from the road to innovation

This post emerged as the second part of the answer to the initial question of a fellow business developer at a large non-US social network: “how can we better monetize our social network“. As I’ve mentioned in the first part, there are several important questions to be asked before focusing on finding answers to the monetization question. The first and foremost question is: do you have a viable business model in place. Many companies that have successfully acquired a larger user base struggle because in fact they do not. In the previous post I have guided you through an action plan and now let’s look at the question “where can I build a business model for my social network” as opposed to “where can I monetize”.

Monetizing means either optimizing a currently working business model or adding new revenue streams. In the former case, monetization is often an issue of either adding value added features to improve conversion or add a differentiated product, or plain and simply an issue of setting up and improving marketing and sales. In the latter case, monetization is again very close to finding a viable business model – exactly what this post is all about.

Let’s start by looking at the business models of some of the currently dominant social networks. The estimates of the revenue contributions are mine – some of them grounded in conversations with founders or executives of those networks, others grounded in personal guesstimates. Let me know if you have more accurate data and I’ll love to include it!

• MySpace: >90% Advertisement/Promotion. They did have a search/contextual ad deal with Google that brought them almost one billion, but Google massively overpaid.
• Facebook: >80% Advertising, 10% Virtual Goods Marketplaces. Facebook is good at targeting and makes accessing the ad platform extremely easy!
• Linkedin: Advertisement ~25%, Marketplaces ~40%, Market Research ~20%, Subscriptions ~15%. LinkedIN has targeted ads as well and has achieved amazing results with their recruiting tools, in particular the job platform and the corporate solutions around jobs.
• XING: Subscriptions 80%, Advertisement 5%, Marketplaces 15%. XING monetizes fantastically on subscription and successfully opens new channels, like jobs and ads.
• Plaxo: >90% Subscription. Plaxo still is in part in a tactical mindset, but after being acquired by Comcast makes contact brokering and people search the basis of its business model.
• Tencent: >70% Virtual Goods Marketplaces. Tencent has an amazing story – it’s fantastic how big the margins are in virtual goods and how this Chinese company is a role model for virtual goods monetization. Respectively, the market value of the company is adorable.
• Sermo.com: >70% Market research/Information brokering. Sermo aggregates discussions of its audience consisting mainly of MDs and brokers insights to e.g. hedge funds who will then use these to steer investments.
• Ning: provides social network infrastructure, SaaS model. Therefore, Ning compares more with e.g WordPress.com and is not at the center of attention.

Let’s now focus on the major revenue streams and identify how to build a business model. The major revenue streams are subscription, advertisement/promotion and marketplaces (this includes virtual goods).

Subscription

A subscription model always centers around a core value which usually profits from the characteristically present network effects. At XING and Linkedin for example, a core value is finding people, since social networks are at the core a people directory that facilitate connections and communication. Subscription models successfully place restrictions on either fundamentals or on virtual factors. Fundamentals are e.g. restricted search results, restricted access to people data, or restricted communication. Virtual factors include things like status in the network (e.g. ambassador badges etc.). Foursquare is doing a pretty good job currently at establishing core virtual factors that might be monetized at one point.

There is an interesting twist on a subscription model: the core user is not necessary the target subscriber. The target subscribers are from specific user segments that particularly benefit from the value created by the network. For example, recruiters get a huge resource and it comes almost for free (compared to e.g. access to traditional people databases).

Considerations:
In the search for a viable market in establishing subscription plans, these considerations may help:

