June 2015 round-up

All good things come to an end, and the sabbatical was no exception. As of July 1st, sabbatical consummatum est.

As I entered the final month of my sabbatical, I told myself that I ought to not stop being on sabbatical before it was over. And, that was exactly what I did: June was a full-on month, with talks, training, meetings and writing.

My #5pm pictures tell me that there were a lot of paperwork and music practice this month. And rain… though, rain is the last thing on my mind, today, which is the hottest day of the year, so far, in England.


This month’s highlights are captured below. I look forward to hearing yours.


The second survey that I mentioned last month has now been piloted. Changes were made, and the survey is now live.

I have been commissioned to do some work on the use of social media by business to business (B2B) organisations. And, after presenting at Britmums Live 2015 about the consequences of sharenting*, I have been sketching an empirical study on this topic.


* This talk looked at how data and metadata shared online by parents creates a digital legacy for their children and contributes to the development of profiles of their children as consumers


I submitted one paper on the implications of digital technology for strategy development in small and medium organisations, and worked on another one on customer screening. I had one paper on sentiment analysis published, and one on customer management accepted. And the Surveillance book has been featured in Oxford Brookes University Faculty of Business Research Newsletter.


Still, no teaching, as I am on sabbatical. Though, I have been planning the module that I will be teaching next semester, and someone gave me a very interesting idea for an assignment.


Lots! I surely made the most of my last month of sabbatical, on this front.

The month actually started on a negative note when I had to cancel stage 2 of my statistical course to attend some meetings. But, then, I caught up on training, with the following:

  • IMG_7393Staff development week at Oxford Brookes University, where I attended a session on resilience and came back with a handful of ideas to implement at course and module level. I also attended a session on using mobile data to develop customer insight;
  • A session on research impact – both designing impact into the research, and communicating it to project sponsors;
  • Research presentation and one to one chat with Professor Vicki Morwitz, at Said Business School;
  • IMG_7369Academic writing workshop at Loughborough University;
  • ERSC workshop at Nottingham University on the implications of technological developments for the Financial Services Industry – e.g. how it creates new competitors, and how it changes customer perceptions of fairness. I also gave a talk at this seminar, on the uses of social media data to segment customers.

What were June’s highlights for you?

Use of social media for segmentation in the financial services industry

On June 25th, I am delivering a talk on the potential and pitfalls of using social media for segmentation. This talk is based on research done with 11 financial institutions (1 credit card company, 3 insurance providers, and 7 banks), and some of the findings (early stage) are also discussed here.

Here are the slides:

Comments and questions are most welcome, as always. And…. if you are in Nottingham and see this post before June 25th 2015, join us :-) It is free. More information here.

Studying sentiment on Twitter is… complicated

Emotions are key to explain and anticipate consumer behaviour, and sentiment analysis offers marketers a way of measuring and summarising those emotions. Emotions displayed on social media conversations, in particular, are very appealing for research, as these platforms offer many opportunities to listen to the conversations in real time, with minimum disruption for the individuals expressing those emotions and in a cost-effective way.

Despite its promise and popularity, the sentiment analysis of social media conversations is neither a simple nor a straightforward process. Yuvraj Padmanabhan and I investigated the vulnerabilities present in processing and analysing Twitter data concerning consumers’ sentiment towards coffee. Our study, which has just been published in the Journal of Marketing Management, showed low levels of agreement between manual and automated analysis:

sent analysis manual vs automated

Source: Canhoto and Padmanabhan (2015)



Focusing on those messages where all types of coders agreed, it is interesting to see that they are most likely to reflect positive emotions, as exemplified by these tweets: ‘Found a euro cent on my walk and have a great cup of coffee in hand. Monday is already off to a good start’ and ‘Feeling much more alive this morning now that I’ve had my coffee’. Similarly, emotions that were clearly positive, like ‘joy’, showed higher rates of inter-coder agreement than those that showed passive emotions.

