I just watched this video of Matthew Thomson, VP of Platform at Klout, describing what I consider a cautionary tale for API providers.
He describes 4 main points for a successful API strategy: (1) Adapting APIs to business needs, (2) More developers are not always better, (3) Segment developers with pricing and (4) Evolve your API. Those sound fine and are generically sound policies. But at the 4-minute mark of this video, he lays out a core conflict that every API provider should carefully examine.
Klout has determined that the business model that is working best for them so far is in being the direct broker between brands and their agencies who want to reach Klout “influencers” as determined by Klout Score, True Reach, Amplification Probability and Network Influence. Well, if that’s where the money is (and who would know better than Klout?) then the developers in their community will want to build businesses to play in that space, too, placing them in direct competition with the mother ship.
So, if I understand what he is saying correctly, they are essentially making it more expensive for developers to use the API as compensation for the erosion of their core business. It’s a legitimately sloppy model for an early stage company — to keep their options open while they determine which branch of the business is more sustainable and poised for growth. Yet the inherent conflict between the two models (direct B2B and open API) is problematic on a strategic level.
Take a look at their terms of service. This sentence is so broad that I would not really want to build a business on such an API: “You shall not… use the API in a way that harms the interests of Klout, the Website, any of its affiliates, the API or its program, or other users of the Website.”
So… how is running a Klout Score campaign in competition with the Klout sales team not harming the interests of Klout?
Are you competing with your developer community? And if so, are you doing so fairly and transparently?
Living in the bleeding heart of the Eurozone crisis (il Bel Paese, aka Italy) as events spin out of control, fear reigns supreme and the crumbling of the old model accelerates, I can hear my grandmother’s voice sighing “Oy vey (אױ װײ)“, oh the pain…
The interlinked webs of huge banking and multinational corporate interests that express themselves through the financial markets have surfaced to protect what is theirs and are showing that they are more powerful than sovereign states.
I hope to post something more thoughtful about all this in a few days, but until then I thought I would post this little sculpture that sums up some of the emotions that we’re all feeling right about now.
What about you?
Dear Blog, I miss you.
First, I wanted to set this up with a couple of assertions, that location is a signal, as John Battelle defined it, and that this signal will be extremely useful when wrapped around social objects, in the way that Jyri Engestrom intended the term way back in 2005, and it’s every bit as true today.
This is a slightly more structured way of just saying that location becomes meaningful in context.
I think we can all agree that Color flopped its launch. It chose proximity-oriented photos as the social object upon which to base the serendipitous creation of affinity groups. The hope was that this activity would be so engaging that people would be motivated to invite more people to use the app, they’d use it very frequently in many locations and Color would thus have access to a hyperlocalized two-way channel into the lives of their users.
The idea is that they would then use this so-called anonymous data to create user profiles and a rich database from which to launch advertising, local promotions and news-oriented feeds. I say so-called because they neglected to understand just how identifiable photographs of faces are! (Yes, I’m looking at you Facebook.)
Another surprising oversight given the data-driven nature of the founders is that “for the average person, knowing their approximate home and work locations — to a block level — identifies them uniquely.”
Isn’t it amazing how fast anonymity breaks down?
As if this wasn’t enough, Color’s original user interface was unintuitive in the extreme and absolutely required that you use the the application with at least one other person. So, it flopped big time.
Now that Color has quietly withdrawn from the scene, it’s back to the drawing board to roll out a different application that will feed their hungry proximity algorithms champing at the hyperlocal bit, not to mention their investors looking for gorgeous pivot. How will they deal with privacy and can they find the secret sauce to make me want to share my location with nearby strangers?
My prediction is that they will not. Their approach is all wrong; it’s backwards. You cannot define yourself “much more of a research company and a data mining company than a photo sharing site,” as Bill Nguyen did and expect to have the wild imagination and fire in your belly to create an amazingly compelling social application that lots of people will love. He has some interesting ideas about the social stickiness of proximity, but it’s all wrapped around how much data he’s going to collect and sell to advertisers.
My next post will explore another proximity application…
All online businesses need to be looking for faster, wider and deeper distribution of their content or services. Think about your own strategy. Does your online product or service already have an API? Should it?
Your business is a platform.
In order for this model to work, you will need to attract top independent developers to apply their magical brains and coding chops to your business and to encourage a developer community to grow within the heart, mind, soul and body of your business.
You can call it a developer program, but do not mark it as a technical or support task, put it down as a primary business development strategy because that’s what it is, it’s rainmaking.
What you’ll want to do is to create the conditions for a community to form, thrive, produce and grow.
Think gardening, not engineering. Don’t manage, cultivate.
The goal is to populate your ecosystem with energetic, creative and technically gifted individuals, groups and companies that are highly motivated to invest their own time and money by building new businesses on your platform. As the successful services built on your API’s become important distribution channels for your business they will probably, in the aggregate, grow larger and faster than your own branded distribution.
