Social networks have been all the rage for over a decade now. In the past years, Facebook has consolidated their leadership in personal networking and Linkedin has become the undisputed leader of business networking.
A side effect of this consolidation is that business models have also consolidated, leaving out a lot of potential applications and use-cases that do not fit the current models.
Both Facebook and Linkedin have business models that focus on monetizing their respective social graphs through advertising. LinkedIn also monetizes search, analytics, and recruiting.
To protect their data and business, they have largely cut off API access for third parties. Consequently, building third party applications that use these networks has become very hard, if not impossible.
At the same time, building a new social network has become harder due to the fact that most people already have their social networking needs covered by the existing offerings. Most wannabe new social networks face the ’empty room problem’, where they don’t become interesting until enough users are using it.
Inbot was founded on the idea that sales is fundamentally a social process where building relationships and trust between people is the key to success. Yet, Customer Relationship Management (CRM) systems used today by most salespeople are anything but social.
CRM is mostly used to manually keep track of conversations with customers. Relationship data is shared only within the sales team, and when sales people change jobs, all aggregated data in the CRM is left behind and they take their social network with them.
Marketing automation has emerged as a software-based mechanism to help companies to generate leads in a more automated fashion. The problem with it today is that the spam and noise generated by these applications is deafening, and making everyone harder to reach.
Initially, Inbot started out as a disruptive play to make CRMs find and provide links to new business opportunities. Over time, we realized that we should focus solely on social lead generation, and decouple it from teams and companies that CRM vendors target.
Since last August, we have rolled out a community that we hope will one day rival that of Linkedin but yet works very differently.
The Inbot People Graph
The Inbot People Graph is in some ways similar to existing social networks but very different in the way it works and evolves.
Most social networks create their social graphs based on friend requests that establish verified connections between members of the social network. Both Linkedin and Facebook operate on this model.
The main purpose of friend requests on social networks is to protect the privacy of the users. Most social networks use the confirmed relations that they have to authorize actions within the network such as sending a message, sharing a photo of a cat, etc.
If you don’t join the social network, you are not a part of its graph and you provide no value to its members. This also creates the empty room problem mentioned earlier.
Bring your own friends
Inbot doesn’t have or need friend requests. It is not an app where you have to sign up and then convince all your friends to join as well. Instead, the Inbot people graph is an emergent graph that we mine from data that our users provide to us about their relations in the real world. So, if you connect some data sources, we’ll figure out who your friends are in Inbot and in the real world.
Our graph is as good as the data we have. We receive data when our users connect their contacts, emails, messages, calls and calendar events on their phone, and use Inbot. This builds a picture of their real world relationships. The Inbot People Graph is a global graph where the relations are emergent rather than managed. Our system knows who trusts whom based on this data.
The Inbot People Graph has some interesting properties. In technical terms, it is a weighted graph and the connections are bi-directional. A common pattern in existing social networks is to connect opportunistically. You may have 5000 friends on Facebook or Linkedin but that doesn’t mean you have 5000 trusted friends in the real world and also you might be calling or emailing regularly with people not represented in those networks.
Adding weights to these relations makes a huge difference. It means we can look at the nature of the relations that people have in our graph and identify the connections that really matter. This changes over time as you start conversations via email, make calls, or get introduced to new business relations.
We can tell apart recent contacts from contacts you interacted with years ago. We can determine the nature of the relation as well by looking at which channels you use to connect and by looking at how your connections are connected with other people you interact with (clustering).
An important benefit from our emergent graph model is that our network evolves in a way that makes the graph interesting for our users much earlier than is the case with the classic invitation based mechanisms of existing social networks. Right off the bat, we know everybody our users care about through their contact books. They don’t have to be Inbot users for the system to know this. The graph grows much faster than the number of users. As new users join, links in the graph get stronger and everybody benefits from this. As new data is absorbed, the weights on the graph keep evolving.
A typical user brings on average from hundreds to several thousands of contacts. Particularly active people may bring in tens of thousands contacts. This means that as our user base has grown slowly over the years to several thousand, the number of relations we are tracking has grown to many millions. Every time a new user joins, more links and data points form between the contacts automatically.
As our user base grows, we’ll eventually know virtually everyone that matters to our customers on this planet; which is effectively anyone with a heartbeat. What’s better, the contacts in our people graph today already represent everyone that our collective user base knows and cares about.
We have so far not made much of an effort to grow our user base due to our strict focus on perfecting the technology first. Despite that, our user base has evolved organically to support our business case. We’ve solved the empty room problem.
The Inbot community
While anyone can sign up for Inbot, we have two special types of members in the Inbot community: vendors and ambassadors. These members are our most active users. Vendors are companies that we help to find new customers, and ambassadors are trusted entrepreneurs, executives and investors who help them get introduced to those customers.
We use the Inbot people graph to find links to new customers for our vendors. We charge our vendors money for this service. When they successfully conclude a deal with a customer that our ambassadors introduced them to, we share a cut of the revenue with the ambassadors.
Currently the Inbot Ambassador community spans 40 countries on all major continents, including main startup hubs and industrial centers worldwide. Through this community, we provide access to hundreds of thousands of decision makers across major industries.
Unlike marketing automation vendors, Inbot does not make money from spamming people. We make money when our vendors convert the introductions from our community to deals. This is important. We don’t sell lists of contacts that our vendors can spam. In fact, our members’ contacts are not available via any other mechanism than their voluntary introduction. And they never will be. We don’t serve spammers and actively work to keep them out of our community.
The only thing that Inbot sells is bona-fide, warm introductions by our trusted network of ambassadors. Introductions to potential customers are provided by ambassadors to vendors on a basis of good faith. Ambassadors only profit when we profit, which is when our vendors convert our ambassador’s introductions into actual sales to, presumably, happy customers. Everybody wins in this scenario.
Our community members are highly motivated to make good introductions, because everybody loses if the quality of the introduction is low. Low quality introductions can erode the trusted relations we maintain with our vendors and ambassadors. These relations are all based on trust.
The cornerstone of our business model is the quality of the introductions. We replace spam with trusted handshakes. We identify the most interesting potential introductions that our ambassadors could possibly make for our vendors, pre-screen them, and then introduce the ambassadors to the vendor for actual introductions. This ensures we don’t waste our ambassador’s time unless we are confident they can make a good introduction and that we supply high quality introductions to our vendors.
The key to success here is how fast we can automate the curation and delivery of high quality target customers to our vendors and ambassadors. This is a hard problem that has taken us a long time to solve, but we are getting better at it thanks to the Inbot People Graph.
Bootstrapping the people graph
People form relations in all sorts of ways and a lot of these relations can be deduced from raw data. Inbot has long provided the ability for users to do this using your email communication on email and call data. But there are many other relevant data sources that we are adding to the mix, including existing CRM systems, manually curated data, user contributed trust data, crawled web data, the behavior of our users in our system, and machine learned facts from all of the above. We use facts from all of these to calculate weights on the relations in the people graph.
This means that some connections are strong, some are weaker. Some connections are verified, some connections deduced, or even merely suspected. As our data improves, so does our ability to reason with this graph.
The data that we receive can be extremely messy. People’s address books are a mess. There are duplicates, incorrect entries, contacts named “mum”, invalid phone numbers, email addresses, malformed data, and other outdated data. Our servers actively clean up this raw data all the time. We use algorithms as well as manual curation to do this. A large part of what makes the people graph work is simply cleaning up the data and connecting the dots. Once cleaned, the data becomes more trustworthy.
When new users provide us with their contacts, we immediately connect them to our graph. This helps us make sense of determining the people behind the contacts. We also use the graph to clean up the contacts. So, even if you are not in Inbot to do sales, there is immediate value from simply tapping into our graph and receiving a lot of metadata about the people that you have in your address book.
The contact enrichment our members receive include social media profiles, companies, job titles, categories, countries and other information that is not privacy sensitive. We never share confidential contact information such as phone numbers or email addresses. However, we do store this information and use it to establish relations between similar contacts.
Trust based mediation
A network like Inbot has much broader applications than what we are currently doing. There are many scenarios where people tend to prefer trusted referrals. Referrals take place in the context of people’s social networks of friends, family, and business relations. However, this network is exhausted pretty quickly when you use it. Reaching beyond your own network requires mediators, brokers, or facilitators to connect you to more people.
The six degrees of separation rule tells us we are only a few people separated from everyone else on this planet. However, the related three degrees of influence rule tells us that the influence we have through our social network drops off sharply and becomes meaningless after only three degrees of separation.
This means that finding and connecting with people beyond your first degree network is a problem. Mediators solve this by bringing you into their network and effectively adding people to your first degree network through introductions that they facilitate.
Trust is the key thing here: you trust some of the friends in your first degree network and therefore you trust that their referral is appropriate. This is less clear with mediators that you don’t know. Who are they? Can you trust them? Finding trustworthy mediators is exactly the problem that Inbot is solving for matching vendors with potential prospects using our community of ambassadors.
The people graph allows us to scale trust based mediation to a global scale and we can apply it to essentially any business where brokering is used today, not just sales. We keep track of reputation of all the actors in the graph and their interactions. So, whether you are buying a house, looking for new people in your organization, trying to find a reliable supplier, or in need of legal support in a new market, it all boils down to knowing who to talk to.
The Inbot People Graph can provide this as a service to anyone, and for anything and this is pretty much our long term goal. We want to be at the core of the networked economy where people can find each other more effectively than they can today.