Skyttle 2.0 introduces several important features:
– Improved sentiment analysis: the API now extracts larger stretches of text that express sentiment, instead of individual words, we provide context so the user can understand what keywords and entities the sentiment relates to.
– XML annotation: there is an option to receive the original text with XML-annotated keywords and sentiment.
– Cross-document keyword and entity matching. The API outputs IDs for each keyword it extracts, which helps match occurrences of different variants of the same entity/keyword in different documents.
– Keyword/entity extraction, sentiment analysis, and XML annotation can be done in one API call, but requesting either type of analysis will count as a separate request.
– Skyttle 2.0 supports English, French, German, and Russian for now, with more languages on the way. Our support for Russian is a new development and one we are very excited about.
There is still a free usage allowance, but in response to requests for enhanced features there are now also paid usage plans for start-up and enterprise use.
Because of the scale and scope of the changes we have made, the format of requests and responses has changed, please see the documentation on Mashape. A sample client script can be found here: https://github.com/vpekar/skyttle_clients.
The old API will remain in operation until 15 August, we encourage you to migrate to Skyttle 2.0 ahead of that time. Please get in touch with any questions and comments.
For developers working with the API these are important changes. The Boston quadrant is particularly significant.
It is worth listening to what Sippey says there: “we’re trying to encourage activity in the upper-left, lower-left and lower right quadrants, and limit certain use cases that occupy the upper-right quadrant.” In other words, bring your rich media to us, we want to own that bit. Don’t build clients or consumer-facing products, build B2B analytics, support businesses integrating with CRM, work in verticals.
It makes sense if Twitter can improve the horrible timeline experience (too much noise, too few ways of filtering or time-shifting the experience) which they have started to do by daily emails.
Facebook is looking to raise up to $16bn from the markets at a valuation of up to $104bn, money that will allow it to do amazing things. But the valuation it is seeking is based on it already having done those amazing things. And it simply hasn’t. Here is a researcher’s perspective on the IPO.
1. Facebook advertising must be a great revenue driver right?
No. Click-through rates on advertising are VERY low by industry standards. Source: http://www.computing.co.uk/ctg/news/2175024/facebook-advertising-failing-click-users-study
2. But all those big brands have piled in. They must know what they are doing, right?
No. Brand pages are not working well for big brands. Millions of Facebook likes do not allow individual FB users to be targeted by brands. In fact the brand’s updates soon disappear from the consumer’s newsfeed unless the consumer continues to engage. And less than 1% do. Source: http://adage.com/article/digital/sexy-brands-struggle-low-engagement-facebook/232993/
Bellwether consumer durable brand General Motors announced yesterday it was pulling its $10m advertising budget. WPP’s Martin Sorrell highlighted the “limited success” of Facebook for advertisers, noting that it hadn’t appeared to effect the IPO.
3. But Facebook is all about viral marketing, right? You don’t have to pay for media, because if the content is cool enough, people will pass it on on its own.
No. The maths doesn’t work here. Facebook messages that succeed in creating a “global cascade” in the jargon – i.e. go viral – tend to have a large number of starting points and relatively short chains (numbers of people passing the message on). In other words, something can go viral in Facebook if a lot of people hear about it from outside simultaneously (because of external advertising or news coverage). Facebook itself is not the means for this to happen. The limitation is the small size of each user’s social graph (the number of their Facebook friends). Source: http://snap.stanford.edu/class/cs224w-readings/sun09contagion.pdf
4. Hey, sourpuss, they’ve got 900m fricking users. That’s 70% of the world’s internet population. I mean – that’s got to mean something, right?
Yes, it is very big user base, but each of those users is valued at $100 or more. And at the moment, they aren’t doing that much for Facebook. In fact average revenue per user (ARPU) is falling at $1.21 per quarter, down 7% year on year. Verizon, for comparison has quarterly ARPU of $160. That’s revenue, not profit. And with such a high penetration rate it is hard to see where future user growth is going to come from. (It is worth remembering that AOL and TimeWarner once had an overwhelming domination of the internet access market. Things can change.)
5. But still Facebook has all that content produced by its users. There must be a business letting people search through all that great stuff?
Not one that Facebook can exploit. Facebook doesn’t archive or index its pages well, until recently the industry gossip mill suggested that everything was stored in cache. As a result it has had little visibility on its key asset – customer data. See: http://www.slashgear.com/facebook-data-archive-incomplete-say-privacy-hounds-12222664/
Partly in consequence the Facebook search experience has been very poor, making it hard to sell contextual ads. That’s another revenue model blocked to the company.
6. But all of those likes, that gives an advertiser really cool information about a user. That is gold dust, correctamundo?
It may be, and Facebook since 2010’s f8 conference has offered access to very valuable data, but that data requires explicit user opt-ins to be exploited. That means that it is most valuable to application builders like Zynga or Spotify. Facebook have struggled to make that data work for advertisers (see above).
7. And Facebook has really got mobile nailed, correct? I always see people using Facebook on trains.
Mobile is an increasingly important platform for Facebook, but they are scratching their heads about how to make it generate revenue, a fact they were forced to acknowledge in the run-up to the IPO. [More on this and wider challenges to Facebook's advertising model from BTIG's Rich Greenfield]
8. Hey smartass, look at the valuation of the business! All those clever people on Wall Street have run the numbers, they must know something.
They do: Facebook has decided to up the amount of stock it is selling in its upcoming IPO by 25%. But it is only selling a small quantity of stock 422m out of 2.7bn shares (15%) which may keep the demand high. That might provide a short-term bump to the stock, based on scarcity, but in the IPO insiders (who know the business) are selling, outsiders (who don’t) are buying. Source: http://www.ft.com/cms/s/0/80f2e7a0-9f59-11e1-a255-00144feabdc0.html#axzz1v74Q40uw
If you can get the stock tomorrow, buy it and flip it. Because long term Facebook’s quarterly numbers are going to be vulnerable to the issues above. It doesn’t have a proven revenue-generation model to justify that valuation, and without that the numbers don’t stack up.
Facebook can be a great business. It’s just that there is a barrel load of execution-risk in there.
In the meantime they are getting roasted on Facebook and Twitter. On Facebook commenters are noting that their remarks are getting deleted. When an answer is provided, it is cut and pasted. The CEO Philip Clarke is @clarkepatesco. He has remained mute as if these posts were not being made. @uktesco is similarly cutting and pasting.
This is inadequate. It is making a difficult situation worse.
What can Tesco do to address this? Field the director of HR to answer questions. Get him or her on every bulletin and the Today programme. Withdraw from Workfare immediately by announcing a review of how it works. Put similarly senior people to work on Twitter and Facebook. Take action and answer questions with thoughtful responses. Do not cut and paste, talk and engage. Do not delete posts. Expect anger and aggression and be ready to listen to it. Even with this approach Tesco will need a dose of good luck to weather the storm. Without it, they will suffer long term damage, which will effect every decision made by every stakeholder who deals with them, from the bright graduate who might think twice before accepting an offer to the local planning committee reviewing their application. This stuff matters.
[Update] It seems that Tesco have deleted all the posts on their Facebook wall about the issue, included those to which they have already responded. It is more than 2,500 posts and comments. Someone has been very busy. Will this make the issue go away though? It would seem unlikely. [Update 2] It seems that Tesco have not actually deleted posts, but some were removed due to violations. Fair play to Tesco: Jay in customer care says he is “working my way through Facebook’s automatic hidden post feature, and putting hidden comments back onto the wall which are not in any breach of the community rules.” Good for him. Tesco have also clarified their position on unpaid work, which is helpful, if a little tardy. Interesting to see Chief Secretary to the Treasury Danny Alexander as a “Friend” of the group Unfair Workfare on Facebook. If this is not a spoof profile (and appearances suggest it is real) it suggests a significant part of the coalition is already rowing hard away from this initiative.]]>
“Sure we monitor social media but – bleurgh! – we have no idea what to do with the data. There is so much!”
“OK, we know we have a problem with customer service, but the data on that never seems to change. Are we making progress?”
“We are spending a fortune on Facebook campaigns but although the engagement stats look good we have no idea if they add to the bottom line. Can you help?”
We have said for a while that the only real way to answer these questions was to look under the hood – inside the enterprise. We realise that you have to match all marketing spending (not just that on social media) with real-time data from the business. And that idea is at the heart of our new focus. We are putting social data together with all the other data in the enterprise, sales data, survey data, even linked open data from public sources, to create a full picture of ROI within the enterprise. There are some more details about our new approach here.]]>
A branding expert would say that this is no accident: it’s the emotional tug of a successful brand at work. We thought it would be interesting and useful to compare two brands with very similar inheritances: Oxo and Marmite. One has tremendous, contemporary appeal. The other seems to be dying in front of our eyes. And yet it could all be so different.
Oxo and Marmite were both invented by talented German chemist Justus von Liebig (1803 – 1873) who discovered that you can create novelty foods out of the by-products of the new food processing industry.
First came Oxo, created around 1840 when von Liebig devised an efficient method of producing beef extract from animal carcusses and trademarked as Oxo in 1899. Marmite came when Liebig discovered that brewer’s yeast could also be concentrated, bottled and eaten. The Marmite Food Extract Company was formed in 1902.
Both Oxo and Marmite are concentrated versions of other food products. Both were invented around the same time, by the same person. Yet, both brands have followed radically different trajectories over the past 100 and a bit years.
Today, Oxo is owned by Premier Foods (current share price 4.69) while Marmite is owned by Unilever (current share price 2083.00). Oxo is generally viewed as static, characterless and dusty, while Marmite is considered cool, witty, savvy.
What the data tells us
A look at social media quickly determines gross disparity between the brands.
In the time frame 4 October 2011 – 17 November 2011, we have:
This is more than a case of Marmite simply being more involved in social media; the Marmite brand is “healthier” than Oxo. It is more relevant, more current.
Conversations Determine Brand Health
This gulf in reach and appeal between Oxo and Marmite did not always exist. In fact, Oxo saw great success in the 1980’s with its “Oxo Family” TV ads staring Michael Redfern and Lynda Bellingham.
Marmite followed a similarly humble path, with the slogan “My mate, Marmite” chanted in commercials by an army platoon.
Why Marmite Succeeded
Marmite really took off when the brand used social cues from consumers to shape its branding. They realised that its distintive flavour had earned it as many lovers as haters. They played on this in 2009, when they invited netizens to join the “Marmarati“, a secret society invited to try an extra-strong version of Marmite, thus leading to great word of mouth and a lasting crew of loyal brand advocates. Marmite continues to play on the “love/hate” relationship in its branding, running a dual-skinned website and actively fueling the debate.
Marmite recognised that the social web word-of-mouth marketing means everything to a brand and consumer conversations determine a brand’s health. Not only that, but you can actually measure these conversations in terms of currency, advocacy and engagement.
Why Oxo Fell Behind
Oxo, however, has done little to adapt its identity to the social web. Their website is basic. They have no Facebook or Twitter presence. In fact, the brand seems to have little identity. More importantly, Oxo has failed to use consumers to inform their marketing. In fact, they appear to have done little to adapt their marketing to what consumers are actually talking about.
This is a shame: Oxo once had huge brand value and excellent positioning. It represented the family, the warm feelings of shared meals, home comfort. Now it represents very little.
Despite the dearth of conversation about Oxo, what’s there reveals that the Oxo brand IS worth something. It has potential.
The conversation that’s there is very positive and good-natured.
We’ve seen people tweeting to Katy Perry:
“@katyperry crush an oxo cube into ure beans and cook them.. You’ll be california dreaming ; p x”
Footballer Alex Oxlade-Chamberlain is nicknamed “oxo cube”:
Can some explain to me how Rosicky starts again and oxo cube doesn’t even make the bench? I’m beyond confused
Oxo is starting from a good base here. People have warm feelings about it, and even associate the brand with some major personalities.
What should Oxo do?
Oxo may or may not have a big marketing budget and during this time of austerity most brands are in this position. Fortunately, social media is free, and there are low-cost ways to benefit from it.
The thing to do is to opportunistically get involved in these conversations.
Oxo should do as Marmite have done: pick up on the conversational cues that are already there and ride it. Work on the Oxo-ness of Oxo cubes. Send Katy Perry and Alex Oxlade-Chamberlain a couple of boxes of Oxo cubes. Set up a stand selling Oxo outside Arsenal’s next home game.
Oxo needs to build on the cranky, old-fashioned Englishness of the brand to drive PR and conversation. By simply observing existing conversations, Oxo can discover what the brand means to the different people who discuss it, be it slimmers or football fans. The brand should try out strategies and themes of conversation, see what works, test it. When it finds a theme that resonates, build on it.
The key for Oxo is to just get out there, mix it up, and start making itself relevant. Learn from the success of Marmite and make itself relevant for the next hundred years.]]>
cultural shift (1)
bully boys (1)
works bus (1)
property shows (1)
class-coded oscillation (1)
countless young men (1)
record player (1)
black-and-white films (1)
sure many people (1)
bouquet of gladioli (1)
local leisure centre (1)
second-hand copies (1)
humdrum town life (1)
studio audience (1)
fey schoolteacher (1)
guitar-rock lineage (1)
yacht-rock opulence (1)
collective pride (1)
lead singer (1)
sexual identity (1)
dowdy jeans (1)
store ad (1)
prog-rock musicians (1)
spinning lights (1)
open-buttoned shirt (1)
body-popping moves (1)
least favourite (1)
mumblecore film-makers (1)
political refusal (1)
tape recordings (1)
Please leave answers below!]]>