I was recently planning the logistics of an upcoming client workshop. I needed 12 copies of a document printed and spiral bound, and I visited the website of a printing company that we’ve used many times before for such tasks. The website had changed, and unfortunately I couldn’t complete the order. For some reason the website was saying it couldn’t deliver to my address.
I’m pretty sure I know why this is. I live in Portsmouth, on the South Coast of the UK, and to the uninitiated, some Portsmouth postal codes look similar to postal codes used on the Isle of Wight. I suspect some courier firms don’t deliver to the Isle of Wight (or charge extra as it’s an island with no roads connecting it to the mainland). This leads to some online sites (incorrectly) lumping some or all of the post codes together and tag them as an ‘exception’. This is really, really, bad design, but it definitely happens.
I was trying to place the order on a weekend, so I waited until Monday and went to contact the company by phone. I tried to phone shortly after 9, and then again at 9.30, and then again at 9.45. No reply. So, even though I’d used this company many times in the past, I just moved on to another supplier. And in fact, I’ll probably use this new supplier in the future, too. So the original printing supplier has lost a customer and it doesn’t even know that. Plus, it missed the opportunity to get feedback about the defect on their website… I wonder how many other cities/postal codes are affected? How many other sales are being routinely lost?
Considering The Customer’s Pivotal Moments In Process Design
As a business analyst, this experience made me think about process and operational design. While the example above was an example of bad design, it is impossible to design an IT system, interface or process that truly caters for every situation, nor (in most situations) would you usually want to. Sure, some call centers might have a process which defines the detailed steps to take if the President of the United States calls from a satellite phone while onboard Air Force One and asks for a message to be passed urgently to the CEO… but not many!
The point here is that there will be certain types of situations that are:
- Predictable, but very unlikely and/or uniquely complicated
- Difficult (or impossible) to predict, with unknown levels of likelihood or complexity
- Unintended, where with the best will in the world (and lots of testing) still something unexpected has happened which has led to an unintended consequence
The first set (predictable) are deliberately not fully catered for by a process as they are either so unlikely that spending time specifying them is overkill, or they are so uniquely complicated that anything beyond broad guidelines can’t be issued. I’d imagine that large companies have a “respond to media request” process which ensures that any inquiry from a TV station or newspaper gets to the right person. The broad process will be structured, and the response will likely be logged in a consistent way. However, how the response is formulated is probably somewhat variable, and more likely subject to guidelines and principles than a strict process. Responding to a request for a photo of the CEO to accompany a “top 10 CEOs” article is likely to be somewhat different to responding to notification that a documentary will be airing showing evidence of corruption within the company!
The second set of (difficult or impossible to predict) conditions can’t be catered for as they are unknown, or the effort of trying to predict them is so great that it is prohibitive. The final set (unintended consequences) are, by their nature, unpredicted! The key here is to find them when they occur and rectify not just the individual case, but the root causes. Taking my printing example, had I got through to the first printing company, I suspect they’d have quoted me via phone and manually processed the order. Great—except the website is still faulty and swathes of other customers might be affected. Understanding what needs to change to prevent the issue happening again is key.
So, what aspects can be considered when designing customer journeys, IT systems and/or processes to cater for these types of situations?
Flexibility, Feedback and Responsiveness are Key Factors
Assuming an organization wants to handle these types of cases, it’s key to design processes with feedback mechanisms built in. Feedback should of course include opportunities for customer or user feedback, but it can also include feedback generated by the process itself.
Take the printing company example I mentioned earlier. As a nationwide printing firm, they are almost certainly finding that there’s been a minor drop in Sales (Portsmouth is a relatively big city, but probably not big enough that the drop in printing sales would ring any warning bells) and the distribution of where they are sending parcels has changed. A curious analyst diving into the data might say “hmmm, it’s odd, there are entire cities where we are no longer sending parcels… maybe we should look into that”. Making sure diagnostic data is captured and examined is important, and this is so much more than just performance data.
It’s also important to ensure there’s a viable support option and, yes, this does usually mean ensuring someone can speak to (or communicate somehow with) a human being when they need to! That support person or team needs to have sufficient autonomy and be empowered to raise issues for investigation. A team that just “raises tickets” and passes them on to others is unlikely to cut it.
Finally, it’s important to note that processes will need to change and this should be expected. Building in responsiveness to the environment is important. Expectations will change, the way people communicate will change and so forth. By designing processes with this in mind, and ensuring they are owned, reviewed and adapted when needed, is a small but important step towards agility. As BAs, we can often nudge towards this way of thinking, and every step in the right direction is a good thing!