Delta Airlines employs more than 80,000 employees worldwide and has a fleet of more than 700 aircrafts and has 160 million customers.
It has a reputable social-media operation. Or so we have read...
In an article written by Dennis Schaal, Inside Social Media at Delta Air Lines - A Behind The Scenes Look on June 10, 2011, he mentions that they have a control room just for social media at their headquarters, and that they knowingly don't respond to all tweets, stating: "The social media staff doesn't respond to every tweet about the airline, and does its share of apologizing to customers."
At PAKRA, we decided to dive a little deeper and experiment with the Twittersphere for Delta Airlines and see exactly what is going on. We wanted to get few questions answered:
(1.) Since Delta knowingly doesn’t respond to every tweet about the brand, would they be more likely to respond to tweets that areincluded in the “Dreaded” hash-tag #Fail search? As you know, if
this hash-tag is included in a tweet about your brand, the customer is probably quite upset with your brand and you are at risk to lose them as a customer. Surely, Delta will respond to all of “these” tweets. Right?
(2.) Why does Delta not respond to all tweets? As we see from the article by Dennis Schaal, Delta has about a dozen "Social Assist Agents" on a 24x7 follow-the-sun support schedule. What stops them from handling more tweets, especially potentially harmful ones like those that include "#Fail" in their tweet?
We analyzed all tweets in a randomly selected time period that included the name "Delta" with the hash-tag #Fail in the tweet text.
(1) Response Time:
Only 26.3% of the tweets during this time period actually received a message back from Delta where they tried to assist the customer in some fashion. This means that 73.7% of the tweets that we analyzed went unanswered by Delta. Delta averaged 9 minutes and 38 seconds response time to the ones they selected to respond to.
(2) Types of Customer Complaints:
We looked at three broad categories: flight related issues, baggage issues, and other. "Other" included any varying type of complaint ranging from an issue with Delta.com to bad customer support over the phone.
Most of the complaints (51%) were in relation to a flight related issue. This means the customer tweeted about a delay, conditions on the flight, the crew or something else closely related.
(3) Patterns on how and why a customer's tweet received a response:
- Was there a pattern in which tweets got answered and which one's did not?
- Were there any specific criteria involved that decides who gets a response and who doesn't?
- Did it depend on the type of tweeter?
The model demographics of the customer included in our sample: A young professional of 32 (range 25 to 40 years) years of age, male, residing in the United States, with 651 followers on Twitter.
We found that there is no pattern or any specific criteria in the way tweets were selected for response by Delta. An example is highlighted below.
Tweet was responded by Delta:
Tweet that was not responded by Delta:
The first customer’s tweet seems less critical to respond, even though in our opinion should still be responded to. The tweeter, of course, doesn’t like that he is being delayed, but is making a joke referencing “The Rock” (Dwayne Johnson) and a movie he stars in, in reference to the flight being delayed because of a broken DVD system. The second customer’s tweet shows that he had 5 disappointing flights with Delta and even goes as far as telling the world he misses US Airways.
Which of these seems to cause Delta more damage? I would be willing to bet that tweeter #2 (no response from Delta) will soon no longer be a Delta customer.
None of the following variables have any correlation to "which" customer will get selected and addressed by Delta:
- Type of tweet
- Demographics of twitter profile
- Twitter handle's follower strength
- Timing of tweet
(4) Big "Aha" Moment:
73.7% of the tweets went unanswered. Should a Delta investor or their executive management team consider 73.7% as the size of potential lost customers? Does this information make other airlines want to woo them away? If you were an operations manager, will you take that risk?
We ranked each tweet on the following metrics:
0: Non Existent; 1: Poor; ..., 5: Excellent.
Perfect score is 20 with the implication that the higher the score, the better the grade.
We graded Delta Airlines on:
(1.) How well did they “operationalize” their service recovery and customer service processes using social media
(2.) How well did they promote their brand --- "Keep Climbing" with the 70,000 followers (that's Delta's potential customers) of the Twitter handles included in our sample.
PAKRA's grading system comprises four metrics: (See more on grading system at: PAKRA #OccupyCX Grading System)
(1.) Average Response Time: How quickly did Delta respond to tweets?
(2.) Missed Opportunities: How many tweets did Delta miss or not respond to?
(3.) First Issues Resolution: Was the issue resolved in a manner that appeared apparent to the customer?
(4.) Quality of Response:
- Did the response include empathy?
- Did Delta agents follow procedures/policies in their response? (Such as: "get them off the air", "no apology guarantee")
- Did the response include any information that made the customer realize that the agent understands their issue?
- Did the response overcome the objection? Was there an attempt to assist the customer?
- Was the response "humanized"? Does the response appear as though it was written by a human and not an automated / canned message?
Delta Airlines' scores:
1. Average Response Time: 3
- Delta responded to tweets with an average time of 9 minutes and 38 seconds.
2. Average Missed Opportunities: 1
- Delta left 73.7% of tweets un-responded.
3. Average First Issue Resolution: 1
- Only 28% of the 26.3% (7.4% of all tweets analyzed in our sample) tweets that were responded to showed some acknowledgement by the customer that the issue was resolved. However, none promoted Delta at the end.
Note: Dennis Schaal, in his article, mentions that most of Delta's exchanges take place through DM (Direct Message) after the initial contact. While this helps Delta with the tweeter with the issue, it can still harm the brand, as the public opinion of Delta may still not have changed. If the resolution is done in a public fashion, that may have a bigger positive impact on the brand.
4. Average Quality of Response: 2.96
All responses adhered to Delta policies and seemingly to their process. However, model responses from Delta's Social Assist team completely lacked personalization with almost no acknowledgement of the issue.
Delta scored 7.96 out of a possible 20 points and receives an operational maturity grade of "Awareness".
- Our questions to Delta Airlines’ Customer Service Managers:
(1.) What are the metrics of quality for the messaging? What are the leading metrics that measures good performance of your business process using social media channels?
(2.) What processes have you put into place that address concerns from customers, who are reaching out to you via social media?
(3.) With 70% and higher missed opportunities and seemingly no tweets that “promote Delta” after receiving response, are you concerned about the possible dilution of your brand equity?
Questions for you:
(1.) Have you used Twitter
to communicate with Delta or any other airlines? If so, what was your experience?
(2.) Would you use social media to communicate your customer experience rather than voice-calling customer service and/or completing web-surveys?
Disclaimer: This study’s intent is to identify opportunities for improvement for companies that are using social-media channels to optimize customer experience. The analysis is based entirely on publicly available data.
Follow us on Google+