Some of you might have noticed that Monday’s Logistics Viewpoints email arrived a bit later than usual. This happened because the email service we use, Constant Contact, was down unexpectedly for several hours on Monday. The service failure threw a monkey wrench in my daily process and disrupted my work plans for the day. Sending out the daily email, however, is not a “mission critical” task for me. In other words, the impact of sending the email an hour or two later than usual was minimal in the general scheme of things, and I was still able to get other work done.
But this experience got me thinking: what if I was a transportation planner and my software-as-a-service TMS went down unexpectedly from 7:00 AM to 3:00 PM on a Monday?
I haven’t come across any “horror stories” of SaaS TMS failures. Maybe they’ve occurred and I just don’t know about them. But I suspect these occurrences are very rare. In Constant Contact’s case, for example, one customer posted a comment on Twitter that this was the first time they’ve had an issue with the service in five years. Also, compared to the IT departments at most companies, SaaS providers generally have better backup systems, processes, and controls in place to prevent disruptive failures.
Still, this experience reminded me that procuring a service is very different than buying traditional software. There are certain factors that customers need to consider as part of the evaluation and contracting process. In his guest commentary last month, Cliff Lynch outlined several of these factors, and 3PLCEO posted the following comment shedding additional light on the topic:
“I would add a couple of key indicators that demonstrate performance on some of these dimensions. First, is the provider SAS 70 Type II certified? If the provider has not undergone the rigorous testing of internal control procedures, it should raise a red flag regarding their commitment to excellence. Second, regarding technology, has the logistics or hosted technology provider outsourced the data center to a world-class provider and facility that can withstand hurricanes, tornadoes, and power and communication outages, and is there a true disaster recovery backup site that can be operational quickly in a worst case scenario? Too many providers cut corners by operating production data centers in a normal office environment. In other words, make sure your provider is not all talk and no action.”
In terms of contracting, here are three items customers should pay attention to:
System Availability
What is the service provider’s minimum commitment on system availability? How is system availability defined and measured? Below is sample wording I’ve come across in the past:
[VENDOR] will use commercially reasonable efforts to provide system availability of no less than [PERCENTAGE] during any twelve month period. [VENDOR] will publish statistics with respect to the overall system availability at least once per calendar year. System Availability is calculated based on the number of minutes of uptime during any twelve month period as a percentage of the total number of minutes during such period. For the purposes of this calculation, all scheduled maintenance periods shall be included as uptime as will any other periods during which System Availability is affected by acts of God, war, governmental action, strikes, riots, fires or any other cause beyond the reasonable control of [VENDOR].
Data Repositories
How much data—a month’s worth, two month’s worth, more—is available on-line? How often is the data backed-up, in what format, and where is it stored? For how long does the service provider archive the data—one year, three years, five years? Can the service provider destroy any data after the archive period ends, and what is their process?
Scheduled Maintenance Periods
How often does the service provide perform system maintenance and upgrade activities? Are these activities performed during the weekend? During what hours are these activities performed, and how long do they last? Are customers notified ahead of time of scheduled maintenance activities? How much notice is given?
**
As a side note, this experience also underscored for me the growing role of Twitter in business. When I couldn’t log on to Constant Contact, I conducted a search on Twitter and discovered that many other folks were having the same problem (you can read the tweets here, some very informative, many just complaints). I then posted a tweet alerting Logistics Viewpoints followers that the daily email would be delayed due to a problem with Constant Contact. Shortly after my posting, I received a reply from Constant Contact via Twitter apologizing for the inconvenience and letting me know that they were working on the problem. I also received a reply from a Constant Contact competitor encouraging me to consider their service offering.
As I wrote in a previous posting, business users (just like consumers) are using social media sites to comment (vent, hype, critique, etc.) about business-related products and services. If you don’t have a pulse on what customers and others are saying about your company, positive or negative, true or false, then you are missing opportunities to promote your products and services, or to “set the record straight” and protect your brand.
