The place where the Customer counts

Free thoughts on CRM, Business and the next big thing

The place where the Customer counts - Free thoughts on CRM, Business and the next big thing

Social Business Forum 2012 – Big data bring big value to the Social CRM

Here it is, my presentation at the Open Conference in Milan at the Social Business Forum 2012. Great experience, great people, lots of insight and lots of new points of view about the evolution of topics like Social CRM, Social Business and Analytics. I need some time to handle all this information and to elaborate some thought to express on a new blog post (stay tuned cause outcomes will be surprising ;)). A big “thank you” to the deus-ex-machina Emanuele Quintarelli (that I called the Event Man), Emanuele Scotti and Rosario Sica and all the OpenKnowledge guys for their invitation.

In my speech I wanted to give to the audience – even if the topic was maybe a little bit “tough” – some information about a topic that, from a hype phase, is now consolidating and that I’m sure will bring lots of value to various aspects of your business strategies and, in particular, to Social CRM.

What’s it all about?

Big data is about technologies and practices that allow us to handle huge amount of data coming from a variety of sources typically in a stream structure. The phenomenon of social platform adoption and people collaboration together with a new set of antennas/sensors embedded in different products and contexts (first of all mobile phones and Rfid) let these data explosion become a reality giving new challenge to the companies hungry of infomation about their customers. The three drivers that typify Big data are volume, variety and velocity. These respectively bring with them business opportunities and risks as well: richness of information and lots of noise, completeness of information and too many sources to be handled, up-to-date information and risk of information loss.

Big data sources inside Social Business ecosystem

In a social business ecosystem where all actors exchange information of all kinds, everyone is characterized by three classes of attributes.

  • Myself: all data typically defining my unique profile.
  • Myworld: all data representing my experiences with products/services I typically use for personal or professional use.
  • Myrelations: all data identyfing my membership to different communities (friendship, interests, hobbies, professional links, etc.).

Let’s get a Social CRM definition

How Big data are useful for Social CRM? Before we have to give a definition of Social CRM and the most appreciated one was given by the most famous CRM expert Paul Greenberg which highlighted the strategic perspective of CRM (social or not).

The shift from CRM to Social CRM

Starting from the three famous CRM pillars (collaborative, operational and analytical ones) we can see that shifting from a traditional to a social approach we have gained, for sure, more channels where conversations occur (direct to the companies or within people excluding a direct invovment of companies), new activities and multi/cross-channel operational approaches and, finally, what about analytical improvements?

The Big data funnel for Social CRM

Now we have the opportunity to start from the real human experience that, interacting with different touchpoints, leaves tracks in the shape of raw data (i.e. transactions, interactions through traditional channels, web & social, location-based). The challenge for modern companies is to take these continuos data streams, transform them in information (giving meaning to data) and make insight (extract piece of information that is relevant for the business cause is actionable).

Now we are plenty of “human” data

From a customer perspective the three classes of attributes Myself, Myworld and Myrelations can give us a really complete view on his essence (more than traditional profiling and transacational attributes).

What’s pratically changing?

Segmentation practices must change cause we can enrich our understading of customers through new information that characterize them not as simple targets but as beings that naturally express themselves belonging to communities (or tribes if you want).

So it’s time to really understand your people

Now you have the opportunity to really understand customers’ attitudes, experiences, emotions and opinions. You are more close to understand their real nature as human beings.

How can we handle it?

But there’s a problem. Big data, with their specific characteristics, make difficult to address our attention to the right information, and “right” means useful for your customers and your business. Mainly in the case of unstructured data handling (i.e. text) you need human intervention for decision making but because of our cognitive limitations we need more and more some help to support us in processing, understanding and selecting information.

What’s in it for me?

Thanks to descriptive and predictive analytics we have all the methods and analysis that can support us in many businessapplications, especially in a proactive way.

Can we trust Analytics?

But lots of companies are suspicious about analytics accuracy and think that only high values of accuracy bring significant outcomes. Wrong. First of all people confuse precision with accuaracy ( and that can create dangerous misunderstandings but, more than that, the impact of analytical errors strongly depends on the business contexts and priorities so it’s normal that you maight focus your attention only to some kind of errors and drop others depending on your business objective.

An example for Social Customer Service

Well, it0s clear directly seeing the slide that correct analytical handling of structured and unstrctured data can provide proactively more information at your CSRs in order to better serve your customers fulfilling their expectations and improving their satisfaction.

