jump to navigation

Game mechanics applied to social media: “Easy to Learn, Hard to Master” March 28, 2007

Posted by jeremyliew in Consumer internet, Internet, social networks, start-up, startups, user generated content, web 2.0.
18 comments

Jay Gould (President and co-founder of Bolt.com) recently commented on one of my posts and posted himself about Amy Jo Kim‘s fantastic work on game mechanics and how they apply to social networks and user generated content. I was first exposed to her thinking during her presentation at the 2006 Etech conference last year (I’m currently at the 2007 Etech conference). Its truly thought leadership work and I urge you to think about how it might apply to your business.

Kim outlines the “five key mechanics of game design”;collecting things, earning points, providing feedback, exchanges, and customization; and then thinks about how these can apply in other consumer applications, especially those with a user generated component.

I’d leave it to you to read her specific material, but I think the broader idea of applying best practices of game mechanics to shape behaviour on social media sites is very interesting. Basically, the key question you’re trying to answer for a first time visitor is “what am I supposed to do here?”

One often repeated game design adage is “easy to learn, hard to master“. This is true of the best social media sites. At Myspace you quickly learn that the game is to get the most friends. At Piczo you quickly learn that the game is to have the coolest pages. At Stylehive it is to submit cool and fashionable clothing and furniture. At Yelp it’s to write useful, funny or cool reviews.

New users understand the game because of the smart application of Kim’s mechanics of game design. They see what you “score points” for (i.e. what metrics are measured on each page), what is “collected” (i.e. what badges and other icons are displayed on each page), what actions or things prompted “positive feedback” from other members of the community (what is “celebrated” through comments and for what users are thanked), and they look to the sites navigation and design as to cues as to what they should be doing.

This argues for an “application specific” approach to building social media, rather than a “platform” approach. If you have a platform that can and is used for many different purposes, a new user doesn’t necessarily know what to do when they get there. An “application specific” approach makes the game “easier to learn” because the cues all point in one direction.

Stylehive, a Lightspeed Portfolio company, and ThisNext, are great examples. At their base, they are social bookmarking products, similar to Kaboodle or eSnips, that can be used to bookmark anything on the web. But a casual visitor to the site quickly learns that this site is about clothing, accessories, furniture and fashion, and as a result other items (new stories, clever photographs etc) don’t get added to the hive.

Similarly, Digg as a platform can support any form of media being submitted. But it started out mostly technology news based, and it has controlled its growth by means of its navigational taxonomy. As the user base grows and clamors for more nodes to be added to the taxonomy (or adds new material regardless), Digg expands its scope, with pictures likely to be added in the future judging from user demand.

If you’re running a social media site, ask yourself if a new user will quickly “get the game” and then want to “get into the game”.

Web 2.0 marks the decline of Ebay and Amazon March 26, 2007

Posted by ravimhatre in Consumer internet, Ecommerce, Internet, Lead gen, Search, start-up, web 2.0.
12 comments

Om Malik is on to an important trend in his recent post  regarding the marginalization of Ebay, Amazon and other legacy ecommerce marketplaces with the advent of e-commerce 2.0.  Given the emergence of new and better merchandising technologies, more intuitive and comprehensive product search services, and the proliferation of contextual and performance-based advertising channels,  small and mid-sized merchants are able to establish rapidly growing web outlets more easily than ever before. 

In the first generation of ecommerce,  marketplaces with recognizable consumer brands (like Ebay and Amazon)  could offer small and mid-sized merchants access to large pools of customers.   However, there was a significant premium charged for this access – usually 10 or more percent of the transaction price. Bear in mind that the typical merchant will have total gross margins of no more than 20-30 percent. 

