Nodees, un (gran bel) progetto rimasto nel cassetto

Quando conobbi Roberto Bonanzinga (era il 2009) gli parlai di un progetto piuttosto ambizioso e decisamente fuori dalla mia portata. Si trattava di una piattaforma analisi delle reti sociali all’interno di un’organizzazione, che – attraverso un sistema esperto – fosse anche in grado di suggerire delle azioni per migliorare la struttura del social network per permettergli di supportare obiettivi come la condivisione di conoscenze. Era il frutto di decine di letture sulla social network analysis, sulla organizational network analysis, sui sitemi complessi e sull’allora nascente settore dell’enterprise 2.0.

Ho recuperato una copia dell’application che feci al SeedCamp di Londra nell’agosto del 2009. Ovviamente, la proposta su giustamente ignorata perché non era il tipo di progetto che loro andavano cercando: era ancora alla fase di idea, richiedeva un lungo periodo di ricerca e sviluppo, chi lo proponeva aveva una conoscenza troppo superficiale dell’argomento.

Lo pubblico qui, perché è qualcosa di più elaborato di una semplice idea e perché io non sarò mai in grado di realizzarlo: magari può far accendere una lampadina nella testa di qualcuno. Da questo progetto ho imparato che bisogna concentrarsi su sfide che, oltre a essere molto stimolante e gratificanti intellettualmente, siano anche alla propria portata. E’ una banale questione di credibilità 😉

What are you creating?
We want to build a web based end-to-end solution to provide social network analysis (SNA) and organizational network analysis (ONA) to:
A) organizations that are adopting social business software (ie. Ms Sharepoint, Lotus Connection, Jive, Socialtext…);
B) social network manager, who are running online communities using popular platforms (ie. Ning), white label solutions (ie. Kick App), open source software (ie. Drupal) or proprietary software;
C) professionals and academics who have to conduct SNA/ONA research projects starting from questionnaires or other relational data.
In case A and B, all the platforms that allow collaboration and sharing lack a system to analyse the dynamics of the social networks they are supporting. They generally use web analytics (ie. Ning offers direct integration to Google Analytics), but these kind of stats do not give any information about the social structure of the groups.
In case C, there are stand alone software (some very complex), but it does not exist an end-to-end web based solution that integrates the gathering and the analysis of the data.
For further information on SNA: http://en.wikipedia.org/wiki/Social_network_analysis, http:// www.insna.org/

Let us say you have 15 seconds to pitch your business. Can you describe your business?
Nodees is the Google Analytics for social networks. It is a web based tool that help to understand social structure of organizations and communities and to govern them according to given objectives.

What is really new about that?
SNA and ONA are new management disciplines: non academic books and papers about these topics – ie. the ones authored by Patty Anklam or Rob Cross – were published after year 2000. Today, there is not a complete solution that can help social network managers (inside and outside companies) to easily gather relational data, analyze them and obtain suggestions to govern a community according to given objectives.
Nodees will be divided into 3 components: the Harvester that will gather data; the Analysis engine that will analyse data and draw the graphs; the Expert system that will help the user managing the process and interpret the results. Following, we give a deeper view of the architecture of the system and of state of the art of the available tools that perform similar tasks.
The Harvester is the component that will gather relational data. This will happen in two ways: a) using questionnaires (you ask people questions like: whom do you ask when you need information for your work?). Our solution will include a survey manager on the model of onasurveys.com.
b) capturing actions from social networking software. It is possible to capture many actions (friendship, comments, collaborations on a wiki pages and so on) using different methods: accessing the DB of the social networking software and extracting the necessary data; instrumenting the system with java-scripts like for web analytics; scraping information from web pages.
We will simplify a task that is generally difficult to perform: collecting data and formatting them according to SNA software requirements is the most problematic task of a SNA/ONA project.
The Analysis engine will perform the calculations and draw the graphs. Today, these functions are performed by stand alone sotfware and the most popular ones are Ucinet, Netminer and Inflow. They are quite difficult to use: not because SNA is intrinsic difficult (basic concepts can be acquired in few hours), but because these tools are designed to be used by skilled professionals and academics.
We will embed a great part of the knowledge necessary to perform a SNA/ONA into the software, providing an easy user experience and making SNA/ONA accessible to an average community manager.
The expert system will help the user to: a) define an analysis strategy given the kind of social network they are assessing and the objective of the project; b) interpret the results and give suggestions on actions that can be performed to optimize the social network according to the objectives.

What specifically is your target market and how is it being poorly served today?
Organizations that use social business software.
Social network managers, who run online communities.
Professionals and academics, who have to conduct SNA/ONA projects.

What gives you an unfair advantage and how will you sustain it?
The first advantage will be conquered crowdsourcing the design of the product to the international community of SNA/ONA experts. Since the most expensive an difficult task of SNA is to gather and format the relational data, we will offer a free harvesting tool to the community and ask them help us to refine and localize it. This will maintain design costs low, build relationships with SNA experts all over the world and build a strong reputation. If we will have the resources, we will involve the most renowned professionals as advisors of the company.
The second advantage will be acquired creating a team specialized in SNA/ONA that will: build course ware and a public knowledge base of case histories; help selected groups of community managers to execute their first SNA projects.
The third advantage will come from the expansion of the Harvester that will cover a wide range of platforms and from the release of the Analysis engine that will integrate the process of collecting and analyse data.

What measurements will you use to value your business in a year? What will it be worth? What are key milestones that will account for the growth in value from today to 1 year from now?
In the 2010, the service will be offered in beta and will be free of charge. So our KPIs cannot refer to economic achievements. By the end of 2010, we will measure our success according to the following KPIs:
* Harvester – Survey Manager: online for at least 6 months and fully localized by the community in 10 languages; 100 professionals around the world actively using the system.
* Harvester – Connectors: working connectors for Ning, KickApps, MediaWiki, Jive, Sharepoint and Lotus Connection. 5 pilots for each platform connected to the system.
* Analysis Engine: online for at least 6 months and performing all major SNA measures.

What fact would make the most difference to your confidence that your company will succeed? How and when will you find that out?
The biggest risk is that Nodees we be not able to convert the need for a tool that helps managers to know and govern organizations and communities into the actual use of its services. So, we will be confident that we can succeed if SNA/ONA experts (our first public) and community managers will use Nodees as an important tool in their day by day job.

Who are your main current or potential competitors as well as identified potential new entrants?
We have identified 3 types of competitors:
a) current software vendors (Ucinet, Inflow and Netminer) because they have an established reputation. It is likely that they will maintain their role of niche and specialised solutions.
b) Cross Analytics (www.crossanalytics.com), a project by Rob Cross (one of the major expert and author of two books) that has developed a tool that integrates a web based ONA surveys and a stand alone software for analysis. The business model is mainly based on consultancy (not on saas), but Cross is very oriented to business.
c) Trampoline Systems (www.trampolinesystems.com) or Touchgraph (www.touchgraph.com) that could adapt and extend their software to compete directly.

How will you succeed versus your competitors?
Assuming that we will be able to provide a quality product and service and that our solution will be the first on the market to integrate a SNA/ONA process, the only way I know to succeed is to listening to customers, embed their knowledge in product, support them with a superior customer service.

Planning for the worst is a key to great success. Think hard: what might go wrong? How can you minimize those risks?
The risk is that SNA/ONA experts and community managers will not use Nodees: We will cope with it, establishing strong relationships with our potential customers and involving them crowdsourcing the design process. At the same time, if we will have the resources, we will contract some of the most visible experts as advisors of the company.

It is always good to evaluate all your future options. If you decided to sell your company, who would be the likely buyers for the business, and why?
Accenture, to sell both the software solution and the consultancy around it.
Google, to acquire the knowledge and integrate the tools in Google Analytics.
Nielsen, to acquire the knowledge and integrate their services as they did with Buzzmetrics.

Aggiungo anche un po’ di bibliografia utile, selezionando tra i testi nella mia libreria:

  • Cross, Parker, The Hidden Power of Social Networks
  • Cross, Thomas, Driving Results through Social Networks
  • Mason, Intranet 2.0
  • Wasswerman, Social network Analysis
  • Carrington, Scott, Wasserman, Models and Methods in Social Network Analysis
  • Scott, Social Network Analysis
  • Barabasi, Link
  • Watts, Small Worlds
  • De Toni, Comello, Prede o ragni
  • Johnson, La nuova scienza dei sistemi emergenti
  • Taylor, Il momento della complessità
  • Waldrop, Complessità
  • Prigogine, Stengers, La nuova alleanza
  • Kauffman, Esplorazioni evolutive
  • Buchanan, Nexus

4 comments

    • No, il progetto Dragon Trainer mi era completamente sfuggito. Mi fa piacere notare una certa somiglianza nell’impostazione: considero un successo il fatto che una personalità nello studio dei sistemi complessi come Dadid Lane sia arrivato alle mie stesse conclusioni 🙂
      Per completezza, aggiungo il link al video di Alberto, con cui all’epoca discutemmo di Nodees (lui mi feve avere anche la ricerca che cita nel video): http://www.cottica.net/2011/09/15/dragon-trainer-begins/

  • Ciao, articolo molto interessante, considerando che io ho avviato per due anni un progetto di ricerca su metodologie statistiche avanzate sulle Cognitive Social Structures sotto la prof. Marjtje Van Duijn (la creatrice del p2 model).
    Inoltre come developer sono sempre stato molto attratto dalla tematica.
    Ti volevo segnalare qualcuno che forse ci è andato molto vicino al tuo progetto:
    https://socilyzer.com/

By Nicola Mattina