• Consider customer value: Who is getting most value from network? are there any audiences that might benefit from accessing members (and just do not know of the opportunity)? Try to draw stakeholders on a map and identify the degree of value created from your product. Based on these insights, use the next two considerations below to dig deeper.
• What kind of features can you add to monetize? Analyze user activity and then dig down to the purpose of activity. E.g. if there are vivid discussions in your forums and you find out that poeple are motivated to engage in discussions because the gain status, you may offer special badges, icons, or add special features on profile pages that help members better visualize and communicate their status. For other motivators, e.g. the contributors’ success at generating sales leads, then it is a good thing to look into how you can facilitate sales, e.g. add fields “looking for” and restrict who can search those.
• Can you build a business around niche audiences (while niche does not mean small size…)? What valued added features or functionalities can you build to monetize in niches? Ask yourself, for whom could I generate relevant search results? For whom can I enable relevant communication channels? Who could be interested in analytics that I can generate? And in particular, consider the size of the market. Some examples: People search and restricted communication with non-contacts at Linkedin, XING is mostly beneficial for recruiters and salespeople. Because of the size of the network, recruiters can tap into people databases even bigger than the ones offered by e.g. Monster.com at significantly lower prices. Instead of accessing a people database and paying way beyond a thousand bucks a month, recruiters can now do so for much less tan a hundred bucks a month.Even better: The people data in Linkedin or XING is typically much more accurate, because people update it much more frequently! Similarly for salespeople, their challenge is to identify and contact relevant leads. People search is a strong enabler – also for marketing people who send out targeted advertisement to potential prospects. So it is no wonder that Plaxo is building it’s business model around restricting access & communication to non-contacts as well. Another example is Sermo, which is building a great model around analytics: They provide intelligence from discussions between MDs which can be highly relevant, e.g. for hedge funds who want to get a grip on current opinions on the success rate of a new drug. Every company I know collects market and consumer insights to make strategic decisions – some activity in social networks is an ideal spring of data to gain consumer insights: trends, preferences etc. Ask yourself, do I have data that can be relevant for someone? And if so, how do I make it actionable? Remember to leave the building to get these answers.
• Consider building valued added features and functionalities for your core users. Three categories to consider are: Vanity, status, and voyeurism. Vanity measures such as “who has visited my profile” are powerful conversion tools. Also, some of the most active members in communities are status-driven; so providing visibility, special branding or any other form of advantage can be very successful at making them pay for such benefits! Voyeurism is a powerful lever, but a dangerous one (privacy!!). Imagine you’d have to pay facebook a couple of bucks and in return you could see every photoalbum of every person on Facebook. I’m sure there’d be lots of demand but would it be a good move – judge for yourself…

Some pitfalls to watch out for:

• Payment process: This is a big one. Do not get caught up in payment optimization until you have proof of product/market fit and are cashflow positive. Do not finance large payment projects for subscription products when you are still bootstrapping or when you’re on angel funds. Don’t do it on Series A money either. Payment is a bi*** and will bite you if you go for complicated solutions. So take the easy way. Whatever you do, however, you should definitely have auto renewal!
• Up/download & other standard features: Here’s what most companies struggle with: Many companies believe to add a price table and limit some features is enough to build a business model. In some cases, it is, but try to find out where you can create real value. The people search in a network like XING is extremely valuable for the points outlined below. In addition, ask yourself if there is a large enough market. In the case of people search, there is. So when you discuss features like up/download, etc., always ask yourself: are those really powerful conversion drivers? In particular, consider that this is not a major differentiator in the face of competition: such features are typically easy to imitate. Besides that, up and download eats up bandwidth and storage, both of which costs you real money.
• Subscription waterfall: Keep in mind that viable businesses always attract competitors and subscription is something dynamic. You will have to add new benefits and make other currently restricted ones free over time. There should be some additional value added features that can add over time. If you can’t think of any, you may want to continue looking for more answers.
• Monthly vs. yearly payments: monthly prices allow you to increase conversion because you can reduce barriers through lower prices and lower commitments for your customers. However, you wold ideally always want to go for longer bundles. Three-month packages are good, one-year packages are better. First, although you cannot book the revenue you can book the cash – valuable cash that you can use to pay bills. Second, yearly models allow you to keep an eye on churn and therefore raise the lifetime value of the customer. There’s so much more to say to that, but for now, keep in mind that you should try to bundle longer period packages… If you have any questions, contact me at any time.
• Sales & subscription: Subscription models are not ideal vehicles for a sales organization when prices are low – you cant’t scale a salesforce that sells packages of subscriptions that cost a monthly 15$… A sales should bring you at least $200-300 before it starts to pay to set a salesforce behind it. And in these price ranges, sales will mean telesales. Having a salesrep in the field to sell $300 subscriptions will not make sense. To calculate more precisely when it scales to have a sales force, you need to consider more details. In particular, you need to consider the customers lifetime value. This can be complicated – contact me if you are struggling with this.

Advertising/Promotion:

Advertisement performance has suffered because of a very low purchasing intent of consumers while on the site. I would say the CPM is dependent on two factors: First, the average purchasing intent of the site visitor/user and second the purchasing power per user. You can break this down into user segments you are serving to get a better picture of your earnings potential in advertisement. Andrew Chen has a great piece on monetizing via ads in which he identifies 5 factors that determine your advertising CPM rates. Great read! Evaluate a possible average CPM, then multiply this by the relevant page views (only page views from sites where you will serve ads).