Great opportunities but pay attention to the issues

Finally you’ve to remember that big opportunities bring also big potential issues that you must address. First of all about policies (liability, security and privacy), access (not every data is transparent and available so it’s mandatory to find trustworthy approaches to incentivize data sharing) and technologies (framework like hadoop are necessary to support massive parallel and distributed application processes). But be sure to spread inside your oganizations an analytical culture and to find the best analytical talent that must have not only strong analytical skills but also great expertise in decision making to support top and middle management in this important phase.

Social CRM: cosa c’è oltre l’incrocio?



Se non avete ancora letto l’ultimo post di Graham Hill fatelo subito e poi ritornate qui. Perché? Semplicemente perché Graham traccia una linea netta e cruciale oltre la quale chi si vuole occupare di Social CRM deve fare una scelta definitiva. Riassumendo l’articolo la questione fondamentale sta nel fatto che la tematica del Social CRM si trova ormai ad un trivio: o si sceglie di considerare i nuovi media esclusivamente come dei meri canali di interazione e comunicazione (al pari di IVR, numeri verdi o neri, fax, email, ecc.) o si cede semplicemente all’occasione di adottare delle nuove tecnologie alla moda o, infine, ci si convince che siamo obbligati a rivedere il rapporto azienda – cliente come un nuovo contratto di mutua assistenza attraverso il quale co-creare valore (coadiuvati sempre e comunque da una strategia ben definita, dal coinvolgimento delle persone e del loro contributo umano e professionale e, logicamente, dalle adeguate tecnologie abilitanti). Le prime due strade non sono concettualmente errate, ma hanno un respiro molto corto e prospettive destinate ad esaursi in tempi molto brevi; la terza strada invece include le prime due e proietta le potenzialità del Social CRM verso un nuovo ecosistema su cui basare il proprio modello di business.


L’aspetto che più mi ha interessato del post (che è stato trattato anche dal vivo al recente Enterprise Social 2.0 organizzato da KGS Global e tenutosi a Bruxelles l’8 e il 9 marzo scorso – resoconto raccontato in modalità live dall’ottimo Emanuele Quintarelli nel suo blog) è in particolare la necessità espressa da Graham di elaborare la cosiddetta customer journey map (tecnica di service design e customer experience modeling) ossia lo scenario all’interno del quale l’esperienza del cliente prende forma attraverso l’interazione con le proprie reti di conoscenza e con l’azienda. In particolare è cruciale identificare i cosiddetti touchpoint (virtuali, fisici, umani e automatici che siano) e comprenderne il valore di scambio che li caratterizza. Rientra così in gioco l’osservazione, non solo l’ascolto e l’analisi dei comportamenti, e il mettersi nei panni dei propri clienti, immergersi profondamente nel loro flusso esperenziale per riuscire a rappresentare questo “percorso”. A supporto di questo processo ci sarà, a mio avviso, la necessità di sviluppare le seguenti aree:


Voice Of the Customer (VOC). Sotto questo nome ricadono tutte quelle metodologie propedeutiche al disegno delle aspettative della clientela e che includono tecniche di analisi tra cui il monitoring, fil eedback management, la netnografia, il surveying, il mistery consumer, i focus group ecc. Su tale tema in oggetto vi rimando ad un mio precedente post incentrato in particolare sul tema dell’ascolto della Rete.


Analytics. Investire sugli analytics, e in particolare su quelli che consentono la gestione della consistente mole di informazioni non strutturate che pervengono per l’adozione dei canali sociali, è tassativo per due motivi:

  • da un lato è importante avere una struttura di indicatori e metriche che consentano il monitoraggio costante di cosa si sta facendo e di come lo si sta facendo in termini di efficienza, efficacia e correlazione con l’obiettivo di value co-creation.
  • d’altra parte le tecniche analitiche di sentiment analysis e text mining assumono un ruolo determinante per gestire e comprendere la suddetta mole di dati in input. E’ vero che ad oggi è in atto una interessante discussione sull’attendibilità di queste tecniche soprattutto se si compara con i livelli di affidabilità dei modelli di data mining basati su dati strutturati e quantitativi. E’ quindi vero che ad oggi l’intervento umano è indispensabile in primis per ridurre il rischio del garbage in-garbage out, ma è altrettando vero che è imprescindibile riuscire a filtrare i dati in ingresso altrimenti il carico informativo risulterebbe devastante (senza contare che l’affinamento della ricerca in questo settore migliorerà sicuramente i suddetti livelli di attendibilità).


