I had an interesting insight a few days ago, one that I have had before. I have always loved the phrase “Nature Abhors a Vacuum”. While in the physics term, it normally means that if you have a vacuum, i.e. lack of anything, nearby physical elements will expand to fill it (generally limited to gases and, to some extent, liquids, but this is not a physics lesson and my undergrad engineering was almost 2 decades ago), this phrase has particular application in the psychology, and hence business, realm.
Quite simply, if you don’t clearly inform your customers/partners/employers/employees what to expect, they will make assumptions. There is an old line about never to “assume”, which I won’t quite repeat here, but assumptions can be dangerous or even deadly.
In the technology business sphere, setting expectations, the lack thereof that companies do, comes down to three areas:
- What features a product/service will have
- When a product/service will be delivered
- How a product/service will perform
In the army, in Hebrew, we used to call the second element a “tavlat yiush,” or “despair schedule.” You could put up with any awful post, as long as you knew when it was going to end. If you didn’t, well, nature abhors a vacuum, and people will make their own expectations, which will either be worse, leading to real despair and loss of energy, or better, leading to even worse despair when they are not fulfilled.
The last case is interesting. Most online businesses have some form of redundancy, across stacks of technology in a single data center or across data centers. These redundant stacks tend to be installed for their own reasons – sometimes redundancy, sometimes staged deployments, sometimes cold, warm or hot standby. But they often tend to gain other uses over time that may even eclipse the first. These uses – and sometimes even the original usage – lack clear Service Level Agreements (SLAs) as to what they will provide and when. Over time, the lack of clear SLAs leads to people inside and outside the company *assuming* (that dangerous word again) that behavior will conform to certain parameters.
I have seen this several times in the past few years. The solution is three-step:
- Figure out what one or two metrics of service level people are making assumptions about.
- Get everyone who can figure out what service level the company can provide right now, with no changes, into a room, and don’t leave until you figure out what it is right now. Make sure you can really commit to these today.
- Clearly publicize what today’s service levels are to every level of the company and even customers.
Of course, plenty of people will be really upset; after all, you are shattering their expectations. The most upset may be the CEO (with whom you should check before letting the data out to customers, as it may violate signed agreements). But this step is crucial, if you ever want to get to positive SLAs. Once this is done, then you can begin recovery;
- Get the same people into a room, and figure out what you need to do to improve service levels in at least 2, if not 3, ways.
Now you can go to the CEO and executive team, and say, “OK, you don’t like today’s, I understand that, but that is what there is. Now here are three options for improving them, with budgets and timelines. Which, if any, do you want?”
I have done this process multiple times, both as a VP in technology and as a consultant. It is difficult, challenging, and absolutely thrilling… when your technology division operates at the level people expect.