Make Money Online: Business Plans for Web Apps

Dilbert comic on venture capitalist stupidity

You have a great idea for an App, you’re all set to start building it, and then you realize you have no business plan. So you start asking around: “how do I make money with this idea?” If you came to me, this is what I’d say…

My first thought is that if your goal is to sell a revolutionary concept to a larger company, you may not want to make any revenue. This recent NYTimes article, while skeptical of the approach, shows how common and successful it can be to build companies with no revenue. In fact, many venture capitals specifically advise it once they’re on board (see the article for why).

But if you’re interested in building a longer-term company that is self-sustaining, I would recommend one of the following approaches:

1) The “Freemium” Model

Used by Evernote, Google Apps, and many others, these companies provide a basic, useful service to people for free and then charge for premium versions of the concept. In this model, it’s important to make the free version good enough that users are hooked to the brand. There are two types of freemium models:

A) The Power User Model

This version is employed successfully by Evernote (and Pandora, Flickr, Skype…). As the Evernote founder says: “the easiest way to get 1 million people paying is to get 1 billion people using”. Their ~2.75% turnover rate means they’re doing even better than that. To do this successfully your free product has to be extremely useful on its own. Once you have a large user base, you just need to figure out what your power users (the most committed and avid users) would need to get more value out of the service. Charge them to get it and you can start making money.

B) The Consumer Advocate Model

This version of the freemium model is used successfully by Google Apps (the suite of apps consisting of Gmail, Google Calendar, etc.) and many others. Google does not attempt to convert a certain percentage of their free Gmail users to be paid users. But by building their brand with free users, they can market their paid versions to businesses who are willing to pay and get help with this marketing from their free users. When free users attend a university or are employed by a company, they will be more likely to recommend the use of Google Apps because they know it and trust it. The big idea with this second approach is that if you can win the hearts and minds of the frugal masses, they will be more inclined to recommend your service when among those who are willing to pay.

2) User-As-Product Model

This version is employed successfully by Facebook, LinkedIn, DuoLingo and many others. The important idea here is that users create value by virtue of using the service, and you capitalize on that value by charging other people to put that value to use. While this can mean advertising, it doesn’t necessarily mean advertising (and most often, should not mean advertising).

So if not advertising, how else can you turn your users into a product?

LinkedIn is one great example.  Most of their revenue comes from recruiters and recruiting agencies. These agencies aren’t (just) advertising, but rather are taking advantage of advanced search and organizing mechanisms that LinkedIn has built to capitalize on its users’ online resumes. In fact, LinkedIn makes 20 times as much money per user as Facebook by mostly ignoring advertising. It can do this because it makes money from its users’ information even when they are not on the site (because recruiters can still search that info) whereas Facebook only makes money when users are engaged in their site (and looking at ads).

DuoLingo is another great example of user-as-product, and they have 0 advertising. They are a free language-learning site, founded by a Carnegie Mellon computer science professor. They make money by offering language-translation services to businesses around the world. From their research, they found that they could combine the translations of novice language learners and get results that are as good as professional translators’ (see the video in that link for how they do it). So, as part of the language-learning process, DuoLingo users are translating real-world text from international websites, and these websites are paying for the translations. It’s truly win-win, because their users get free language-learning tools that are as good as the leading competitor (Rosetta Stone) and websites get translations at a fraction of what it would have previously cost.

The key with building something like LinkedIn or DuoLingo or Facebook (or Craigslist…) is creating a free, attractive service where users not only get value from their use, but also add value by their use. If going down this road, it should be clear up-front to your users that they are the product. With LinkedIn, that works great because people who post their resumes online do it with the express purpose of being seen by recruiters. This works for DuoLingo because many people would prefer to translate real-world sentences, rather than made-up ones. It’s less clear how well this works for Facebook, because most users tolerate the ads, but would prefer to navigate the site without them.

That last point is really important: the best user-as-product sites work because their users want to be the product, they don’t just tolerate it. This is why traditional advertising online sucks: People don’t like being distracted from what they are doing; people don’t like being tracked; and people don’t like having their private lives shared without their express permission. And yet, without doing any of these three things, traditional ads are all but useless to advertisers.