Adaptive Case Management (ACM). Dal BPM all’ACM. Dalla gestione di step strutturati, sequenziali, rigidi e inamovibli (seppur automatizzati o quasi) a un nuovo paradigma di gestione delle informazioni e di casi strutturati e destrutturati – tipicamente associati alle attività dei knowledge worker (provate ad immaginare i contesti in cui operano coloro che si interfacciano con la clientela lavorando in ambito Customer Support o di Sales) – che permetta di affrontare tutte quelle situazioni in cui è indispensabile (ri)adattare i diversi task con tempi di risposta sempre più vicini al real-time e possibilmente senza interventi lato IT . Per approfondimenti sul tema vi consiglio la lettura del blog di Keith Swenson, autore tra le altre cose del bel libro “Mastering the Unpredictable“.




Analizzando la figura si possono notare degli spazi di sovrapposizione:

VOC <> Analytics – in questo contesto gli analytics consentono di entrare nel dettaglio dei dati acquisiti con le tecniche di VOC e di comprendere esperienze e aspettative della clientela.


ACM <> Analytics – in situazioni di gestione dei processi in ottica adattiva è fondamentale tener traccia di tutti i casi gestiti e delle modalità operative seguite in modo da poterne valutare, attraverso gli analystics, l’efficienza e l’efficacia. Inoltre tutte queste informazioni e i relativi insight sono indispensabili se visti come strumento di supporto alle decisioni (future) del knowledge worker.


VOC <> ACM – la definizione dell’esperienza del cliente insieme alla comprensione e predisposizione delle modalità di gestione della sua relazione con l’azienda in ottica adattiva (quindi non solo adattamento alle logiche di lavoro del knowledge worker ma anche alle esigenze e aspettative del cliente) consentono di ottimizzare il processo di Service Design.


La sintesi di queste sovrapposizioni costituirà dunque la base conoscitiva e operativa per approntare una corretta Customer Journey Map.

Avere quindi competenze e padronanza nelle tre aree esposte consente una migliore definizione della customer experience del cliente e del raggiungimento dell’obiettivo di co-creazione e mutuo scambio di valore.

Cosa ne pensate? Mi piacerebbe avere un vostro feedback.

Expert POV: Social Analytics interview (part 2)

Here we are with the second interview on Social Analytics; now with Seth Grimes, surely one of the most influential expert in analytics especially the “Text” ones. A brief introduction with his bio and then let’s read his answers full of interesting information.

He is an analytics strategy consultant. He is founding chair of the Sentiment Analysis Symposium and of the Text Analytics Summit series, contributing editor at TechWeb’s Intelligent Enterprise magazine, and text analytics channel expert for TechTarget’s BeyeNETWORK. Seth founded Washington DC-based Alta Plana Corporation in 1997. He consults, writes, and speaks on information-systems strategy, data management and analysis systems, industry trends, and emerging analytical technologies.

Question 1

One of the main problems that rise with the advent of Social CRM is the huge amount of unstructured data coming from the increasing number of social platforms and applications on the web. In your opinion, which is the better way to manage them for the successive insight and intelligence analyses? Will it be necessary to archive them all or do you think that filtering systems/specific access modality can (must) be applied? What do you think is necessary to pick up the more meaningful information in social data and how?

Seth Grimes

Unstructured sources are not a problem; rather they are a challenge, an opportunity. They contain a vast quantity of business-relevant information. They are the key to getting at the “voice of the customer” in real-time, to engaging with customers and prospects and stakeholder as routine business interactions move online.

Businesses need to find, filter, and harvest social intelligence that will help them further business goals… and ignore all the noise, all the conversations that don’t relate to them. Listening platform, semantic (meaning-based) search, and text and social-network analyses can all help, but you don’t have to do it all yourself. Set up your own systems if you wish, but most business are going to find “as-a-service” solutions to be very attractive. Software as a Service (SaaS) makes getting started easier, faster, and cheaper, and in addition, many of the service providers have already archieved years worth of social and online information.

If you do need direct access to social and online information, you may be able to get what you need from services such as Moreover, Spinn3r, and YellowBrix.

Question 2

During the last ITExpo Symposium 2010, Gartner has identified the 2011 Top 10 Strategic Technologies. Among these there are three that are directly connected to the Analytics world on which I’d like to hear your opinion. Let’s start with the most generic one: Social Analytics (Gartner says: “Social analytics is an umbrella term that includes a number of specialized analysis techniques such as social filtering, social-network analysis, sentiment analysis and social-media analytics”). Which do you think will be the possible developments of these techniques and their main support to the several aspects of Social CRM (advocate/influencer identification, proactive lead generation, etc.)?

Seth Grimes

Gartner’s definition of Social Analytics is a good one… for now, for 2010. The issue is that few businesses exist only socially, really only the social platforms themselves. For every other company — including online vendors such as Amazon, eBay, and Netflix — social-media success is (or should be) measured not by followers and page views but rather by sales conversions. So the challenge becomes linking social mentions and interactions with Web-site/store visits, inquiries, purchases, and customer experience.

Online advocate/influencer identification, lead generation: All good and getting better. But I want and expect to see the term Social Analytics disappear — I want to see the term Social CRM disappear — in favor for general, comprehensive customer/market analytics and CRM where social is just another touchpoint, just another channel.

Question 3

Now, what about Next Generation Analytics (Gartner says: “It is becoming possible to run simulations or models to predict the future outcome, rather than to simply provide backward looking data about past interactions, and to do these predictions in real-time to support each individual business action“)? Which will be the main Social CRM area that will benefit from the use of these predictive techniques, and how? Can you make some practical example?

Seth Grimes

Next-generation analytics (NGA) is comprehensive, pulling data as needed from the full spectrum of touchpoints and sources, including textual and geolocation data. NGA fuses information from these disparate sources, perhaps using semantic data-integration techniques, and it delivers information in flexible form both for self-service, visual analyses, including via mash-ups and for embedded predictive systems.

The underlying Big Data will make for more personalized customer interactions, for better targeted marketing, for customized products and services.

There are a few social-media analytics out there who tout an ability to predict future levels of social conversation. That’s ridiculous and it’s meaningless. I couldn’t tell you what new signer will be hot in 2 years, nor which nascent platform will be the next Twitter.

What we need instead is the ability to read “intent signals” in social messaging: phrases, patterns, and behaviors that suggest that someone is shopping for a product or is dissatisfied with a service and about to change providers. Link those signals to CRM analyses — does the customer have a projected high lifetime value or is he/she a chronic complainer whom we’d just as well not have as a customer? — and there you have true Next Generation Analytics in the service of unified — not just social — Customer Relationship Management.

Question 4

And finally what about Context-Aware Computing (Gartner says: “A contextually aware system anticipates the user’s needs and proactively serves up the most appropriate and customized content, product or service“)? How to really trigger coherent and proactive actions through social insight?

Seth Grimes

If Content is King — and I believe it is — Context is Prime Minister, responsible for the correct interpretation of content, for guiding the appropriate response to those “intent signals” I discussed earlier.

But really context awareness is only a part of a larger need. That larger need is situational computing, which takes into account not only current context but also past history (from behaviors and transactions), profile and identity, and also customer-user interests and preferences. It’s a mistake to blindly treat your community as content/product/service targets. Not everyone wants that. Many people will find Gartner-esque “anticipation” to be invasive, to be downright creepy.

Question 5

Do you believe that there are other important analytics techniques/technologies not considered in the Gartner forecast? If yes, which are and what are their business potentialities in Social CRM?

Seth Grimes

Gartner has long been a market-follower, reporting innovation only when it become established enough to hold some appeal for Gartner’s larger-firm clients. Further, the majority of Gartner’s work is published only to paying clients, and I’m just not convinced that Gartner gets social-media as users. All this is to say that I don’t closely follow Gartner closely and I’m guessing that there surely are important analytics elements not considered in Gartner’s forecast.

Question 6

In you opinion, there are vendors proposing interesting social descriptive and predictive analytical solutions? And how this market will develop in the next few months?

Seth Grimes

I know of a number of vendors in my speciality area, text analytics, that are focusing on those intent signals I discussed: OpenAmplify, Attensity, and Janya for instance. There are a number of other providers with strong (and getting stronger) sentiment analysis, semantic search, and complex data integration capabilities. IBM has one of the broadest set of semantic-social-enterprise analytics solutions, backed by some very deep technologies, and SAP is similarly on the cusp of bringing these forms of analytics to its enterprise customer base. It’s great stuff, with a constant flow of innovation.

Thanks again to Seth and to his valuable contribution. Stay still tuned for the next experts’ POVs and if interested on the topic, have a look at the previous interview with Jim Sterne.