Considerations:
In the search for a viable market in establishing advertisement/promotion models, these considerations may help:

• Is there a viable market for me in advertisement, anyway? Only if you can think of reasons why you’d be able to achieve high returns from ads you should make this a core of your business model. Ads is a nice addition, but in most cases, ads do not scale.
• The straight forward ones Display, contextual and keywords: Although have have the option to manage you inventory of display ads yourself, it’s probably best to choose an ad network. There are many posts on the web that will guide you through the process and give you helpful tips. For contextual ads, I would go for Google AdWords. Again, theres many articles on the web that will guide you through the process. What I feel has some unrealized potential is keyword ads. As far as I am aware, there is no social network that runs it’s own keywords. But consider for example Linkedin. People go and search for expert answers on Linkedin all the time. Wouldn’t it be natural to have their own keyword ads? Keywords work well for searches about specific answers or topics. Keywords don’t work well for people searches (names). The trouble with keyword advertising is that it is quite some technological nut to crack! Nevertheless, it is worth considering.
• Consider branded experiences or a special product for a brand. If you reduce supply of a special feature (make it scarce), you can attract high prices. Let’s assume for example that Facebook would only allow 10 companies a month to start a fan page (as opposed to making them entirely free like they do). The result would be: Companies would pay fortunes to have the exclusive opportunity to leverage their brands on Facebook. StudiVZ, a German Facebook, did this quite successfully – they provided branded pages at placement prices well above $10k/day. Of course, the size of your network and the activity in the network drive the prices. So this may be a viable road to explore. Ask yourself if you can leverage the interests of certain companies, brands or products in a special way. A note of caution: in most cases, this is a lemon squeeze model, which means that it is a way to monetize what you have, but probably not in a sustainable way. In most cases, providing brand experiences will not be a scalable option and will not be your vehicle of choice to build a business model.
• Consider Lead Generation. Lead gen can be a very powerful way to monetize and may support a scalable business model by itself in the case a company can prove to consistently deliver better leads over time than alternative sources. Qualified leads can earn high prices, $10-20 are no rarity. In order to find out if you should build your business around sourcing leads, consider factors like: where’s the demand for leads? how big is demand? is it growing? What assumptions can be made about the behavior of the size and growth of demand over time? What other sources of supply are there and are we positioned to outperform in terms of quality?

Some pitfalls to watch out for:

• Overestimating the returns from ads: There typically is a low purchasing intent in social networks. CPMs are extremely low, available inventory is extremely high and platforms like Facebook make it very easy even for small companies and individuals to post ads. So it wont be easy to fill your inventory and it wont be easy to achieve acceptable CPMs. If you have high unused inventory, consider advertising for your own company in the unused spots. Use the ad space to promote new features, raise awareness for your premium products, etc. In order to do so, you will need to have an ad network that is willing to grant you unused inventory for your own purposes.
• Scarifying user experience: Many users are used to ads and you can safely assume to introduce ads will discourage people from using your site. However, there are some caveats. First, advertisement is a cultural thing and it depends if you launch ads right from the start of if you choose to launch ads later. In some markets that are more defensive against advertisement you will need to have a very clear communications and PR strategy before introducing new ads. The second corollary is, of course, that at a certain point, users get fed up with ads and ads actually destroy the user experience. There’s a fine line, be aware of it.

Marketplaces:

The best example for successfully established marketplaces are the jobs marketplaces from Linkedin and XING. Both networks have created an intersection for supply and demand that fits to their site culture. While these “classifieds”-like markets are more traditional, they typically scale well in a sales organization. Sales comes at the expense of margins, of course, but there might be other models you can play with that will compensate some of that. In the Asian markets, networks like Tencent, Mobage-town and gree.com have a lot of fun with virtual gifts marketplaces. Facebook’s revenue from virtual gift is rumored to be in the several millions as well. While virtual gifts may not work in all environments, they are awesome in terms of margins – and good margins are awesome for nice valuations.

Considerations:
In the search for a viable market in establishing marketplaces, these considerations may help:

• Consider sourcing and bundling products and services and making them available in your own marketplace. For example, Ancestry.com is a special type of social network that allows you to create family trees. They provide you with access to records that they have collected and aggregated from all around the world. They basically provide access to a unique source of information and they build their core competence around collecting and aggregating historical people data and creating and connecting family trees. I took this example into the marketplaces section even though Ancestry.com is monetizing through a subscription model. By choosing subscription, they effectively restrict access to their marketplace – which is one special subform of marketplace monetization. Ask yourself what kind of products or services you can source, aggregate or combine for your users, and you may hit a very valuable, sustainable and profitable business model! Try to identify potential partners, think of possible joint ventures or alliances, most of all, dare to think big. Here some keywords you may want to consider for this excercise: SMS, voice calls, ticketing, printing, data, reports, virtual goods, …
• Consider opening marketplaces for third parties. Enabling a marketplace for third parties is a second option – it is basically a form of advertisement in many cases. Ask people: What would you sell in my social network? Then ask yourself: Can I bring enough relevant traffic to such an offer? Can I do it repeatedly? How does traffic and traffic relevance behave when I increase supply? For marketplaces, consider placement fees (e.g. per ad or bundles), performance fees (per click, per impression, etc.), and include time of visibility or accessibility in your pricing model. Also, you may identify product variations that would justify different pricing to engage in price differentiation.
• Consider transaction or service fees for member-to-member marketplaces (classifieds). Transaction fees can be a great source of recurring revenue – in particular when looking at member-to-member transactions. Consider the istockphoto model (ok, it’s not a social network). istockphoto allows you to buy stock photos at low prices directly from the photographer who uploads them. For the transaction, istockphoto takes a certain percentage of the price. The classical example for this, of course, is eBay. For every transaction, eBay gets its share as the market maker.
• Gauge the feasibility of virtual currency. There is no real standard in virtual currency (amazingly so, actually), but there have been many attempts (e.g. linden dollars, opensocial currency, facebook credits). Virtual currency can be a great source to capture transaction costs. In particular, if you own the currency, you are practically the central bank yourself. Powerful and a little bit scary maybe…
• Focus on value created, and explore numerous potential markets. It helps to gather user feedback, because there are always visionaries out there who love to share their visions for your product with you! Sometimes, these are unexpected. Listen very carefully to them!

Pitfalls Marketplaces

• Overestimating your ability to generate relevant traffic. You need to know your usage numbers and need to be have some experience with results of prior traffic routing experiments. You can easily test your ability to generate relevant traffic – so do it before you make a business model decision and use the results to support your choice.
• Forgetting hidden costs: do not forget about costs of sourcing, prospecting and acquiring supply! Many times, you will have to educate the market, in particular when you offer new products. Companies purchasing habits change infrequently (e.g. once or twice a year when budgets are discussed) and even if they understand benefits, it is hard to get them to a trial.
• Payment: Beware of sophisticated payment structures or features (see also payment in pitfalls for subscription). Establishing virtual currency has many pitfalls. Unless you have significant resources to invest, focus on established technologies and/or piggyback on the developments of others.
• Missing feedback and lacking feedback processes: Consider the following for feedback: while uservoice and getsatisfaction are good tools, they inhibit capturing a lot of valuable feedback that gets lost when customers decide against making the tradeoff between clicking through to the feedback forum or not to do so. You will want to make it as easy as possible to drop you a line – therefore, provide a one-click feedback option (like e.g. on startuply or at teuxdeux after login or a service like kampyle). My suggestion would also be: Don’t call it feedback. You may want to use something like “ideas? submit here” or “yell at us” or similar. Feedback is too corporate in many cases (see here). Of course, a downside of providing instant feedback access is that you will you get a lot of support requests like this. Take this into consideration and establish a process that handles multiple user feedback, ideas, support requests, bug reports, etc. Make sure you follow up on submitted feedback – at the least send a thank you not via email. In general: Cater for all the feedback you can get, take it as the most valuable currency there is – even more valuable than money.
• Not considering cannibalization: Marketplaces can cannibalize each other. Consider the following: You have a jobs marketplace with paid placements and later launch a free classifieds market for members. Since members can now also post job ads in that classifieds market, you cannibalize your jobs marketplace. Make sure to clearly analyze and plan for potential cannibalization effects.

Although I know the above offers just a limited view on many more aspects, I think there are some good starting points. So what do you think, are there other considerations that should be included (I’m sure there are many!)? Did you make any experiences which should be added to the pitfalls for each model? Have you heard of other models that perform well as business models? Please let me know in the comments!