A YEAR ON IF ONLY BA HAD REALISED INDIVIDUALIZED RESPONSES VIA SOCIAL MEDIA DON’T WIN HUGE ONSITE CRISES
By Anthony Olabode Ayodele, Chartered PR Practitioner (Chart.PR) MCIPR
It had happened before! British Airways (BA) once found itself in a similar situation. Around a year ago:
September 5th, 2016, to be precise, the airline’s passengers were stuck in the airport at the check-in-process stage due to a – wait for it – ‘undisclosed IT glitch’!
The problem which started in North America by the next day saw the negative impact on BA’s operations, spread worldwide!
Fast forward to May 2017!
Not only was the experience and official explanation to what BA passengers suffered (at UK Airports: Heathrow and Gatwick) on May 27th similar, the airline’s management adopted approach towards checking the problem was virtually the same!
Then British Airways had dwelt mainly on responding to people individually via Social Media such as Twitter! Many customers complained of lack of information. There was no authoritative voice at the scene - is the airport - available to address the situation. The May 27th fiasco reflected the same tactical approach. BA's CEO, Alex Cruz, admitted as much when, while explaining three days later on T.V his absence in the news since the crises broke out he said he was 'busy communicating with passengers via Social Media (tweets and videos)!
There were some improvements from the September 2016 approach though. It seems BA may have learnt from Delta Airlines, who, when suffering a similar problem early August 2016, ensured that their official website had information on the matter.
As early as on the first day the May 27th reported I.T glitch started, Cruz, in addition to individual twitter responses
had posted a general video response of him addressing the issue on BA's website. Additional video response were posted May, 28th and 29th, giving those who were able and in the position to access BA'a website, up-to-date official responses on the situation.
The problem is when crises strike not all of those who experience it know what to do or where to go to. The onus lies on the organisation facing the crises to take all explanations and probable solutions to where those affected most and need same can get them! In this instance, it would obviously be at the scene of the problem and/or in the news! Well, the later, as much as it is a national affair!
Not doing this is where BA erred.
Customers stranded at the airport or those who reckon they will be inevitably be stranded soon begin to panic. Newshounds covering such beats hope the situation might escalate and give them an easy cover story! Not having an authoritative voice at the point of crises or even updated information relayed over the electronic screens at the scene allows a bad situation go awry: 'By the time Cruz appeared on television tensions had risen to such a degree that he was put on the defensive. BA had allowed other parties (frustrated passengers and unions, mainly; ...) to provide the narrative'writes Seth Arenstein for PR News online on the matter.
Beyond, all mentioned, the subsequent negative online commentary the lack of a solution generates, and how long this lingers on for, is a major reason why companies like BA should avoid poor crises management of such a delicate situation. In 2016 when the BA power outage incident occurred the problem didn't last that long. It was over within 24 hours and by September 6th everything was restored and up and running.
The situation, post-May 27th, was different. Three days later it was still hectic for many and BA was still at loss explaining where on its website customers could get required helpful information. Cruz, made an appearance on BBC explaining and defending his position not to resign while noting he spent more time addressing the issue via Social Media.
Below, based on analysis done by PR and Digital Marketing outfit, Ideas Concepts and Products (ICP Media) - using Google Cloud Platform (GCP's) Machine Language Natural Language API tool - one gains an insight into the variance in social media sentiment towards the organisation established over the separate incidents based on two randomly selected articles (also calculating the Net Promoter Score (NPS) of each ) posted online, shortly after each incident occurred.
Then we look at the simple elements of a crises Comms Plan as detailed in a webinar tagged: 'CIPR Webinar A Crisis Quick Fix' by Magnus Carter FCIPR towards suggesting how BA could/should have identified and addressed the matter.
British Airways’ Check-In Glitch Causes Major PR Problems
Written by Mark Renfree shortly after the September 5th incident (click title to follow URL link to article) he notes that a lack of proper communication left a 'more lasting impression' than spending a few hours delayed at the airport'!
The results returned by GCP's NL API as analysed reveals :
The entire document returned a negative sentiment score (for an understanding of terms such as score, magnitude, salience etc used in this article read: MACHINE LEARNING: DOES GOOGLE CLOUD PLATFORM’S NATURAL LANGUAGE API SERVE AS THE MISSING LINK FOR MEASUREMENT IN PR? ) of 0.2 with a magnitude of 7.8 (1.56). - see screenshot image below.
The article presents a lot of negatives (as seen by the red squares in the 'document and sentence level sentiment' analysis) with the most negative statements (two) both rated at 0.7 (score) 0.7 (magnitude) (- 0.49).
Both statements, reproduced below and following each other in the main text, reflect how impersonal BA's responses via Twitter to clients complaints were:
NEGATIVE STATEMENT 1: 'Anna Walters Hewson (@AnnaHewson1) September 6, 2016, While the company responded directly to many complaints made on Twitter, the New York Times said the responses closely resembled statements made to news media'.
NEGATIVE STATEMENT 2 'Each response on Twitter was nearly identical—an interesting take on communicating with a hostile audience'.
Including the title - which is considered even more negative than statements in the text, returning a - 0.8 (score) and 0 .8 (magnitude) (-6.4), the negative sentiment BA's handling of the issue has attracted via this review of the situation, can easily be seen.
To calculate the Net Promoter Score based on the sentiment in the article (which is derived from subtracting the negative sentiments from that of the positive sentiments ) the total negative sentiment for the article (including the title) is -3.36 (being the sum of the negative scores - red squares - multiplied by their corresponding magnitudes.)
While, the positive sentiment in the article is just 0.45 (being the sum of the positive scores - Green squares - only two in this instance - multiplied by their corresponding magnitudes).
The only two statements considered positive in the article and their corresponding score/magnitude are provided below:
POSITIVE STATEMENT 1: 'Unlike other outages experienced by airlines like Delta and Southwest, British Airways focused on responding to people individually on Twitter' - Score (0.6) Magnitude (0.6).
POSITIVE STATEMENT 2: 'The undisclosed IT glitch is taken care of and flights seem to be resuming their normal schedules'. - Score (0.3) Magnitude (0.3) .
Hence, the positive sentiment is calculated as: 0.6 X 0.6 = (0.36) + 0.3 X 0.3 = (0.09)
= (0.36) + (0.09) = 0.45
The NPS for the article is:
0.45 - 3.36 = - 2.91
The entire article has a positive sentiment of (1.56) and a negative Net Promoter Score (which, when defined in this case, indicates the writer's commendation or condemnation of how the organisation handled the crises) of - 2.91.
A Chart of the sentiments (score multiplied by magnitude with negatives in orange positives in blue) is seen below:
 SECOND ARTICLE
The second randomly selected article reviewed BA's May 27th, 2017 incident, titled:
Under the 'Document & Sentence Level Sentiment' the NL API returned for the entire article returned a score of -0.1 and magnitude of 23.7 (see image with scores above). Multiplied, this gives the article a negative sentiment of - 2.37.
Two statements in the text had the highest negative rating of 0.9 (score) and 0.9 (magnitude) each. They are reproduced below:
NEGATIVE STATEMENT 1: 'Having an IT issue is never good, but having one during a busy holiday weekend is worse'.
NEGATIVE STATEMENT 2: 'The airline owes it to the tens of thousands of passengers it inconvenienced or stranded to conduct an unbiased investigation into what happened and to report back on the findings.”'
Overall the article had a total of 21 negative statements- marked out with red squares (the title was considered more subject friendly this time around returning a neutral score and magnitude of zero each) with a total negative sentiment (sum total of score multiplied by magnitude) of -10.96.
On the other hand, the article returned 12 statements considered positive! The most positive being:
MOST POSITIVE STATEMENT 'That’s true, and the apologies were well crafted (see the first video) and included the appropriate apologetic tone'.
The statement follows up on a previous one in the text alluding to fact that even though BA didn't provide adequate information at the airports, Cruz had used the right language in the video(s).
The statement returns 0.9 for a score and magnitude each giving, when both are multiplied, (0.81).
The total positive sentiment for the article (sum total of score - green squares - and corresponding magnitudes) is 8.33.
Hence, the Net Promoter Score for the article is: 8.33 - 10.96 = 2.63.
In all, the entire article has a negative sentiment of (-2.37) and a negative Net Promoter Score of - 2.63.
From the analysis, one can see - from these randomly picked articles (written by separate authors after each incident) dealing with a similar issue almost a year a part, BA has not improved its online sentiment regarding how it has handled the crises.
As produced for the first article above a Chart of the sentiments for the second article is produced below (with negatives and positives charted separately due the charts scale). Click on image to view an enlarged version :
 ENTITIES SENTIMENTS
Also of interest would be the Entities analysis for each article showing the reflected sentiment for key words of interest. (Again, check the article linked here, for meaning of relevant key terms such as Salience, Sentiments and Magnitude.)
In the article reviewing the September 5th, 2016 incident among the first set of entities analsyed (see image below) the first mention of 'British Airways' (item 2, top right) has a Sentiment: score of zero, a magnitude of 0.7 and a salience of 0.07; producing a neutral sentiment when multiplied - considering the zero score.
The first mention of 'flights' (being what BA sells) in the write up, however, has a Sentiment: score of - 0.1 and magnitude of 0.1 (- 0.01) being negative, and a salience of 0.03 (- 0.0003).
Thus, the write up has a neutral sentiment first mention of 'BA' and a negative sentiment mention of 'flights'.
The May 27th (first set of Entities produced in image below) article, in its first mention of the organisation (BA) has a sentiment: score of zero, magnitude of 1.3 and salience of 0.01. The second mention of passengers (who make the flights possible) however produces a negative sentiment with a sentiment: score of -0.8 and magnitude of 0.8 (-0.64) and a salience of 0.01 (0.0064).
Going through the entities (keywords) analysis for the entire article one can see the changed sentiments for each one.
Following Magnus Carter FCIPR advice in his CIPR Webinar, 'A Crisis Quick Fix' British Airways, through it's In-house PR department or PR agency, in contrast - in addition to ensuring management representation is found constantly at the scene to douse the tension - should have conducted a crisis audit anticipating potential crises: asking ‘what if' questions such as 'what if most passengers are too perplexed to check BA's social media platforms for instructions' during the crises?
Not doing such from a PR perspective can only see the organisation establish further negative sentiments towards its brand.
Consulting a PR Agency/Practitioner/Consultant that has a strong inclination towards Sentiment Analysis (such as Ideas, Concepts and Products) to manage its online sentiment would be the best next natural step to take.
Anthony Olabode Ayodele is a UK - based PR Practitioner and Digital Marketer. He is also the CEO of Ideas Concept and Products (ICP Media). He is a Chartered Institute of Public Relations MCIPR Accredited PR Practitioner and an Institute of Direct and Digital Marketing IDM CPD 2017 Award recipient.
Contact PR and Digital marketing outfit, Ideas Concepts and Products (ICP). to have a Sentiment Analysis conducted on Social Media articles Social Media comments, and posts related to your organisation or a general idea of the temperature of your Social Media outing and presence.
We can also help you to evaluate different opinions and articles as they relate
to each other on your organisation/person and using a baseline of zero (for score and magnitude) as a benchmark, we will help you to conduct a campaign to beef up positive sentiments in your favour designed to inclusively fetch you earned and shared media that will return you to an overall positive Social Media sentiment score.
We can also conduct in-house training for outfits and practitioners on how to go about using the Google Natural Language API tool to do same.
CONTACT: bodeayo67@yahoo.co.uk, ideasconceptsandproducts@mail.com Mobile: 00447495337259 Landline: 00442072532133
|
Comments
Post a Comment