The Microsoft outage on November 13 raised questions not only about the reliability of major technology providers, but also about how their partners manage the consequences for customers.
iCatalyst director Andrew Devitt was frustrated by some of the criticism directed at Microsoft following the disruption. The incident also prompted iCatalyst to rectify a few aspects of its preparedness.
Devitt shared his perspective in a post on LinkedIn and accompanying blog. Here, he also shares his view with us on the industry’s reaction, how iCatalyst responded, and what technology partners might take from the experience.
Q. You pushed back against some of the criticism of Microsoft following the outage. What do you think that criticism got wrong or overlooked?
Andrew Devitt, iCatalyst: After reading a recent techpartner.news article covering the recent Microsoft outage, I felt some feelings of frustration with the response which was then exacerbated by some of the responses I noticed across my network and even within our team.
While I think many critics conflated issues, the article spoke particularly of the dangers of single vendor concentration risk. I think this should become a risk/reward discussion. As co-founder of a Microsoft Business Applications business, I am sensitive to my bias on this topic so I took the better part of a week to reflect on my belief in what our business stands for and the faith we put in one specific vendor in Microsoft. This is true of our internal operations: we’re a real Microsoft shop. We practice what we preach. A few days on from the outage my view hasn't changed: one well-secured Microsoft tenant, backed by Microsoft's significant investment in R&D, security and AI, beats a collection of independent tools, adding in complex integrations with many points of failure. Spreading across more systems sounds safer, but it can add more ways to fail. Is the risk of single vendor concentration higher than spreading that same gross risk across multiple smaller vendors without the same means to respond in an incident?
I agree with Astha Singh and Dr Joseph Sweeney’s comments in the article: the answer is layered resilience: having quickly deployed failovers in place. Some of which should absolutely be human!
Q. You said parts of your response weren't perfect. What specifically didn't work as well as you'd hoped, and what did the experience expose about your own preparedness?
Andrew Devitt, iCatalyst: In my personal and professional network I witnessed the pile on. The default response was “thanks Microsoft for the day off”. Lots of AI-generated images about Microsoft being down and anyone with a Microsoft-shaped axe to grind getting very active on LinkedIn.
We closely monitor and support our customer's Business Applications across Microsoft Dynamics, Power Platform and Azure, including Co-Pilot. Our monitoring, alert and notification strategies in this space are very good. We probably had a blind spot in a couple of specific areas: relying on agents for monitoring/notification and ability to communicate with customers on multiple different platforms.
Like everyone, we’re rapidly deploying agents across our business and driving incredible improvement and efficiency.
Friday really brought home the importance of humans in the loop and humans on the loop.
This then brought us back to our core values. We talk a lot about being high touch. We like face to face meetings and we value our team being in the same location as our customers. So we hit the phones and we hit the road to meet with our customers where necessary. Thankfully the biz apps we support were up so it was a fairly simple call to our customers: “You focus on your big issues and rest assured your ERP/CRM core business application is up, we are monitoring it and will keep you informed. One less thing to worry about.”
Honestly, as a 5-year-old company some of our plans and processes around failover weren’t as solid as they should’ve been. We already had backup and continuity plans but we found a few holes in our process. We rectified that quickly and I think it cost us about an hour (thankfully non-critical for our business). We’re now making sure our customer facing teams are fully across those plans, particularly the communication plans.
Q. What had you done before the incident that helped manage your customers' expectations, and were there any gaps you hadn't anticipated?
Andrew Devitt, iCatalyst: I’d expect our customers are used to the sound of our voices and seeing us in person. That is part of our brand. So their expectation should be that they will hear from us with useful information. Our Managed Service customers are also very used to our Support Portal and we were publishing regular banner updates on that portal throughout the day. Thankfully, our Microsoft Dynamics back end was unaffected on the day. This was important because it is where our customer contact data sits. Without that, we would’ve struggled to communicate as we did. We’ve taken that as a learning.
Q. Have you made changes following the outage that you think other technology partners should consider in their own businesses?
Andrew Devitt, iCatalyst: Obviously having Business Continuity and Failover plans for all of your critical systems so they can execute immediately.
We quickly implemented an SMS alert service (opt-in obviously) so we can get critical alerts out through one more medium. It wasn’t something we had in place previously and it is now just another option for us in communicating with our customers. Make sure you have every possible communication method and I think also train your customers to have multiple communication touch points with you. Make it the norm, not the exception. All of that is important on a tough day. Don’t only do it in an emergency, make sure they are used to face to face meetings with you. Make sure they log tickets on your portal as well as via email. Make sure the communicate with your team via Teams and Phone Calls.
We’re also investigating sensible and secure methods to have failover access to customer contact data. This is about some layered resilience, much like Dr Sweeney advocated for in the article.
Andrew Devitt has been in the ERP and Microsoft Business Applications channel for nearly 30 years, having worked across ANZ, the UK & North America growing Microsoft Partner organisations. After 12 years with UCX Eclipse / DXC, Andrew founded iCatalyst with Dennis Chan and Justin Martin. Now one of the premier Microsoft AI Business Solutions, iCatalyst employs over 80 dedicated Microsoft professionals with offices in Perth, Sydney, Melbourne and Brisbane.
techpartner.news provides a platform for guest opinion articles such as this one to foster debate within the IT industry. The views expressed in these guest opinion articles are those of the author and do not necessarily reflect those of techpartner.news or its publisher.