Let’s get practical. I see five basic phases that must be addressed in order to benefit from a developer program (assuming, of course, that your API actually solves a problem that merits solving): creation, activation, retention, productivity and growth.
Creation: In more traditional business lingo this would be called lead generation or customer acquisition, but here we are concerned with preparing the fertile, accessible and attractive top soil in which to grow a developer community. You will need to:
+ clearly define the value proposition to developers;
+ identify a developer profile likely to understand and to be excited by your business and figure out the best channels in which to strut your stuff: niche blogs, social media channels, industry events, personal networking and SEO for starters;
+ plan self-hosted events and contests; and
+ provide great tools so that developers can kick the tires!
You must have someone in your organization that is fluent in the social skills of a tummler, AKA someone that knows how to guide, encourage/reprimand, engage/ignore, shout/whisper, cajole and generally spark online conversations and communities. You won’t get past this first stage without a tummler!
You are creating the culture in which the community will define and nourish itself. Read: Communities of Yes.
Activation: Once you have the attention and interest of a community, your focus ought to shift — make the community aware of its own existence by offering online and real world opportunities for members to get to know you and each other.
This is the most delicate and difficult phase. The quality of your business proposition and the awesomeness of your technology must be amplified as you make individual developers feel like they are part of an active community, that they have a real opportunity to create a viable business, are important to your company, are respected for their creativity, technical abilities and feedback.
Your program should:
+ allow for fast and easy registration;
+ permit cost-free to access to experiment, no obligation;
+ require payment only for deployed apps with no lockin;
+ offer quick start guides, examples, implementation ideas;
+ have a wide variety of devices available for testing at your hacker events;
+ provide support and direct channels to your team; and
+ be built on a reliable, robust and standards-compliant technological platform.
Sponsor developer pitch contests, hackdays and participation in industry events. Keep in close touch with your developers throughout their participation in order to suss out hiccups before they become problems, turn crazy ideas into tangible products and discover use patterns.
Retention: If you can achieve all this, you will have also built the loyalty necessary to retain the attention of developers that have become part of the community! This is an incredibly valuable driver of the entire program because these developers will begin to spread the word to others in their social and professional networks. The developer community itself can be great generator of buzz and organic marketing.
Productivity: Some developers will innately understand the market, their product ideas and the value of the API as good or better than you do. They will ship successful products and provide incredibly valuable suggestions, improvements and feedback. They will also inspire other developers and create healthy coopetition.
For these developers your job is to make sure that they know you are listening. Respond quickly and positively to their requests and questions and proactively add features that you know they will appreciate.
There will be other developers just as keen to work with you, but who’ll need more guidance and support. This can be very time consuming, so you must make sure to have tools such as wikis, forums or knowledge bases that allow these interactions to be as public and searchable as possible to reach the maximum number of similarly situated developers.
In the beginning is much better to have a lower number of high value success stories than many low value ones. The high value successes, over and above their direct business value, will bring in many more new developers just by word of mouth.
Track the statistics of your developer community based on the metrics that make most sense for your business. At a minimum you should be able to analyze the number of registrations, number of apps and projects, apps per developer, developer churn, support activities, most effective marketing channels and word of mouth growth. Look for bottlenecks and figure out how to improve your performance. Don’t forget to also keep tabs on your brand’s online buzz and reputation in the world outside of your own community.
Growth: If you’re firing on all cylinders, your organic growth numbers should already be happily trending upward. You can encourage further growth by getting better at the activities that I already described and by thinking about new ways that you can make your developers be more successful.
Here’s a list of other things you could consider doing to up your game:
+ promote top developer projects in your company’s marketing campaigns;
+ provide help with app store submissions when needed;
+ open doors to industrial partners that developers may not have access to — mobile operators, handset manufacturers or media companies;
+ provide expert sessions at your hackdays; and
+ provide a place for non-developers (B2B customers and end-users) to post job requests and product ideas. Then help put them in touch with the best developers in the community.
Want to see how it’s done? The most recent Music Hack Day in NYC is a great example. Check out the list of sponsors and the API’s that were hacked. And you don’t need to be in NYC or Silicon Valley to have a successful event, hackers can be found everywhere! Look what Max Ciociola e Stefano Bernardi did in Milano: HackItaly was a mega-successful event.
The beauty of this kind of program is that the direct economic benefit is just the start. Your developer community is also resource for market and technological research, recruiting new talent and strengthening your place in the larger ecosystem. And it’s fun
Despite evidence to the contrary, I’m thinking about tiramisù. I just got to Barcelona for MWC 2011. It’s not that I am craving dessert, you see I’ve got pull on the brain. (Tiramisù literally means pull me up, but it can be translated colloquially as cheer me up.)
Pull. In moments of transition, we struggle to come up with models to explain what is happening, and inevitably we are stuck using the concepts and terminologies of the past to describe the future. And to synthesize in order to simplify and get to the heart of it.
And so it is with this pull thing, four letters that will help us to figure out how to turn our world upside down, and it’s about time.
But isn’t there somebody somewhere doing a bunch of pushing (of information into the system) so that pulling things in is worth the effort? And won’t other verbs also be useful here? Attracting, repelling, bouncing, reflecting, tilting, reaching, passing, catching and chasing are among others that come to mind. Some of this pulling, and maybe one day most of it, may be done lazily, and by this I mean it will be the exhaust of other activities re-examined, newly interpreted and then repackaged for a particular purpose or context. A simple example of how this done today would be how credit card companies track your usage patterns and when your exhaust smells funny, out pops a block of your account.
So, I will be running around MWC looking for the stuff that is most interesting to me right now — technology and applications that allow mobile devices to be more aware of what’s physically near with simple, lazy, ways to hook into the internet of things. Mix that with a pull model of information flow and I’m suddenly paying attention.
The curation is king meme is having a day in the sun. This is the kind of article that editors love to publish — it declares a winner in a zero-sum game between the old champion Content and the contender, Curation. The article posits that Old King Content has been ousted “because it isn’t scarce. It’s everywhere, it’s overwhelming, and it’s gone from quality to noise.” Of course there’s too much for any one person to slog through, but if content is of secondary value, then what are we curating?
But Rosenbaum has got it very right on one crucial point:
He also seems to be saying that this victory of curation is a recent event. I’d take issue with that. The first Yahoo! directory was, at the time, state-of-the-art curation, not particularly scalable, rather unsocial and opaque, but it was curation. And, of course, Google’s revolutionary insight to use inbound links as a fundamental page ranking metric took people-powered curation into new territory.
Right on. Let’s drill down into the humanity angle.
The word curate derives from the Latin curare, meaning “to take care of“, something that we humans do for people and objects that are meaningful to us.
The Tummelvision crew (@heathr, @debs and @kevinmarks) have been exploring this human side of how digital communities and conversations are born and thrive. They’ve adopted “Tummel” as their verb of choice. It’s a Yiddish word used to describe the act of catalyzing others to action and I highly recommend tuning in to their show.
Just as Suw Charman-Anderson blogged about this way back in 2006, it still is fresh and true today:
Curators already exist. Some are people: Bloggers who sift through tonnes of stuff in order to highlight what they like, and who, if you have the same taste as them, can be invaluable to discovering new things to like. Some are aggregators: Site that gather lots of little bits of stuff and present them in aggregation and help us find the bits that the majority find to be good. Some are algorithms: recommendation systems and search.
What would I do without Kevin Marks’ curation?
In Three Tips for Curating Your Audience, John Sviokla writes about taking care of your audience, also known as customers if you happen to be in the business of selling stuff. (I’d take issue with the term audience for these reasons.)
Taking care of your curators means finetuning the curation. In socially-powered services like Facebook and Twitter we curate the people that we choose to friend or follow by keeping them interested in what we have to say and we also adjust our flows by un-friending or un-following in a more or less constant feedback loop.
And now, let’s bring on the ants!
Did you ever wonder how ants manage to organize themselves into those superhighways of efficiency to bring the food back to the colony? Well, that’s stigmergy in action and this survival-oriented curation mechanism is how we create our own superhighways of inter-linked information I actually wrote about this back on October 16, 2003. Take a look at “What do blogs and ants have in common?” to learn more. It’s a much shorter post than this one!
What are the “economics” of curation?
I was talking about this notion of online generosity with my friend ocrampal the other day, and he questioned whether curation was primarily generosity-powered as I was saying. In fact, I think he’s right, the answer is more subtle. Social networks and community-powered curation works not so much from generosity as from the very good deal that participation offers.
Let’s call it fair price curating. In other words, contributing to the system is the price you pay to get value out of the system. More often than not I get more out than I put in. It feels like I am getting my curated information below cost — that I’m gaining much more than I’m spending in terms of time, effort and satisfaction.
I guess that’s because once everybody contributes their info bits, it costs the system very little to redistribute the information in infinitely variable combinations as requested by huge numbers of people and ever more sophisticated technological tools.
I haven’t yet read Clay Shirky’s new book Cognitive Surplus: Creativity and Generosity in a Connected Age, but I intend to and would be curious to see how he sees it.
Both questioner and answerer must have a satisfactory experience for the service to work. In fact the service really hangs on the experience of the answerer. The questioner has an immediate and present need or interest — not so the answerer. His or her motives for participation have to be incentivized or contextualized by other means.
Are earning a good reputation and earning increased self-esteem by virtue of knowing the answer sufficient social drivers? Time will tell what the sustaining factors really are as these services grow and evolve.