A new approach to advertising would look at how advertisements can act as a service for the user. And in fact this is already being done successfully by a number of websites. If I do a Google search for Pizza Hut, it’s nice that the first result is an advertisement for Pizza Hut’s official website. While Pizza Hut’s website is also the first result that shows up organically, that can’t be counted on. So, Pizza Hut and Google are doing me a service by ensuring I get what I’m searching for without any hassles (this could have come in handy for Facebook a couple years ago when they weren’t the first Google result for “Facebook login”). Another example of advertisements-as-a-service would be Foursquare‘s rewards program. Users get coupons for telling their friends where they are (“checking in”) and businesses get the benefits of word-of-mouth advertising. Foursquare users want to be advertised to, because that means they’re getting free stuff (or recommendations).

So how do you figure out which approach works for your business?  First, write down all the different sub-populations (make them as specific as possible) that could get value out of your idea (e.g. “recent college graduate newlyweds in the housing market” or “Small-Medium Businesses looking to expand into new cities”, etc.). On a scale from 1-5 mark down how much value each group could get from your idea (it’s easiest to do this if you have a lot of sub-populations so you can figure out relative values). Then rate each sub-population on a scale from 1-5 on how likely they are to spend money in general (income is an obvious factor here, but age is as well, and there are many others that aren’t necessarily related to how much someone has in the bank).

Now, plot the results using a scatter plot, where one access is labeled Value Received and the other is Willingness to Spend (see images, below).

Scatterplot Demonstrating the Power User Model

If the Value Received and Willingness to Spend ratings grow in tandem at a steady rate (i.e. if your graph looks like the image above), then you should go with the Power User Model. The trick with this model is to think about what value you could provide to those who are willing to spend money that you previously ignored because you didn’t know you could count on a steady revenue stream.  By doing this, you are pushing everyone to the right (everyone receives more value).  The big spenders are the first to get more value, but because they’re giving you money, you now have more money to develop the entire platform.  Think about Skype – they provide calling to phones for a cost (a value for the big spenders). And the revenue they make from this can be used to make call quality better for everyone.

Scatterplot Demonstrating the Consumer Advocate Model

If your sub-populations cluster around 5 and 1 (i.e. if your graph looks like the image above) then you should go with the Consumer Advocate Model. Here, you are looking for ways to grease the wheels of fewer but bigger sales. You know that the big spenders get a lot of value from your product, but it may not be easy to sell to them.  So if you can push the cluster of frugal people to the right (give more value to free users) you will have consumer advocates who can help you sell to the big players. The example I gave earlier was consumers advocating Google Apps to their employers, but this could just as easily be people recommending a product to their rich friends.  One way to add value is to help frugal people look smart to big spenders. For example, the fact that Google offers Gmail for enterprise customers means I get value as a free user not only from the free email services, but also because I can look like an expert to my boss when I help our company in a purchasing decision.

Scatterplot Demonstrating the User-as-Product Model

If those who get the most value are least likely to pay (i.e. if your graph looks like the image above), don’t fret, because you can still have a successful User-as-Product business. In a user-as-product business the key is figuring out how your free users can add value to your site to make it more appealing to those who are willing to spend money. Looking at the graph above, you want to use the users in the bottom right cluster to push the cluster in the top left further to the right.  If there aren’t any groups in your list who rate high on “willingness to spend” (perhaps because you assumed they’d find little value in your concept so didn’t add them) you’ll need to branch out.  Think of LinkedIn.  Originally, their users were the people posting resumes. These people do receive value from posting their resumes online (because they would show up in Google search results) but it wasn’t until LinkedIn added recruiters to their user base that they could use their free users to give value to a “willing to spend” group.

It’s important to remember when looking at all three of the above scatter plots that it’s much easier to push people horizontally than vertically.  That is, it’s easier to add value for your users than to change their general willingness to spend money.  What’s important then, is figuring out which users need more value and then deciding whether that value should come from your product or other users.

Have a question?  A quibble?  Let me know, below!

This entry was posted in Web and tagged , , , . Bookmark the permalink.