Tagged with data

The latest developments in social media

I recently spoke at the Bath and Bristol Marketing Network regular event and shared some of the latest research in social media. The audience was a mix of small businesses hungry for social media lead generation tips and larger businesses and agencies looking for the next trends in social and how to apply social data to their business.

A very enjoyable event and some great questions were thrown my way to answer – if there are any more or you would further info on SDL SM2 or Social Intelligence services, please do get in touch.

++ Take a look at this blog post about some of my talk by @AgencyNurse >> Read ++ 

Tagged , , , , , , , , , , , , , , , , ,

The social media data behind christmas

xmasSoV

xmasDomains

XMASTHEMES

Looking at the online data behind the first week of December with the keywords ‘Christmas’ and ‘Xmas’ we can see certain interesting themes and therefore, opportunities.

As you might expect, the highest mentions of the search terms are to be found on Twitter and the like but that doesn’t always mean that is where you should focus your marketing activity. If we look at the second most popular destination for discussing Christmas, we can see that there is capital in the Forums. Forums have stronger networks, higher levels of trust and greater opportunities to convey your opinion beyond the length of a Tweet.

The conversations that take place on the likes of Mumsnet, Money Saving Expert and Offer of the Day are all about getting value for money, the biggest discounts and sharing this information with the network of community members to deliver tips and tricks to make it a more affordable Christmas.

When you drill down further into some of the conversations by platform, in the instance of the niche TES forums, there is a wealth of activity discussing what represents a sensible gift for teachers to give their pupils and lots of ideas on where to get the right gift for the right price. If you cant see the opportunity in that then give up now.

When we look at the themes of the online conversation, we can see that social, commercial and financial mentions are at the fore, religion is up there too but the underlying theme of Christmas lies in the bottom line.

Tagged , , , , , , , , ,

Using social data as a predictive indicator of business functions

Lick your finger and stick it in the air to determine which direction the wind is blowing. Pick a stick from the riverside and set it adrift downstream to judge the pace of the flow. These rudimentary attempts at predicting an outcome have as much value as judging a book by its cover. Sure it looks to be telling you something from all outward signs but underneath, predictivewhat is going on at the core?

Ever done that thing where you kick the tyres of a car to judge the suitability of purchasing it? You can make casual judgements but you can’t have a grasp of what lies beneath or which process in the grand scheme of processes are going to falter. Or can you?

Launching a product to market isn’t about the launch event itself, that is but a bit part of a string of processes and certainly not the end point of a product’s life. Tracking trends in healthcare is going to help you identify production demands for specific medicines and by location but behind the curve. If you could track online social data to witness an unfolding emerging trend or to identify a micro-process in product development that might be a roadblock, then being ahead of the curve will help you to act in the appropriate manner.

Data is the key to all of this insight and the affordability of social media data and the cost savings it delivers against in situ focus groups and other traditional research methods is where the value can be found. Many businesses place their faith in data based algorithms and indexes that deliver insight on a variety of business functions. NPS and CSAT scores are two of the most prominent measures of performance, both regularly viewed by investors and CEO’s as static scores, social data offers not only a less expensive and live view of business data but also indicates where a change within the business needs to occur, something neither NPS or CSAT offer.

The power of social media and the pockets of data that make up tweets, blogs, forums and other digital instances afford businesses a great opportunity to make informed business decisions. Want to find out more?

Webinar | Using social data as a predictive indicator

Tagged , , , , , , ,

What should I be measuring via social media monitoring?

Social Media Monitoring with Alterian SM2

‘Likes’ ‘Retweets’ and ‘Reblogs’ are all well and good as a shallow metric but not much more. Use them  as a guide indicating how effective your social activity is progressing.  Report and monitor the things that truly matter to your business, the deep metrics that deliver consistency of reporting and value in terms of business critical KPI (Key Performance Indicators). If your boss is only asking for these shallow metrics then you will need a considered and logical argument up your sleeve as to why they are not the right ones.

The things you should measure are the very behaviours you are looking to alter by implementing a change to your existing marketing repertoire. That is why you undertake a campaign or a certain strategy. To make something better. To improve leads, brand awareness, renewals or whatever it might be.

Measure the uplift, the money in the bank and the cost per renewal against the cost of acquisition. Compare and contrast. Benchmark your activity. Go historical if it paints a better picture but measure the numbers where the needle shifts from the start of new activity and where it ends up as a result of your considered marketing campaign.

Tagged , , , , , , , , , , ,

The APPrentice App challenge: What social media said…

The Apprentice has returned to the UK for another series of businessmen and women exchanging clichés, back-stabbing and showcasing “the best” of entrepreneurial Britain…and I love it! Perfect trashy TV, the twists and turns (and not just when it comes to the back-stabbing) and the mild peril of Sir Alan in the Boardroom are all part of what has made the show a success over the years. 

This week’s episode was particularly interesting as the challenge was for the two teams, boys against girls, to come up with a mobile phone App that would get the most downloads within a 24 hour period. 

Both teams went for soundboard type Apps, a quick and easy type of application that wouldn’t push the limits of developers but would raise a smile, if only for 30 seconds, when you press a button. In the instance of the boys’ App, Slangatang, was pitched as a “local vocal” soundboard that generated mild insults and sayings that were delivered in a variety of regional accents. The girls created an App that generated a random, yet irritating noise and went by the name of Ampi Apps. 

The power of television and social networks means that there is an abundance of data to be gained from those that Tweet-along during the show or fire off a missive on an App developers’ message board soon after transmission ends. So let’s take a look…

Volume

96 per cent of the conversations yesterday and this morning took place on Twitter, further illustrating the match-made-in-heaven relationship that has blossomed between the two media formats. 

35 per cent of the noise generated via Twitter was discussion around the girls’ Ampi Apps product and 65 per cent was attributable to the boys and their Slangatang. To reiterate this point on volumes by product, Slangatang was the only one to break into the Top 10 UK trending topics during the show. This is fairly interesting considering the actual outcome.

Sentiment

The social data indicates that sentiment around the Slangatang App registered 20 per cent as very negative and 15 per cent as very positive. Ampi Apps, the one that was downloaded the most, registered a 48 per cent score in expressions that would be considered as very negative in sentiment and an 11 per cent score for the other end of the scale, very positive. 

What can we deduce from this? Knowing that it was the media pick-up on the Ampi Apps that saw them over the line, there are certain questions to be asked. How many people deleted it in the blink of an eye? What ratings did the App receive at the time? Of course, the name of the game was to shift quantity; the brief never said anything about the quality of the App.

What did you think of the result? Did you contribute to the social media data during the show?

 

Tagged , , , , , , , , ,

Are we getting a bit too sentimental?

I can say with 100 per cent level of accuracy that when it comes to monitoring social media, sentiment analysis is one aspect of monitoring that we could talk on all day.

How accurate is yours – do you have stats on this? Are you 83 per cent accurate or 95 per cent? Lest we forget, one person’s positive is another person’s negative and so on.

Sentiment analysis is subjective and can often only be as good as the source dictionary, NLP list or human that makes the call. Other pitfalls are associated with accuracy across long form and short form content. A tweet by its very virtue of being only 140 characters is likely to render a higher level of accuracy in assessing sentiment than a long-form blog post of complex sentences, where multiple expressions of sentiment may be expressed within one sentence, paragraph or article unit.

This isn’t to say that sentiment analysis is wholly redundant. Alterian SM2 utilises a proprietary set of technologies including word parsing, weighting, proximity and Natural Language Processing to enable the most accurate and customisable sentiment analysis. It also allows you to develop and define customisable dictionaries, manually review and override sentiment, conduct mining in C# script and select multiple language sentiment. All that said, sentiment analysis should be judged with a pinch of salt. 

Sentiment accuracy can be improved through data cleansing, combined machine and human analysis and by using custom dictionaries that are specific to the linguistics of the area of study that you are conducting. The automated analysis of sentiment may not always be something you would bet your house on but it can give you great guidance as to areas that need further human analysis and exploration.

In the same way that social media monitoring isnt just seeking out negative comments about your brand and responding to ‘neutralise a threat’, social media monitoring is made up of many parts, of which, sentiment is one – a useful one but not THE one. Never forget the business need you are actually looking to address with your monitoring, be that looking to instigate a sentiment shift or otherwise.

Do we get too hung up on but one part of what social media monitoring is about? How important a feature of a social media monitoring tool is sentiment analysis for you?

Take a look at our Social Media ROI series

 

Tagged , , , , , , , , ,
Follow

Get every new post delivered to your Inbox.

Join 198 other followers