Conversely, problems arose when:

  • A negative sentiment was expressed but this resulted from the absence of coffee;
  • A sentiment was expressed but this referred to a different object (e.g. shift work), not the coffee;
  • The sentiment was not explicitly expressed but rather implied through cultural associations such as having coffee out, or through emoticons or abbreviations;
  • The user employed irony or sarcasm.

There were also cases where the sentence contained the word coffee and expressed a sentiment, but did not refer to the drink itself and, thus, should be excluded from the corpus of analysis.


In our study, not only were multiple types of software used, but also software products from both a commercial and an academic origin were employed. There were no marked differences in performance between the various products, indicating that this is not a failure of one product or the other but, rather, a challenge presented by the subject matter (emotions and sentiments) and by the channel, with its technical limitations and very specific culture and netiquette. These challenges are accentuated by the fact that the segments of text available on Twitter are very short, rich in abbreviations and slang, and often with typos or grammatical errors.

sent analysis manual vs automated 2

Source: Canhoto and Padmanabhan (2015)


So what?

These vulnerabilities have a number of effects on the use of automated tools to analyse sentiment in online conversations.

The first effect is that the problems with classification of tweets lead to an inaccurate representation of the overall sentiment towards coffee, both in terms of sentiment polarity and in terms of emotional state. The second effect is that segments of text that should be excluded from the analysis because they do not relate to the topic under analysis – coffee – are retained in the corpus of data, possible skewing the results. Given that so many commercial and academic research projects rely on the automated analysis of sentiment data, these findings raise concerns for the quality of those insights and subsequent decisions.

One of the reasons why using qualitative data analysis software may improve the credibility of a qualitative study is that the software enables researchers to make visible their data coding and data analysis processes. This is not the case with most automated sentiment analysis tools, given that the coding and analysis process is performed by algorithms strongly guarded by the commercial organisations that sell these applications. It is also concerning because researchers in search of speedy and inexpensive customer insight are unlikely to assess the robustness of the automated tools prior to using them, as we did in this study.


What can we do about it?

Our study does not aim to discourage researchers from using automated sentiment analysis tools, or Twitter data. Instead, our message is that researchers need to spend considerable time familiarising themselves with the technical and pragmatic aspects of communication in the social environment, and with the characteristics and limitations of the software that they may use to analyse social media data. To improve the classification of tweets, sentiment analysis needs to take into consideration the social context within which the conversation takes place, for instance by looking at the tweets before or after the one being coded, or considering wider patterns (e.g. more negative tweets on Mondays). Moreover, analysts need to consider the cultural connotations of the object that they are studying, including international variations – for instance, in Japan the consumption of coffee is associated with the idea of foreignness, whereas this is no longer the case in the United Kingdom. Additionally, it is important to keep developing dictionaries that reflect the specific syntax and style used in social media conversations, or even software solutions that, in the first stage of analysis, replace commonly used abbreviations with their formal equivalent – for instance, replacing BRB with ‘be right back’. However, it must be recognised that as language and communication styles are constantly evolving, these dictionaries and tools will never completely reflect the full variations and nuances in social media communication. Moreover, they will struggle to capture sarcasm and highly contextualised uses of language – for instance, teenagers using the term ‘sick’ to refer to a very good experience.


You can access our paper here.


What challenges have you faced when using Twitter data to study sentiment, and how do you deal with them?

Consumer Data and the ‘War on Terror’

The book that I co-authored with Kirstie Ball, Elizabeth Daniel, Sally Dibb, Maureen Meadows and Keith Spiller, has been featured in ‘Research Reporter’, the research newsletter of the Faculty of Business at Oxford Brookes University. The original article is here. Transcript below in case the link does not work for you.


surveillance bookSurveillance, consumer data and the war on terror

The government’s increasing requirements for businesses to support national security programmes has had significant implications. Ana Canhoto’s new, co-authored, book helps businesses prepare for the challenges they may face.


If you have flown in and out of the UK in the last couple of years, you might have noticed that you need to provide the airline with your passport details well ahead of departure, unlike when you travel between other destinations.


Likewise, you may have noticed an increased amount of scrutiny from your bank about the funds entering or leaving your account. These behaviours are the result of regulations that have been imposed by the UK government on commercial organisations, as part of its initiatives to combat crime and terror.


In order to support national security programmes, more and more businesses are required to share information such as customer behaviour. However, our research shows that these requirements are not in favour of the business’ commercial interests.


For instance, those surveyed in our study have been forced to buy expensive information systems in order to collect and process the required data. This has impacted on the businesses’ profit, the shareholders’ income, and has meant changes to the products’ prices.


It has also impacted on operations and ways of working, what customer information is collected or how certain transactions are handled.


We spoke to organisations who felt the requirements have created delays in their service delivery and impacted negatively on customer satisfaction (some customers have found the data collection intrusive).


The requirements have also created tension amongst staff, and additional work requirements, particularly for those in a customer facing role. Many participants also felt that the requirements conflict with the nature of their jobs, such as the emphasis on privacy and secrecy that are characteristics of the financial industry.


Institutions deemed to have made a sub-par effort in complying with the regulations are liable for heavy fines or even imprisonment. So, over time, firms have found ways of adapting to the government’s surveillance requirements.


Some firms have even managed to use the heavy investments and process changes to their advantage. For example, as a source of additional customer insight, or as an opportunity to provide superior customer service.


As more and more business sectors are enlisted in the government’s efforts to fight crime and protect national security, other organisations will be facing similar issues. Our new book The Private Security State? Surveillance, Consumer Data and the War on Terror can help organisations understand the challenges that they will face, and prepare for them.


Going along with customer lies can be good for business

Yesterday, child #2 lost a tooth. At night, he carefully placed the tooth under his pillow, for the Tooth Fairy… even though he knows, a-hem, the truth about Father Christmas and the Tooth Fairy. And, sure enough, this morning, when he checked under the pillow and spotted something left by the Tooth Fairy, he gave me a knowing wink, and flashed a big, toothless smile.

This episode reminded me of this study, which found that consumers that lie in order to get a particular outcome (e.g., a prize, a service, a refund) are more satisfied when they get what they want, than those that obtained the same outcome without lying.

The reason for the increased satisfaction, according to the authors, is that when

liars are busy lying, they have fewer mental resources available for other tasks. One such important task involves using feedback from the listener to update one’s expectations about how the conversation is progressing“.

So, when liars get what they want, they are more likely to be surprised by the outcome than honest customers. Pleasantly so.

It seems that in business, as in life, there are some rewards from pretending to believe in (fairy) tales. When would you go along with a customer lie?

What Nash’s work means for Marketing

We learned, at the end of last month, that the mathematician John Nash had died, with his wife Alice, in a traffic accident. Nash’s work on game theory, for which he won a Nobel prize alongside John Harsanyi and Reinhard Selten, has relevance to many areas of public life, from international trade negotiations to war strategy. And, as it turns out, it is highly relevant for marketing, too.

Here is how.

Thinking of decision-making as a game

Game theory conceptualises decision-making as a game with two or more players. Like with any other game, my choices impact on the other players’ outcomes and, thus, the other players adjust their behaviours according to my moves.

This means that, when we make a decision, we need to factor in how the other players might react to our actions. For instance, when weighting in the possibility of cutting prices, I need to think about what my competitors will do (e.g., start a price war?), and what that will mean for my profits.

In game theory, the modelling of the other players’ likely reactions is based on the expected payoffs of various alternative actions, for myself and the other players. In reality, this information is rarely available, at least completely. What we do have, however, is the company’s reputation (e.g., for always matching price cuts, for bringing innovative products to market, or for delivering great service). These reputations work as signalling mechanisms, telling us the likely actions or reactions of the various players, encouraging competitors to try and occupy different segments of the market, rather than fight head on.

Thinking of interactions as a non-zero sum game

Prior to Nash’s work, the analysis of negotiations was based on the principle of pure conflict. That is, the assumption that the two (or more) players have no interests in common so that one player could only improve their situation by leaving the other one(s) worse off.

However, Nash’s work framed negotiations as ‘non-zero sum’ games, in which the players have some complementary interests. So, they will maximise their gains by engaging in a mixture of competition and cooperation.

For a marketing example, think about the smart watch market. They can improve their situation by fighting each other for market share (for instance, through price cuts, or by tying customers in long term contracts or proprietary technology). But, they can also gain by cooperating with each other on things like agreeing technical standards, or educating consumers, in order to grow the overall market for smart watches.

In summary: 1) think about how other people will react to your actions, and 2) if you want to get a bigger slice, make a bigger pie. It’s quiet simple, really; and that is the brilliance of Nash’s work.

May 2015 round-up

A belated round up, as the last few days of May (and first ones in June) were very busy, including a couple of big pieces of work due around the same time. Murphy’s law!

Funnily enough, my #5pm pictures reflect exactly that: work features heavily in this month’s pictures, either directly (i.e., I was sitting at my desk, working), or indirectly (i.e., I was doing something else, like shuttling the kids to extra curricular classes, but I had taken work with me).


What did May bring you? My highlights are summarised below.


One survey piloted, and another almost ready to be piloted, meaning that June will be a month of data collection.

In addition, I had a really positive meeting with a company working in the customer service arena to discuss avenues for collaboration. We identified two streams of work that we could do together, including specific projects for the short and the medium terms.

I also had some discussions with colleagues in another university about developing a research grant application to work together on analysis of Twitter data. As you may remember, from my March 2015 round-up post, there is huge pressure on academics in England to attract research funding, which has some merits but is an extremely time consuming and frustrating exercise.

So, in terms of research pipeline, May was a very exciting, and very promising month.


Two papers submitted; and big revisions made to a third paper, followed by submission.

Whipped a fourth paper into shape, which is now with my co-authors for input; and contributed to a fifth one, where I am not the lead author.

DarkSideCover_zps9e4e743aAlso made final, minor corrections to the ‘Dark Side of CRM’ book, which is coming out in September. I am completely biased, of course, but I think that this is a really interesting book, and I cannot wait for it to hit the shelves. It is an edited book, which brings together the views from Customer Relationship Management (CRM) experts from academia and industry, to define and explore the dark side of CRM across the themes of customers, relationships and management. The book has lots of examples, and really brings to light how bad behaviours by both firms and customers impact on perceptions of trust and fairness, loyalty, endorsements and, ultimately, profitability.


No teaching this month, as I am still on sabbatical.


Quite a few learning opportunities, this month.

At the start of the month, I joined (virtually) in Google’s Education on Air event, and was delighted, entertained and, above, all inspired. You can catch some of the events here. I absolutely loved James Sanders’ session “How to Survive a Zombie Apocalypse” session, which was a seriously fun way to get people to work together, using tools such as Google Maps, shoutkey.com and, my favourite, iftt.com. Check the video – I think that James’s ideas are applicable outside of the classroom, too. Maybe for your next meeting?


A few days later, I had some training about how to investigate and respond to student complaints, which was a great reminder of the benefits and importance of putting things in writing.

IMG_7235Then, on the following week, I went to London for a workshop on ‘Researching Digital’, where various scholars shared experiences about (you guessed it) researching digital phenomena. We heard from work in areas as diverse as crypto currencies to digital drugs.


At another level, I learned a lot about British History, from Boudica and the Iceni tribes, to rationing and wartime shelters, by helping child 2 prepare for his end of year exam.

And last, but not least, I learned a lot about myself and about my work. Three completely different and independent events required me to discuss with others the ‘why’, ‘what’, ‘how’ and ‘so what’ of my work. And that, in turn, forced me to think very carefully about those same questions. It was a really great exercise to reflect on why I made certain choices at key points of my career, and what the consequences of those choices were; but, also, to define where I expect my path to take me. I am feeling reenergised by knowing / remembering where I am heading.

What were May’s highlights for you? Do you have a clear path ahead of you, or is it all looking a bit unclear at the moment?