app after the first week or so anyways.

Freemium isn’t the only option to allow people try the app before buying it. There are free trials that allows to use the app for several days and then asking to pay the full price to continue using it.

I think Apple would likely reject such a design as being a "demo" version, which is explicitly forbidden--unfortunately.

Good point. Free trials are currently available for Newsstand apps only: https://developer.apple.com/library/ios/documentation/Langua...

Has apple ever given a reason why they dont allow demo/trial apps?

A time limited trial seems like the perfect fit for productivity apps, not sure entertainment based apps would want this model. You have a day/week/month to use the app in its entirety, if after the trial you value the product you purchase it, otherwise it no longer works.

I don't think they ever gave a reason. Personally I like how it has forced developers to come up with more creative product offerings than time limited demos. While convenient for developers, time limited demos suck as a user experience IMO. The whole "trial expired! buy now!" experience is very grating.

I am fairly certain the "trial expired" experience is exactly the reason. They're also trying to make a commodity out of software and having an app store half full of trial versions doesn't help them with that goal.



The startup world has pretty much destroyed future chances of making a living from an application. Sure, the top 10 apps in the apple store might be making good money, but most apps that charge $1.99 will most likely not pay a living wage to the people that created it.

The same thing happened to the music industry. The total amount of revenue has dropped 10 fold since 1999. I know a few independent artists and they can't really make a living on just music anymore. They need to sell t-shirts or have a day job. So in the end, free music and file sharing has decreased the overall cost of music, but it really only hurt the small and independent artists.

Even look at things like Spotify. 1,000,000 plays nets an artists something like $100.

It has forced many artists to go with large labels because the large labels will take all of the risk and actually pay them some sort of a living while they are touring.

Software is going the same way. With all of the open source out there (and more coming out every day), it has allowed companies to hire software mechanics for pennies instead of software engineers for a living wage. Why? Because all of the parts that need to be engineered are given out for free. Development now in most businesses is mashing frameworks and libraries together. It really doesn't take a computer science education to get the job done.

Business owners are also becoming more tech savvy as the older generation retires.

I've never seen so many smart communities of people put themselves out of a job so quickly.

To me, apps are just an interfaces to a paid service, not the actual service itself. So, I might have a service that you pay X amount of money to use and the app is something free that is thrown in as a benefit.

Totally agree with your comments about the music industry, however I don't think the same applies to software development.

If many businesses can now get by with "mashing frameworks and libraries together" then that is a good thing - it means problems have been solved in a repeatable way that many people can benefit from, and the real software engineers, rather than doing the same old work over again can refocus on new challenges and create the frameworks and libraries of tomorrow.

Is that really a bad thing? The only people putting themselves out of a job are people trying to earn a living working on problems that were solved a long time ago.

Suggesting there are fewer jobs now for software engineers than any time in the past is a bit of a stretch!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact