Facebook ha appena lanciato l’IPO che valuta la società 104 miliardi di dollari. Facebook ha circa un miliardo di utenti, e ha fatturato [solamente] un miliardo di dollari nel primo trimestre del 2012. La valutazione dell’IPO sta in pieni solo se si ipotizza una crescita significativa del fatturato.

Facebook può crescere ancora in termini di utenti, ma presto cominceranno a pesare i vincoli legati, se non altro, al fatto che la popolazione della terra è finita.

Più facile pensare ad un incremento significativo del fatturato pro capite, oggi solo quattro dollari l’anno per utente, circa un ottavo di quello che riesce a fare Google.

La crescita potrebbe venire anche da altri servizi non direttamente connessi all’advertising. Facebook potrebbe diventare – e in parte già lo è – una fantastica piattaforma di gaming.

Ciò detto, è difficile giustificare i 104 miliardi di dollari di valutazione senza ipotizzare una importante crescita del fatturato pubblicitario.

In realtà negli ultimi mesi il fatturato pubblicitario per utente è leggermente calato. Sempre più utenti accedono a Facebook via mobile e Facebook al momento non eroga pubblicità agli utenti connessi via mobile.

A causa delle ridotte dimensioni dello schermo di uno smartphone, inserirci dentro pubblicità potrebbe peggiorare l’esperienza utente, e questo per Mark Zuckerberg è fuori discussione. E’ facile prevedere (vedi acquisto di Instagram da parte di Facebook) che Facebook troverà presto il modo per distribuire contenuti pubblicitari su mobile senza impattare sulla godibilità del servizio.

Più preoccupante è la decisione di alcuni advertiser – in particolare General Motors, l’annuncio è di qualche giorno fa – di non fare più pubblicità su Facebook, in quanto giudicata “di scarso impatto” sui clienti.

GM – terzo advertiser degli USA, dopo Procter & Gamble e AT&T – spendeva circa 10 milioni di dollari in advertising su Facebook.

La cosa interessante – oltre al fatto che alcuni concorrenti di GM, ad esempio Ford, hanno opinioni diametralmente opposte sulla efficacia della pubblicità Facebook – è che GM ha deciso di non toccare i 30 milioni di dollari di spesa a supporto della creazione dei contenuti “gratuiti” legati in particolare alle fan page di GM.

Questo perché, sempre secondo GM, Facebook continua ad essere un tool molto efficace per ‘engage’ i clienti. In effetti le fan page dedicate alle auto riescono ad interessare, ad “engage”, un numero piuttosto alto di fan.

In media circa l’un per cento dei fan di auto interagisce con la pagina ogni settimana (commentando i post, o con un “mi piace” di un post,…). Oltre il doppio di quello che riescono a fare, ad esempio, le pagine dedicate alla moda ed ai prodotti di lusso. Solo le pagine dedicate allo sport, alle star e ai personaggi politici appassionano di piú.

Anche in termini di crescita del numero di fan, le fan page dedicate alle automobili sono fra quelle che crescono piú velocemente, poco meno dell’un per cento a settimana.

Fra i diversi gruppi automobilistici, Fiat-Chrysler conta il maggior numero di fan, circa quindici milioni di cui circa la metà sulla fan page dedicata alla Ferrari.

General Motors ha circa sette milioni di fan, distribuiti fra pagine “corporate” (la pagina GM), pagine “marchio” (ad esempio la pagina Chevrolet) e pagine “modello” (la pagina Chevy Camaro).

GM spende circa 4 dollari l’anno a fan in costi legati alla creazione dei contenuti delle varie pagine cui corrisponde un tasso di crescita del numero di fan e un tasso di “engagement” in linea con quello della industria.

Ottenere tante interazioni dei fan sulla pagina è importante, in particolare se si è presa la decisione di non fare piú pubblicità. Infatti Facebook decide quanta visibilità dare ai post della pagina sulle bacheche dei fan sulla base del numero delle interazioni dei fan.

I post di una pagina che non riesce a convincere i propri fan ad interagire vanno rapidamente a sparire dalle bacheche, salvo naturalmente che la pagina non voglia “sostenere” i propri post con un investimento di advertising.
Da qualche tempo Facebook offre ai brand la possibilità di avere una specie di “visibilità garantita” – in cambio di una fee Facebook garantisce che una certa percentuale dei fan vedranno i post della pagina.

Su Google, per essere visti, un brand può “comprare ranking” – attraverso Adwords o facendo ricorso a tecniche di Search Engine Optimization – o cercare di rendere il contenuto piú rilevante possibile in modo che Google lo mostri prominentemente in modo organico. Generalmente, per stare tranquilli, i brand fanno entrambe le cose.

Su Facebook una pagina, per fare apparire i propri post nella bacheca dei propri fan, può lavorare sulla qualità dei post, cercando di generare interesse e interazioni, o “comprare impressioni” acquistando pubblicità, ad esempio sotto forma di “sponsored stories”.

La differenza è che su Google si va per cercare un informazione (quindi la pubblicità di Adwords spesso quasi quasi fa comodo), mentre su Facebook si va per interagire con gli amici, e quindi la pubblicità “aiuta poco”.

General Motors deve aver fatto i conti e concluso che per ottenere un certo livello di interazioni con i fan ed un certo numero di impressioni era meglio spendere i soldi in contenuti che in pubblicità. Non è però affatto detto che il ragionamento valga in generale, in particolare per fan page dedicate a prodotti emotivamente meno “interessanti” delle automobili.

Here is the new version of our brandbook engagement dashboard. It is a bit squeezed in the blog. You get a wider version here.

Use – aka click on – the top charts to select the pages you are interested in analyzing. Then set the parameters on the right. On the bottom, a second dashboard shows the pages that have had larger positive jumps in their number of engaged fans relative to the previous 30 days – a possible indication of cool posting going on.

Enjoy.

Sometimes the talking about value that appears on Facebook pages is not updated by Facebook, that just leaves the value of the day before (the number of fans instead changes regularly).

For 99.999999999999999% of the world population this is both irrelevant and uninteresting.

For me it is not, since I use the metric to calculate the number of engaged fans (talking about minus new fans in the last seven days) and I might get incorrect results for those “repeated”days.

Starting today Brandbook’s engagement dashboard filters out the talking about data of the days when this data is not updated by Facebook.

We added some new visuals to our engagement dashboard, aggregating pages that belong to the same group. Since companies might have similar Facebook policies across their different pages/brands this view could deliver interesting insights.

This is the company ranking by average number of fans in the period (from November to now).

This is the ranking by average weekly increase in fans.

And this is the ranking by engagement.

Here are essentially the same numbers in the form of double scatter chart. Some groups do very well growing their fan base – Ferrero seems to be an example. On the engagement side however, you do not really see that many outliers, if you exclude the case of large sport fan page, and Anheuser-Busch, with its Guarana Antartica page.

As a page admin you might be interested mostly in how many impressions you get, how many people see your page. Having an impression means the post showed up on the news feed, not necessarily that it was actually seen – but it is certainly the first step.

Facebook rewards more impressions to pages that are more engaging, probably with a twist (probably because I have no hard data to prove this assertion). Facebook considers new fans as a measure of engagement, and new fans can be “bought” with ad spend. By buying ads on Facebook, you can increase your fan base and increase your edgerank number.

In exactly the same way, on Google you can pull up your “organic” ranking by buying adwords, since Google counts organic and not organic clicks to judge the “relevance” of results.

We define “engaging” as talking about minus new fans (basically we try to measure the fans’  feedback to the pages activity).

Is it “bad” if I have a high talking about number made up of a lot of new fans (some acquired through ad spend, other organically) but a low “engagement” number?

I think this depends on the “cost” of increasing “real” engagement compared to the “cost” of sustaining a given level of impressions via ad spend.

For instance, some pages might find out that the only way of increasing comments and likes is to post entertaining content that has little to do with the brand. For some pages/brands this might make perfect sense, while for others (more “specialized” products and services) trying to engage fans by posting about the Super Bowl might not be a sustainable editorial choice.

As a side note, there is little indication (as a very rough analysis of the talking about ratio might suggest) that large pages are less engaging than smaller ones (dark red bars). The difference in the talking ratio (blue bar) is simply tied to the weight of the “new” fans in the equation (orange bar) –  it is easier to grow 1% per day if you are smaller.

Having many pages supporting the same brand, especially if the all post in one language, costs more, and might not have significant advantages. I just noticed that Skittles “killed” its 3 million fan UK page and merged it into its main page, even though the main page only grey by less than a million fans.

I believe data (as collected, for instance, by Brandbook) can make these choices and trade-offs like these easier.

We added a new feature to our engagement dashboard that shows the most engaging pages of the day. This means the pages whose engagement rate has at least doubled in the last 7 days relative to the last 30.

Just a word of caution.

If page A has increased its engagement ratio a lot over the past week, this should mean that is has done something cool that works. Therefore you can learn something from page A and possibly apply it to your page, right?

Not necessarily.

While we try to separate “talking about” in “engagement” and “new fans” this is not always possible (sometimes the people talking about figure is not updated by Facebook for 2-3 days, while the number of fans changes).

Many of the pages that have spiked in engagement have also spiked in terms of number of new fans acquired per week. This might be correlation – something very engaging the page did convinced a lot of non-fans to like the page, or causality – the increase of fans due to an ad campaign has created some increase in engagement or is partially “read” as an increase in engagement.

So do not confuse a spike in engagement with a spike in the number of new fans per week that the page is getting. Both are great news, but the increase in fans might have little to do with something “engaging” that the page has done, and more to do with (an effective) promotional spend by the page.

The viz allows you do see the top engaging pages in different days (above those of the 22nd of February, below those of the 20th). This should allow you to distinguish “peaks” of new fans (probably tied to promotional activities, that of course can be very interesting and engaging and be much more than a simple ad campaign) from “peaks of engagement” – probably tied to some kind of more-effective-than-average interaction of the page with its fans.

Possibly.

I have taken the daily talking about/fans ratio of the last 40 days or so of about 700 pages, mostly with over 500 thousand fans. I have categorized these pages by product category. Every categorization is subjective, and a different categorization might have given (partly) different results.

The first chart shows the relationship between the talking about/fan ratio of pages relative to their pages size. There appears to be a somewhat negative relationship. While it is obvious that the bigger the page is the more difficult it will be to obtain a given percentage increase in the number of fans, the reason why a larger page is necessarily less engaging is a lot less obvious.

So the second chart shows a different ratio, the engaged/fans ratio, on the vertical axis. It is calculated by subtracting the last seven days increase (or decrease) of fans from the talking about number. The engaged/fans ratio is less sensitive to page size and might give a better idea of how effective a page is at engaging its fans.

I agree that there is a relationship between “engaging” and new fans – the more a page is engaging and the more new fans it will probably attract (all other things equal) and the more new fans a page gets, the higher its “engagement” edgerank parameter (all other things equal) and thus its impressions. Nevertheless, IMO, keeping new fans and interactions by existing fans separate helps understand what is happening.

This third chart chart shows the engaged/fans ratios for each category of pages. The size of the ball indicates the total number of fans of the pages included in the category. The horizontal axis indicates the average size of the pages in a given category (to check whether the difference in talking about/fans ratios is due to different average sizes of pages in different categories). The highlighted blue ball represents the 92 “confectionery” pages, that have a total of over 200 million fans. This is my third largest category after one that takes together all fashion, clothing and luxury, and a second that puts together media, news and internet services. “Confectionery” contains chocolate snack pages, chocolate spread pages, candy pages, bubble gum pages, and, yes, salted snacks. I would not have expected this category to be the one with the lowest number of talking about/fans and the lowest “engaged/fans” ratio. My candidate would have rather been “non food FMCG” which is where you find (surprisingly more engaging) washing powder fan pages .

Splitting the confectionery category up by product type doesn’t help much answer the question of why this category fares relatively badly in terms of engaged/number of fans ratio. The most important sub-groups in terms of number of fans (chocolate snacks, cookies, candies and mints,…) hover between 0.2% and 0.3%, while salted snacks, spread and chewing gum pages have ratios of between 0.05% and 0.15%.

Difficult to tell why this (ie chewing gum pages significantly less engaging than candy pages…) would be. Then again, I have no idea of what’s cool and engaging about washing powder pages either.

As further food for thought, here is the comparison between the confectionery and eateries category. I chose the latter because I see no big reason for one to be more engaging than the other, and because they have similar average fan page sizes (3 million) and similar total number of fans (200 vs 140 million fans). For some reason, eateries fare better both in terms of engagement and in terms of total talking about ratio (horizontal axis in charts below).