At the end of the day…
Curation is the central issue of our socially networked culture. This mutual mediating, to riff on Kevin Mark’s coinage, turns the firehose of our collective activity streams into a refreshing drinking fountain.
Call it curating, crowdsourcing, stigmergy, page rank, social networking, lifestreaming or a thousand other terms, its objective is to deliver information and entertainment that may be useful or surprising, silly or serious, funny or annoying — when, where and how we want it.
Thanks also to JP Rangaswami for thoughts that inspired this post:
Curators add to relevance by stripping away the irrelevant and the unneeded and the shoddy.
In order to improve consume-ability and relevance, curators need the tools to do this. There are two ways these tools will come about, the “nice” way and the “nasty” way. In the nice way, the producers and distributors make it easy for people to point to, package and pass on the relevant pieces.
For digerati traveling in Italy, you know that there is dearth of wifi spots to be found. As a public service to some poor soul that searches “wifi Corniglia“, I’m writing this post. In the piazza of the main church, with a convenient stone bench all around the circumference, you may connect to “Tiscali” without a password.
Location, location, location
If you haven’t heard about Foursquare, Gowalla, Latitude, Scvngr, MyTown, Loopt, Geodelic, Brightkite, Where, Check.in, and could care less about other location based services, just enjoy the Pongo illustration and don’t bother reading any further! (I want to mention Layar here, even if it is not strictly in this category of applications. If you haven’t seen this yet, you really should check it out. It’s described as an augmented reality browser. Using the geo-location, compass and camera functions of your mobile device, it displays information about the stuff around you as you aim your camera from place to place.)
This guy with the crown has checked in here lots of times, using an application on his mobile phone. In order to keep him coming back here, I let him earn a crown and the title of mayor. That may be enough for the honeymoon period, but I’ll probably need to start offering him free drinks as long as he keeps coming back and keeps on broadcasting his presence to his friends.
What is the business model that will emerge from this scenario? I’m just going to wait and see, quite a few come to mind…
Hyperlocal, not hype-local
But there are three startups that are coming at the issue from a different perspective and I wanted to mention them here, Lasso, PaperG and BlockChalk. They all are focusing on the hyperlocal experience, the places where we spend most of our time and, not incidentally, most of our money.
Lasso and PaperG are quite interesting from a business point of view because they are sitting at the crossroads of a very real problem — local newspaper advertising revenues are shrinking, yet local media still have a sales force to sell advertising (and other promotional services) to small local businesses that a digital pure-play would have a hard time reaching due to the high overhead of selling to small advertisers.
First up, Lasso, who are building a kind of self-help hyperlocal ad server with a dash of Google AdWords:
Lasso is a platform positioned to enable local media companies to reach small and medium-sized businesses with attractive new offerings: integration with social networks like Facebook and Twitter, and distribution through customizable widgets throughout a media company’s website…
“We [Lasso] have combined internet marketing expertise plus newspaper DNA. An effective product has to have the right metrics, and in terms of landing it appropriately, it has to be done in terms that newspaper ad sales people understand,” Treadaway said. “If people don’t feel like they’re getting value, the churn will be very high. The product demos very well, looks very slick by newspaper standards. Most of the response has been: ‘when will you come train our people.’ (from LostRemote)
PaperG takes a different approach:
PaperG is testing a software system called PlaceLocal that automatically generates ads for local businesses by crawling the Web. The system scrapes the Web for basic information about a business such as its address, phone number, and opening hours. Even if the business doesn’t have its own Web page, data can often be pulled from third-party services such as Yelp or Google Maps. The system then uses semantic analysis to find and extract photos and positive reviews, and it builds an ad automatically using Adobe’s Flash software. The business owner or newspaper ad sales representative can customize the ad, so if PlaceLocal didn’t choose the best photo or review, it’s easy to select another.
Lee expects PlaceLocal to help representatives sell ads in the first place. “The sales rep can have a beautiful ad designed for every lead sheet,” Lee says, “which makes a real difference in the conversation.” (from MIT’s Technology Review) (The NYT writes about it here.)
BlockChalk is taking a much different approach, the social media classic — build traffic and figure out how to make money once they get traction. The only reason I’m putting them in this post is because of their unique usage model. You don’t have to sign up to the service to use it, and you remain anonymous until such time as you decide to share your identity on a case-by-case basis with other users of the service.
Since anonymity opens the service up to spam and all sorts of abuses, and denies the service a registered user base and profiled social graph, I am really interested to see what kinds of checks and balances they build into the system. But the really cool part to watch will be to see what kind of behavior users will adapt on their own in this wild west of a service. What will be the equivalent of the Twitter hashtag, what cultural norms will people develop and adopt?
That is if it lasts long enough for a culture to develop…