Like many net-based ecosystems we’re now witnessing the emergence of an open environment to replace first generation “closed”  marketplaces or communities.  Instead of listing on Ebay or Amazon and relying on their brand to attract customers and their standardized merchandising and search to drive purchases, a merchant can now easily build a product website that will drive organic traffic from vertical and horizontal search engines picking up their unique product content and also utilize a variety of performance based advertising channels including comparision shopping lead-gen sites (the top 10 sites delivered over 100 Million shopping leads to merchants in January 2007)  as well as  search engine keyword marketing to acquire new customers.  These channels are less expensive and drive significantly more customers and purchases at higher margins than legacy marketplaces.  

From a VC perspective, we believe a key requirement to making this work is the emergence of next-generation product search services that tame the Internet’s infinite shelf-space and provide consumers with truly comprehensive product search results through an interface that is highly intuitive and digestable.  Several start-ups are intensely working to solve this problem such as TheFind (LSVP portfolio company) Become, and ShopWiki.  Let us know what you think of their services. 

“High Concept” startups March 26, 2007

Posted by jeremyliew in Consumer internet, Internet, start-up, startups, web 2.0.
14 comments

I missed two posts earlier this month with the same general theme, that the best way to come up with a new web company is to take an old idea that works, and recycle it.

Marc Hedlund (founder of Wesabe) says to create a new company, take a UNIX command and webify it:

…find an old UNIX command that hasn’t yet been implemented on the web, and fix that. talk and finger became ICQ, LISTSERV became Yahoo! Groups, ls became (the original) Yahoo!, find and grep became Google, rn became Bloglines, pine became Gmail, mount is becoming S3, and bash is becoming Yahoo! Pipes. I didn’t get until tonight that Twitter is wall for the web.

The comments to his post are worth reading if you are into UNIX humor – it mostly went over my head. (I also skip the Ubuntu posts of Digg!)

Jason Kottke says (excerpting – go to his post for the full text):

…take something that everyone does with their friends and make it public and permanent. (Permanent as in permalinked.) Examples:

* Blogger, 1999. Blog posts = public email messages.
* Twitter, 2006. Twitter = public IM.
* Flickr, 2004. Flickr = public photo sharing.
* YouTube, 2005. YouTube = public home videos.

Both Jason and Marc are essentially advocating “High Concept” Startups.
In addition to curent online and software/OS behaviour, another good place to mine for new “high concept” ideas is the offline world.

“High concept” movies are movies that can be summarized in one sentence, often referencing another well known movie or book e.g. “Superfriends in 19th Century London“, “Schindler’s list in Rwanda“, “Jaws in space“, or “Heart of Darkness during the Vietnam War“.
[rockyou id=61482296&w=400&h=300]

Hence by analogy the “High Concept” startup (e.g. “List serv on the web“, “Public IM“, “Flickr for Video“, “Myspace for Baby Boomers“, “Bumper stickers for the web” etc.).

There is a lot to like about this approach to building a new company. The first is that you know that you’re targeting a large user base because you can observe an existing large market. The second is that adoption can be very rapid because you’re taking advantage of behavioural metaphors that people are already used to – they “know what to do”. (The importance of this concept can’t be overemphasized. I’ll blog more about this later). The third is that monetization models can often be predicted by analogy – a low CPM massive scale advertising business in client form will likely be a low CPM massive scale advertising business on the web. Past can be looked to as prologue.

The challenge of the “high concept startup” is that it is often not all that novel. There may well be multiple other companies targeting the same idea. There is nothing wrong with competition, but this creates a greater level of murkiness in the water so that often the “best” product is not the winner. Sites with early adoption – whether due to Distribution, Virality or just plain luck, can end up pulling away from the others. In all startups, execution is at least as important as the idea itself, but this is even more true for the “High Concept” startup.

Google CPA will crunch lead gen arbitrageurs margins further March 21, 2007

Posted by jeremyliew in advertising, Consumer internet, Lead gen, Search, start-up, startups.
19 comments

Today’s release of Google’s Cost-Per-Action (CPA) beta has generated a lot of attention. Most are focusing on the impact on affiliate networks such as Commission Junction or Link Share as the test is currently confined to Adsense ads that show up on the Google Publisher Network.

I’m waiting for the other shoe to drop. The next logical step is to have these CPA ads show up as Adwords next to Google’s search results.

This presents a direct and present threat to many lead gen businesses, especially those that rely on CPC to CPA arbitrage as their business. I posted on the future of lead gen in January, where I noted that, simplifying substantially, lead gen comprises three processes:

1. Acquiring traffic (e.g. from paid search, organic search, brand advertising, banner advertising, distribution deals etc).
2. Converting traffic to leads through a form-fill process
3. Finding the highest value for a lead among multiple buyers (ie having a network of advertisers and knowing who placed what value on each lead)

Google’s current beta will essentially eliminate the arbitrage opportunities in part one of this value chain. Companies driving the majority of their traffic from organic search and (long term) distribution deals will be less affected, as will those who add value to the process by qualifying users and directing them to the best matched vendors as leads. But those whose core competencies are in clever media buying will be pressured because a CPA model shifts the risk out of buying CPC and CPM media and converting to lead forms.

There are a large number of lead gen companies that have grown to over $100m in revenue. These have grown to their current size by being well managed, and building multiple sources of traffic and an efficient mechanism for matching leads to their highest value.

Smaller “mom and pop” lead gen shops that depending on buying traffic through banner advertising and CPC advertising to landing pages and selling these leads to a small network of buyers will find their margins under increasing pressure if their clients can disintermediate them through Google’s new products.

UPDATE: Some very insightful responses posted in comments that I will attempt to summarize as “you’re assuming more efficiency exists than actually does, thats why this will still create a lot of value”. Its a fair point. If you read this in RSS, its worth reading the comments.

Time Rich or Time Poor? March 19, 2007

Posted by jeremyliew in Consumer internet, Ecommerce, Internet, Lead gen, Search, social networks, start-up, startups, user generated content, VC, Venture Capital, video, web 2.0.
62 comments

Broadly speaking, there are two types of internet users, Time Rich (more time than money) and Time Poor (more money than time). I’d speculate that many of the readers of this blog fall into the Time Poor category, but the vast majority of internet users fall into the Time Rich category. If you’re starting a new internet company, its important to know who your audience is, and to make sure that you don’t let your own experience and that of other Time Poor people guide you wrong.

Time Poor

Time Poor people use the internet to get things done. They are very task focused, and their favorite websites help them use their precious time more efficiently. Great examples of websites built for the Time Poor include search engines, first gen comparison shopping engines (trying to find the lowest price as quickly as possible), ecommerce and lead gen sites where the purchase is more functional than emotional, and many of the “social news” websites that filter the news for you.

If you’re building a website for the Time Poor, your focus should be to minimize their time and pages on site. As a result, business models around e-commerce, CPC and lead generation are good matches for these sort of site – it aligns both user and site around getting to a transaction as quicly as possible. Depending on what you do, you may even be able to charge a subscription as well.

Time Rich

Time Rich people use the internet to kill some time. They are bored. They are willing to be diverted and entertained. Great examples of websites built for the Time Rich include broad based social networks, targeted social networks, picture sharing sites, anything celebrity related, anything sports related, social shopping sites (recreational shopping), social discovery websites that suggest new sites to you, all video websites and causal games websites.

If you’re building a website for the Time Rich, your focus should be to give them options to explore. Links density is the name of the game – more links means more clicks. Suggest a next click at any natural pause point, and keep people clicking within your site. Stimulate communication and community – it keeps people engaged and coming back. Give people reasons to bookmark you and come back often with fresh content and evergreen favorites.

You’ll likely monetize through advertising – sponsorship and CPM as well as CPC. Subscriptions may work for you too if you have certain features held back. If the products you sell are bought spontaneously, then ecommerce may also work for you. But don’t fall into the trap of creating extra pageviews for your own benefit and not that of your user (e.g. by splitting articles across multiple pages, or creating extra steps in a process to edit a profile page) as your users will wise up to your game soon enough. Time Rich does not mean unsophisticated. Your users spend enough time on the internet, and on your competitors sites, to know what are the best practices.

Know your audience when you build your site, keep the target clear, and you’ll have a better chance of meeting their needs.

UPDATE: New visitors, if you liked this post try the second most popular post, Three Ways to Build an Online Media Business to $50m in revenue

Three ways to build an ecommerce business to $500m in revenues March 16, 2007

Posted by jeremyliew in Consumer internet, Ecommerce, Internet, start-up, startups, VC, Venture Capital.
31 comments

Since my earlier post on “Three ways to build an online media business to $50m in revenues” was well received, I thought I’d examine the e-commerce industry as well.

The margin structure in most (physical) ecommerce businesses is dramatically different from that of online media businesses. Whereas online media businesses can enjoy gross margins of upwards of 90%, and net margins (at scale) of 50% or higher, many ecommerce businesses have gross margins in the 20-40% range and net margins (at scale) in the 5-10% range. As a result, ecommerce companies have to grow to a much bigger top line to achieve the same value. We’ll target $500m in revenue to get to net income in the same range as a $50m revenue online media business.

There are three ways that an ecommerce company can get to this scale:

1. Build up $500m in sales in a single vertical. You’ll need to ensure that the vertical that you’re addressing is large enough in online sales to accommodate your size – books, music, consumer electronics, shoes and groceries are all good examples. You’ll likely need to be number one in your category, which implies industry leading cost structures. You’ll probably be holding inventory and operating multiple distribution centers, dealing with returns and generally operating a very large scale business that gives you certain margin advantages because you’re one of the largest retail channels for your suppliers.

You’re probably able to spend on building a brand (vs performance based marketing only) and you likely think about customer lifetime value. Hence you may be willing to pay more to acquire a customer than you’ll realize from your first transaction since you sell a product that is bought frequently. As a result, you obsess over customer service because you need your customers to have a great experience and have confidence and recall to buy from you again in the future – ideally by typing your URL in directly into their browser.

$500m in a single category is a lot. In 2005, according to Internet Retailer, only four pureplay ecommerce companies exceeded $500m in online revenues (Amazon, Newegg), Overstock and Netflix. The remaining 22 companies who had sales over $500m online were very large multi-channel retailers like Office Depot, Gap, Dell, Circuit City and Walmart.

Other companies who likely have already reached this revenue level since then, or will soon, include companies like Zappos, Freshdirect, Drugstore.com and Buy.com

2. Build up $50-100m in sales across each of 10-5 verticals. This could be by being a smaller player in a larger category (such as the verticals discussed above), but you’re likely number one in a smaller category. Say Ski gear, nutirtional supplements, autoparts or power tools; smaller categories than books or shoes, but still pretty big. (Note – the links are to examples of companies that are at or could get to $100m in sales, but they are not parts of companies with multiple verticals each doing $100m in sales).

Since in many cases there are not that many synergies across categories (little or no ability to leverage supplier relationships for example) you may be a result of a rollup to get to critical scale. You may see some ability to leverage your distribution infrastructure, but in many cases the pick, pack and ship needs of different products are quite different and may not support shared infrastructure (small vials of pills looks very different from aftermarket auto parts). You likely hold inventory and operate your own distribution centers, but if you are in a category with large and unwieldy items that often get built to order, you may be able to dropship from your manufacturers.

In 2005, according to Internet Retailer, only 155 companies exceeded $50m in online revenues, and only 45 were pure play etailers, including Blue Nile, Redenvelope.com, Shoebuy and Furniture.com. However, there are a number of companies who are taking this approach, including Musician’s Friend, Provide Commerce/ Liberty Media, and Blue Lava

3. Build up less than $5m in sales in each of 100+ categories. According to Internet Retailer, there were 479 ecommerce companies with sales over $5m in 2005, including Batteries.com, Junonia (plus sized activewear for women), iGourmet, artbeads.com and thinkgeek.com. As you can see, even relatively small niches can sustain $5m in sales. You may be able to rely on your manufacturers to drop ship, and you may need lower levels of dedicated resources against each category with less depth of industry merchandising expertise.

Rather than building a brand, you can rely more on performance based marketing, particularly paid and organic search and shopping engines. You may not even need to be number one in your verticals – if they are big enough you can still win some share of the market to get to $5m.

What is hard is getting to this level of sales across so many verticals. To be able to do this you need a level of shared technology and processes that can be applied across many stores. Winning becomes less about any one store, and more about applying best practices across all the stores. The challenge is in being able to enter a category cost effectively, and to run a store against low volume in a very low cost way. Processes and cost control become paramount because any sub optimal practices get magnified across 100 stores. Although no one has hit $500m in revenues through this approach yet, there are a number of companies who are taking a shot at this approach, including Mercantila (a Lightspeed portfolio company)CSN Stores, Netshops, Niche Retail, and others.

All these models are viable. As in most cases, the first $20m in revenues are the hardest! I’d love to hear from people on any of these paths.

More on building an online media company to $50m in revenues March 14, 2007

Posted by jeremyliew in advertising, Consumer internet, Internet, Lead gen, start-up, startups, VC, Venture Capital, web 2.0.
15 comments

As my previous post indicated, it is not easy to build an online media company to $50m in revenue. Depending on whether you’re broad reach, demographically focused, or can support endemic advertisers, you need to get to top 10, top 25 or top 125 levels of US website traffic.

A couple of interesting studies have come out recently that underscore how difficult this can be.

At the Online Publishers Association’s Forum on the Future earlier this month, Marketspace (a consulting firm associated with Monitor) announced the results of their research which showed that 92% of 2006 gross online ad spend in the US went to only four companies; Google, Yahoo, MSN and AOL. Although some portion of that ad spend was subsequently distributed to independent sites through ad networks (e.g. AOL’s Advertising.com, Google’s Adsense, Yahoo’s Publisher Network etc), that is a big proportion of the total. Furthermore, that is an INCREASE from the 88% that went to those four companies in 2005.

Now According to the IAB and PwC, internet advertising revenues for 2006 were estimated to be $16.8 billion, a 34 percent increase over $12.5 billion in 2005. So doing the math, that suggests that the online advertising that didn’t go to the big four actually DECREASED from $1.5bn in 2005 to $1.34bn in 2006.

For companies in the broad reach/$1 RPM bucket, this probably doesn’t matter much. Ad networks owned by the big four sell a lot of their advertising anyway. But for companies that target endemic advertisers, this is sobering information. To be able to realize RPMs in the $20 range, companies will need to have their own sales force. And if these numbers are to be believed, this sales force is actually competing for a share of a slightly shrinking pie.

These numbers don’t quite match to the numbers in Avenue A/Razorfish’s 2007 Digital Outlook report, which is well summarized at Paidcontent, but they agree directionally. Avenue A says that portals have increased their share of online ad spend by 85% from 2005 to 2006, from 13% of overall ad spend to 24%. (This report breaks out search and ad networks separately – the big four would be a combination of these categories).

It would appear that advertisers are seeking consolidation in their spending patterns.

This isn’t entirely a doom and gloom story – online advertising revenue as a class is still growing at 34%, and $1.34bn of online ad spend among the independents is still plenty of revenue to go around. But it does underscore the need for websites to have a compelling story for advertisers, both about user targeting and about volume of traffic.

Speaking at Web 2.0 Expo; April 15-18 in San Francisco March 13, 2007

Posted by jeremyliew in start-up, startups, VC, Venture Capital, web 2.0.
3 comments

I’m speaking at the Web 2.0 Expo in April. Its a “how to” conference for people who are starting in or working at web 2.0 companies and companies that aspire to be “web 2.0”. The organizers have lined up a nice selection of workshops and speakers to cover seven tracks: Fundamentals, Services & platforms, Web Ops, Marketing and Community, Design & User Experience, Strategy & Business Models and Products and Services.

My session is in the “Strategy & Business Models” track on Wednesday April 18th at 1pm, but I will be around most of the day Tuesday and Wednesday.

Lightspeed blog readers who are attending can get $100 off their registration by using the discount code webex07mk3.

Feel free to come say hello!

Reducing (pricing) friction March 12, 2007

Posted by jeremyliew in Consumer internet, Digital Media, Ecommerce, Internet, start-up, startups, VC, Venture Capital, web 2.0.
3 comments

Josh Kopelman has a good post this weekend about the friction between free and one penny when charging consumers for goods that can be delivered digitally (e.g. articles, video, music, information etc). As he points out, price elasticity is not constant as price changes, but rather there is a huge step function (downward) in demand between prices of free and one penny.

A good example of this is the downloadable casual gaming space. The industry has standardized to a price point of around $20 for the unlimited version of a downloadable game. On average the industry realizes conversion rates of about 1% between the free (limited play) version and the pay version. Why do they not charge less and make it up in volume – after all variable costs are close to zero? Emperical testing has shown that price elasticity is relatively low once someone has decided to pay to play a game. The key friction point is between free and one penny.

Josh provides some different examples in his post which I recommend reading.

How to make the most of your pitch to a VC March 7, 2007

Posted by jeremyliew in Entrepreneur, start-up, startups, VC, Venture Capital.
7 comments

Thanks to a glitch in my Netvibes reader, I got exposed to an old but very useful post by David Cowan on “How to NOT write a business plan“. It sparked me to post on how to get the most out of your pitch to a VC.

The overriding point is that your intention in a first meeting is not to fully explain your business. Its just to get to a second meeting.

You are likely to have around 45 minutes with the VC after pleasantries have been exchanged, laptops booted up and projectors connected. That isn’t long. Resist the temptation to fully explain the intricacies of your technology, or to explain everything you’ve done. Instead, focus on the big “hooks” that will get an investor intrigued in what you’re doing. Most slides take about three minutes to cover, so try to keep your presentation to ten slides, and leave yourself fifteen minutes for questions, and to learn a little about the Venture firm that you’re pitching.

Ten slides doesn’t sound like a lot, but its an excellent discipline that will force you to really crystalize what you’re doing. Its also all you’ll have time for. Its very unlikely that your business is so special that you can’t do this. Its usually more of a comment on you than your business!

David’s post suggests what the 10 pages should be, which I’ll summarize here (but you should read his original post which goes into more detail):

1. Cover slide with complete contact info, and a tagline.
2. A mission statement (that is specific, achievable, but not yet achieved).
3. Team background, including key hires yet to be made.
4. Nature of the problem you address. Emphasize the pain level and the inability of incumbents to satisfy the need.
5. Product overview, including benefits.
6. Elaboration on the technology or methodology you have developed to enable your unique approach. If appropriate, mention patent status.
7. Early customer or distribution progress: traffic, revenue, number of customers/users/whatever, logos, testimonials, other key metrics.
8. Sales strategy. Show the expected channels and cost of customer acquisition.
9. Market size and competitive landscape.
10. Earnings Statement, historical and forecast (including total future financing requirements)

(The Appendix can be as long as you want but shouldn’t be a key part of the presentation – just used to answer questions)

For an internet company, I’d also suggest a live demo. If you’re an internet company seeking a Series A or later and you don’t have a site up (at least in beta) then I’d advise waiting until you do. The technology for most internet sites isn’t that hard or expensive, and its well worth getting something up so that a potential investor can better visualize the product and user experience.

Keep focused on your objective, to get a second meeting, and you’ll find yourself much more “on message” during your pitch.