Death to Freemium, Long Live the Free Trial
Here’s a reprint of my first guest post (in Mashable, nonetheless). Since they took out some of the juicier points (particularly why the App Store doesn’t offer refunds), here’s the original version.
__________________________________
There’s a psychological barrier to paying for online services, and it’s hurting both consumers and businesses.
We pay for things in real life every day. We easily spend $5 on a vanilla latte, but agonize over paying $5/month for a really valuable web service. Freemium business model has emerged as a solution to this unfortunate customer behavior.
There are a many shiny examples of widely successful Freemium companies. With Evernote’s billion dollar valuation, more people will point to Freemium as the business model to strive for. Unfortunately it’s an exception and it sets a bad example. Companies like Evernote and Dropbox are the poster kids for VC-backed success, but not necessarily an ideal model for every online service.
In my experience there are 3 main cases where a Freemium model works:
1. Evernote-like paywall
The way the product is designed, a significant portion of the users will inevitably cross the paywall. The longer you use the product, the more value you derive from it, and the closer you are to hitting the free upload limit. WriteThat.Name is another great example of this, and one of my favorite products. It extracts contact info from email signatures and creates/updates your Google contacts. The free version gets you 40 new contacts per month - 1.3 per day. Since I get more, paying $3/month for this service is a no-brainer.
2. Dropbox-like network effects and viral lift
Dropbox has inherent virality, and it’s value increases with more users (in order for you to share a folder, the other person has to be using the product). It also doesn’t hurt that you have an additional incentive to invite them - permanent extra storage. This is the classic Fred Wilson “free users are your marketing cost” argument. Most network effects businesses have to offer some form of value for free.
3. Spotify-like super annoying ads
Five bucks a month is a small price to pay to avoid the anger you feel after hearing the same commercial every 3 minutes. Of course, it’s a fine line between being a) valuable/annoying and b) just annoying. In case of (b) the user is much more likely to stop using the service than pay. From business standpoint, ad monetization sucks because it requires huge scale to generate significant revenue. So you’re much better off charging a user $5/month, instead of blasting him with ads that may get you a $5 CPM if you’re very very lucky. Sparrow, for example, gets away with charging $10 to remove an ad unit in a particularly annoying spot.
Of course, if your cost structure doesn’t support lots of free users, neither of the above will work, but for most of us user acquisition cost outweighs support cost by orders of magnitude.
The problem is most consumer internet startups want to be in the above 3 categories and probably lie to themselves just a little. But by charging $0 you’re actually anchoring that value in your customer’s mind, making it harder to raise the price later. One approach that seems to work is staying in free beta (or offering $X credit) while you figure out where to put the paywall.
Long live the Free Trial
So what if you don’t satisfy one of the 3 cases above? The model we went with at our company is the free trial after which everyone has to pay. It works beautifully - more than 25% of our free trials convert to paying users, and we don’t even ask for a credit card upfront. The remaining 75% leave, and we don’t have to support them.
“But isn’t this a kind of freemium?!” you might say. Not really. Freemium is when there’s a free and premium versions of your product. A Free Trial is more like a “no-questions-asked” returns policy for the web. I’m fine paying $120 for Microsoft Office upfront - I’ve been using it for 2 decades, and know exactly what I’m getting. But it’s harder for me to shell out even $1.99 for a product I’ve never tried.
“How long should the trial be?!” That depends on how quickly your product can show value. The more the value increases with time, the longer the trial should be. On the flip side, the shorter the trial, the faster you can optimize user acquisition (and the less free users you have to support). For us, the value is seen almost immediately, so we are currently experimenting with shorter trials.
What boggles my mind is that the Appstore hasn’t introduced a Free Trial model. There’s no real reason NOT to do it. If you try an app, and don’t like it, you should be able to uninstall it and not have to pay. It would probably increase sales for paid apps significantly. The lack of a reasonable return policy actually feels kind of scammy. I wonder how many people paid for crappy apps they don’t use, and what the active users/downloads ratio is for most apps?
Death to Freemium
There’s a lot to be said for creating something of value and charge money for it. If you’re not charging for your product, then your users ARE the product. This actually screws the user in the long run (when the company dies or gets sold to somebody who can monetize your users). It also forces you to focus on 2 unrelated efforts: 1) growing your user base, and 2) figuring out how to monetize it. (And please don’t use Facebook as an example of waiting to monetize your audience. You are not Zuckerberg.) There are many benefits to having free users and focusing on hyper growth. But the decision to go with Freemium should be based on math specific to your business - not a pricing philosophy.
And do your part as a consumer - stop being cheap. Just because something is virtual, doesn’t mean it should be free. As we spend more time online, we should get more comfortable paying for value delivered there.
