Southwest Flight 2001 San Jose to Los Angeles. It’s been 2.5 months since Southwest changed its boarding policy from A/B/C groups, with first in-line, first on the plane to sequentially numbered boarding passes and numbered signage in the boarding areas indicating where to stand in clusters of five boarding pass numbers.

I’ve now done five legs under the new system and have come to regard it as a reasonable improvement. I like the Exit Row, so often was willing to sit on the floor for 40+ minutes to be at the front of the rush for the line. Under the new system being early will only let you meet the other people in your cluster of five people, so you don’t need to be in line until just a few minutes before boarding. However, it does take a while to implement any change of this magnitude. In the waiting areas I’ve noticed bewildered passengers not sure where to stand, others double checking that their fellow line standers are in exact order (e.g., stand in line in order of A16 to A21), people asking if they have to sit in seat A21, and people just unhappy with the greater number of signs in the waiting area.

While there may be some growing pains, I think Southwest has done an amazing job with a very difficult implementation. Their method of boarding policy change is a solid example of intertwining social and technical organizational change. There is huge complexity in this change. Not only did they have to revamp enterprise, public-facing technologies (e.g., the on-line check-in system, the kiosk check-in systems), they had to train their personnel, and develop user-friendly information and signage that their broad variety of passengers can easily understand. Southwest’s success metrics are likely are whether passengers think this is what the airline rated #1 in customer satisfaction should be doing and whether their passengers still get on the planes more quickly than those of any other airline. Had Southwest focused only on the software or employees, and not done what they could to help passengers make sense of the whole experience, the result might have been a nightmare for the company and passengers both (this Southwest sponsored blog provides a variety of positive and negative observations about the change).

Contrast this with the experience of using “point of sale” (POS) terminals to swipe your credit or debit card to pay for your purchases. While the glitches we encounter there are much less traumatic than not getting the seat you want, they are a nice contrast to the Southwest experience. I’ll use a major grocery store chain as an example, but these issues seem to cross company lines. This grocery chain had implemented user POS terminals around the same time as did other chains. The POS sit on the little counter we used to use to write our checks. They are physically obvious so are likely to trigger sensemaking – customers see them and realize they are there to be used. This was a change from handing your card to the checkout person and having them swipe the card. Again, the change was physical and so likely to engender sensemaking around the idea that they were for customer use. The user experience beyond this point, however, is not standard. Some systems want the card swiped early, others don’t accept the swipe until after all the items are entered; some systems have the “accept” button on the lower right, others on the lower left; etc. These variations are the result of software and user-interface differences.

The particular grocery store chain I have in mind has recently “upgraded” their POS user interface. The change is all software and so is less likely to trigger user sensemaking around the change than would have a physical change. There was no effort to support the users through the change (most obviously that the swipe must wait until all items have been entered – which interestingly actually increases time in line). As a result, the checkout people have created handwritten signs with instructions. Kudos to the people on the front lines for addressing the social aspect of the change that the implementers did not.

Certainly the above examples vary in many regards (no one company has control of the POS systems, people deal with a variety of POS systems often while they may fly rarely, etc., many of Southwest’s customers use the web interface and so have access to additional information about the change). However, Southwest’s navigation of this difficult process is an exemplar in terms of fully managing a complex process. They have taken the range of stakeholders into account and adjusted both the social and technical aspects. I also expect that they are tracking user comments (see the blog above) and boarding speed in order to know if the system needs further adjustments.

The takeaways for me are that organizational change is difficult – especially when you are involving customers – and that the implementation is likely to go better when both the social and technical dynamics are taken into account. The management field often uses Trist and Bamforth’s studies of changes in 1950s English coal-mining as an early example of managing these “sociotechnical” issues (they changed team structures, equipment, and methods together in an effort toward “joint optimization”). I hope we can have some discussion of more modern examples (positive and negative). I’d also love to hear how Southwest came to manage the process in the way they did. Do they have this social and technical focus for all of their changes?