Skip to content

New Feature for social segmentation

October 22, 2014

We’re pleased to announce the release of Audience Insights;
In one single click, you can understand the segments in social that are relevant to your brands or topics.
See for yourself

Why marketing is missing the point in social

September 18, 2014

Unlearning marketing is a term that has surfaced recently, mainly pushed by the Adobe folks.

I can’t  agree more.  Marketing – as it was done before social media – is archaic and less and less relevant.

If you look at the words of marketing,  you clearly see it was built to address a flat market. The core notions of marketing, that current marketers are familiar with are:  “lists” and “segments”.

Marketers segment people into square boxes or folders, groupclist people with the same geo-socio-demo-attributes in rigid segments.

Marketers  buy lists, email to lists, manage lists.

All this was good before social media. But “social media” enables people to connect online and  make these connections  visible to brands, and to others.

The result is that people in marketers lists are  connected to each other too and that these lists are no longer  lists but  graphs.

And talking about segments, you can’t really segment a network. The proper technique to extract part of a graph/network is “clustering”. Segments have therefore become Clusters and the challenge for marketers has shifted from defining segments to identifying clusters.

Screen Shot 2014-09-12 at 9.47.22 AM

It may not sounds radical but the transformation is similar to traveling on a “flat world” versus on a globe.

Still most marketing executives we meet these days are not ready to question the practices and learning they built for a “flat world”. They frequently come to us with requests such as:  can you find us millenium’s,  wealthy single men 18-25  ?, people that match a specific persona? and can’t take the answer: “this is not how the market is organized in social”.

That’s for the geometry of the market … and  there is more to it:

  • These clusters are “conscious” i.e they know each other, they know they have power as a group and they know that marketers should know so the tolerance for “plain advertising messages” or lack of understanding of the nature of the social web is on the rise. ( for a definition on concious markets, see JF Noubel)
  • These clusters are also made of people with different positions & roles in the network. In this “networked market”   potential clients are connected one to each other in various social platforms and  belong to organic eco-systems with prescriptors, influencers , connectors …

How should people market to tribes? None really knows yet and It will take some time and a lot of failures before we unlearn “flat marketing” and invent how to market to a networked market. My take on this is that a better approach is to learn from  how “sales work” (“consultative selling”) and grow from there.

I’d like to end quoting Adobe’s CMO:

“Imagine getting rid of every marketing process, structure, and hierarchy. Imagine marketing was just invented. What could it look like if we started from scratch right now? No matter where you start, you end up in the same place: making a connection with a customer. Marketers may do it in all different ways, but the intent is the same. Only the means are different. Today, marketing is structured around the means, not the end result.”

Quote From B Rencher (Adobe)

How big is Sharknado?

July 31, 2014

For those who don’t know about it, Sharknado (now Sharknado 2) is either a cult movie or something really crazy/insane.

Here is the trailer of Sharknado 2 and that should be enough for you to pick your camp:  lovers or haters.


But whether you like it or not, it was THE top trend in Twitter today, so we gave it a try looking how really big it is.

To do that, we compared it to a few top series: “The Walking Dead”, “Game of Thrones” and “Under the Dome”. (yes I know, Sharknado is not a serie).

In terms of data points/ methodology , we looked at

- Google Trends  for the # of searches against the name of the serie and at

- our own data set of close to 1M influencers/publishers, measuring the number of mentions of the shows/ movie by the influencers.

Learning #1: It’s big, but it’s not super big .. so there is room for Killerwhalesnado or SharkQuake (with shooting in San Jose).

On Google trends, we see that Sharknado is 10 times smaller than big guys like Games of Thrones or The Walking dead, and roughly 1/2 of a show like Under the Dome. (but Google is late capturing trends  :-) and I guess Sharknado will catchup with Under the Dome shortly)

GTrends 4 series

On eCairn,  we got a similar pattern ( measuring how many of our 1M influencers mention the different TV series). The ratio is even bigger, Sharknado is 17 times smaller than Game of Thrones in the last 6 months.

Screen Shot 2014-07-31 at 1.23.43 PM

The trends are similar. However, we see that “publisher/influencer” trends are less pronounced than the search patterns.


Screen Shot 2014-07-31 at 1.29.01 PM


Learning #2: It did a huge buzz today (July 31th) with more mentions than Walking Dead or Game of Thrones.

We see below the “last day buzz” brings Sharknado a lead over everybody.

Screen Shot 2014-07-31 at 1.31.10 PM

It will be interesting to see if the trend we see from the “influencers” is reflected in the search trend in a few days (and I’ll post an update)


Learning #3:   Sharknado is popular with Movies/TV and Geek and got some pickup in Germany

Looking at what type/categories of people write about the TV shows/movie, we got the following graphs.

Walking Dead

Screen Shot 2014-07-31 at 4.22.14 PM

Comics ( it all started with a comic book) , Horror – really don’t understand this – , and religion!. If you wonder how the hell,  people covering religion (and atheism) can write about Walking dead, here are some examples:

but let’s refocus on data and pie charts….

Game Of Thrones

Screen Shot 2014-07-31 at 3.43.48 PM

Interesting to see the limited pickup in Gaming , and Moms.

Under The Dome

Screen Shot 2014-07-31 at 3.46.55 PM

some mentions in Brazil, France and Spain.


Screen Shot 2014-07-31 at 3.46.46 PM

Horror and some Geek-Tech (13%), marginal coverage in Germany.


May be a call for Sharknado 3 in Berlin???


@ dominiq




Social must be used for Selling

July 8, 2014
  • Markets are Conversations as found in the Cluetrain Manifesto. More than ever those conversations take place online, spamcustomer are actively engaging with each other within the social ecosystem relevant to their passion, expertise or cause they fight for.
  • Business are relationships. 80% of businesses are generated through referrals. Its well known that the true value of a sales person is its network of relations. And it goes beyond sales!

Top social media experts says the same thing here (and they’ve been saying it for many years now):

Bottom line, conversations and relationships are the keystones of social media.

I don’t know about you but cold calling and email out of nowhere are now just spam.

Here’s what an engineer put quite bluntly in his LinkedIn profile:

“I don’t know how to be more explicit about this—RECRUITERS: never contact me. Calling is a giant “fuck you” to me and email is a waste of your time. I don’t read your emails. If you’re not a recruiter and want to hire me, try convincing someone I know to talk to me about you. That’s the only way I’ll listen”

So what’s going on here?

Well, it’s easier to find people, social give people access to others that would never have been accessible. We’re all out there with profiles about who we are, content about our interests and tastes. Most of the time there’s a way to contact us and bang is the spam.

Case in point, here’s what I received today:

“Dear Laurent,

First of all I’d like to introduce myself. My name is (hidden) and I am the Chief Business Officer at (hidden). I have found your profile on LinkedIn and, given your role, thought that it might be a good idea to contact you and offer some services that could be of interest to you or your company”

While polite, this message doesn’t provide any reason for me to take action -> Ignored.

Today, if you’re in sales/biz dev and more, It’s key to try and establish a relationship prior to contact. It isn’t without challenges.

So it goes that way:
If you find me, contact me and I don’t know you -> Ignored
If you find me, contact me with something relevant to some content I published -> Remembered and perhaps we start a relationship
If you find me, convince someone I know to talk to me -> Gold because the someone I trust is telling me to talk to you.

Say, your company sells database administration or video conferencing software

1) Have you sales people listen to the conversations from your prospects. Their conversations is your market. Chances a lot will be on database administration, giving you a chance to engage.

2) Understand the nature of the relationship between you and them. Business is relationships. Find someone who knows you that know them.




DIY beauty

July 2, 2014

eCairn network of frequent expressions

Network of frequent expressions associated with 3 topics relevant to DIY beauty

Social media is teeming with tips, tutorials and recipes for making your own beauty products. The DIY beauty trend ties in with several other areas where influencers are actively relaying and publishing content demonstrating their skills and expressing their creativity thereby encouraging their readers and followers to claim ownership of the process of making their lives more beautiful.

In numbers, beauty influencers and others are building a virtual community of individuals and consumers, i.e. a market, where new values emerge. If those values for the most part stand clear of those of the mainstream industry, they are not necessarily in contradiction with them.


The objective of this study is two-fold:

  •  to make the case for the existence of a new market defined around the values and concerns dear to the DIY beauty influencers;
  • to provide insights on the issues covered within this community of beauty afficionados and identify the ones mainstream actors in the industry can leverage to participate and eventually enter the market.

Indeed, this new space where doing things yourself and doing things together is a defining feature offers great potential for growth and innovation for brands big and small.


To provide data points to measure and analyze the phenomenon, eCairn compiled a tribe of social media influencers interested in this new model of production for products related to beauty and fashion. About 1800 influencers sensitive to our issues were gathered: influencers with an interest in DIY cosmetics and perfume, DIY fashion, as well as minimalism.

Leveraging the eCairn Conversation™ application, we will study our tribe along three axes (or topics) pertaining to DIY beauty:

  • “make your own”: the conversations found on this axis include recipes and tutorials in which influencers share their know-how in the area of homemade beauty products.
  • “socially conscious”: the conversations found on this axis include concerns about the economic, social and environmental impact of producing and therefore consuming cosmetics and other beauty products.
  • “perception of beauty”: the conversations found on this axis include threads about beauty as an expression of the self, as well as beauty as an external ideal, a product of normative edicts.

Our social media analysis will bring answers to the following questions to gauge the scale of the DIY beauty market and define it further:

  • Who are the influencers who are the most sensitive to the DIY issues? What are their favorite social media outlets? What kind of reach do they have?
  • How much of the conversation is devoted to the topics through which the DIY trend is expressed? What are the most commonly used keywords and keyphrases in each of the three contexts and how do they relate to one another?

Want to know more? Download our in-depth study about DIY beauty in social media.

How to know if professionals in a field are influencers?

May 30, 2014


How to know if professionals in a field are influencers?


Here at eCairn we map the social web and constantly run experiments to increase our understanding of the social media. Recently we did an experiment in a ‘trendy’ career in the tech world, Data Scientist.


Our goal was to find out if professionals in the data science field often engage in influential conversations thru most common social media channels. AKA Twitter, blogs and facebook. This test in others words tries to solve the following hypotheses*:


H0: a techy guy into Data Scientist (job tittle/profession) will most likely be influential in the Data Science (DS) ecosystem.

H1: being into Data Science doesn’t make you a Data Science influencer


*The study takes into consideration a set of individuals actively generating content in social media for the past 6 months, and that had mention stuff related to data science as well.


In eCairn we have half a million influencers grouped into 1 thousand groups of interests called tribes. Data Science, in which the experiment was done, is one of the tribes.


The DS tribe has 1365 influencers broken down as follows:



After careful analysis we concluded that groups 1 and 2 above had a strong connection, therefore we regroup them as one.

The network map (below) shows on the on hand that the professionals with a Data Scientist job tittle [group 1] and those that perform Data Science tasks [group 2] within their companies often engage between them. On the other hand that, that influencers from other fields/specialties closely related to DS [group 3], such as Business Intelligence, Statistics, Machine Learning, Mathematics, Data and Big Data do engage with the first two groups, but in a lesser way.

Network DS


Blue: represents groups 1 and 2.

Yellow: represent 3


The red area highlights how some influencer overlap with DS professionals.


In conclusion, we ruled out the null hypothesis for this example and proved there is segregation among DS experts with its influencers. Rather there are more peer to peer conversations.


Here is a close up look into each of the group 3 influencers:


The least overlap is present in the statistics and Math influencers.

This is a sample of the top expressions used by the data science influencers out of 90k conversations:


New York Times – where are the missed clicks and views?

May 21, 2014

nytMost of you have probably already read the very interesting “leaked” audit about the New York Times. If not, I strongly recommend you do: Innovation Report.

One of the findings in this study is that there are many publishers that actually grow audiences re-purposing New York Times content, without the New York Times benefiting from it.

This was intriguing and we decided to investigate, using our database of publishers/bloggers/influencers (name them the way you prefer).


So we built 4 topics that we ran against our organized database of 600,000 bloggers, looking at everything they published over the last 6 months.

Our topics are defined as follows:

  • NYT Either is a mention of either the New York Times or a reference to a nytimes url.
  • NYT Both is an article where both the New York Times (as a brand) and the nytimes as a url are present (the good guys).
  • NYT Url is an article with a reference to a nytimes url but no mention of the New York Times as a brand.
  • NYT Name is an article where the brand is mentioned (New York Times), but without any link to the newspaper => these are the views that the New York Times is missing.

And here are the results. [Note that NYT Either should be the sum of the three others, which is almost the case – my topic definition is clearly not 100% perfect... but close.]


So, what we see is that 37% of mentions (59,687) of the New York Times don’t have a link to the newspaper. Pretty significant.

Looking a little deeper, we can highlight the vertical/tribes where the New York Times is mentioned the most:



No real surprise: Economy, Politics, Books are the tribes with the highest volume of mentions.

Now when we look at where (in which tribe) people use the name (New York Times), without references to urls, we get the following chart:

ImageBook comes first i.e New York Times best sellers.

If we compare and get a percentage of “use of the name without url”, we get the following graph:


Read this as follows: “amongst the Literatura Brazil publishers, only 3% of the people mentioning “New York Times” also include a link to the Newspaper”.

We clearly see that book as a category and language (two brazilian tribes in the top5)are the top problems for link back.

For book, this is related to “New York Times bestselling” expression.  Here is a quick look at the expressions mentioned the most with New York Times in the Book tribe:

Screen Shot 2014-05-21 at 2.51.49 PM

So maybe build a bio database for NYT Bestselling authors and make it easy and useful for publishers to link back?

As for the foreign references, maybe provide a short version in local language for the main foreign market like Spanish and Portuguese. After all, the NYT is advocating learning languages ( – see, I put a link back!). Just in the US, Spanish would be a good addition and as a global reference, it may be worth it to have local language versions.

Diving into another mainstream category where the gap is significant, we looked at the “Journalism” tribe and spotted the good guys – don’t want to get into trouble exposing the journalists that steal from the NYT ;-).

Here is the list of journalists with “high reach” that link back to the NYT:

 Screen Shot 2014-05-21 at 12.03.15 PM

First number is the frequency of link back, second the influence withing the Journalism community, third is a reach indicator.



Get every new post delivered to your Inbox.

Join 2,207 other followers

%d bloggers like this: