About this series: This is Part 2 of the ITIL in Practice series — four posts drawn from real ITIL implementations across nearly two decades. Part 1 covered Incident and Problem Management at Tyco, where a global team of 160 engineers had no consistent measurement or escalation process when I arrived. This post moves forward to AudienceView — a SaaS ticketing platform for entertainment venues — and the Change Management challenge that every fast-moving software company eventually faces: how do you govern change without strangling the development velocity that makes your product competitive?
When I joined AudienceView as Director of Client Support, the team was dealing with a number that stopped me when I first saw it: P2 response times averaging 13 to 14 days. Priority 2 issues — significant customer-facing problems that weren’t full outages but were meaningfully disrupting clients’ ability to run their business — were sitting for nearly two weeks before customers received a meaningful response.
That’s not a performance problem. That’s a structural problem. And when I dug into the causes, the picture became clear quickly. AudienceView was a sophisticated SaaS company — ticketing and venue management software for theatres, arenas, and festivals. The development team was genuinely talented. When something went wrong, everyone mobilised to fix it. The culture was one of urgent, collaborative problem-solving.
The problem was what happened before the fix. Changes were being pushed to production without a consistent review process. Nobody was systematically checking what had changed before an incident was reported. Nobody was notifying customers before changes that might affect their workflows. And there was no formal mechanism for testing changes before they hit a live environment. The result was a steady stream of unplanned incidents — not from carelessness, but from the absence of a process that should have been there.
The case for a CAB process — and why it’s always a hard sell in SaaS.
The Change Advisory Board (CAB) is one of the more misunderstood elements of the ITIL framework. In many organisations it gets implemented as a committee — a group of people who meet regularly to approve changes, most of which are straightforward and could have been handled without the meeting at all. Done that way, the CAB creates friction without creating protection. Engineers rightly see it as bureaucracy, and they find ways around it.
The version I introduced at AudienceView was built around a different principle. The CAB’s job wasn’t to approve every change — it was to ensure that every change was categorised correctly before it moved, and that the appropriate level of review was applied to the appropriate level of risk.
ITIL distinguishes three types of change: standard changes (pre-approved, low risk, well-understood procedure), normal changes (require CAB review before implementation), and emergency changes (expedited path for critical fixes, with post-implementation review). Getting that categorisation right is the real work of implementing Change Management. Most of your changes, in a mature operation, should be standard — routine deployments that follow a known, documented process and don’t require committee sign-off. The CAB exists for the minority of changes that carry genuine risk.
The pitch to the development team was straightforward: the CAB process isn’t here to slow you down. It’s here to protect you. Every incident caused by an unreviewed change is a P1 bridge call, a post-mortem, and a customer-facing explanation. Compared to the cost of that process, a pre-deployment review takes minutes. The math is obvious once you show it — but you have to show it, because the friction of the review is visible and immediate, while the incidents it prevents are invisible.
What the CAB process actually asked.
The framework we implemented answered three questions before any normal or high-risk change reached production.
First: what is changing, and what does it affect? Not a technical specification — a plain description of what the change does and which customers, products, or workflows it touches. This question alone surfaces a surprising number of assumptions. Engineers often know exactly what they’re changing technically but haven’t explicitly mapped which customers will feel it.
Second: have the relevant people been notified? Customers before changes that affect them, not after. This was a cultural shift as much as a process shift — the instinct in a fast-moving SaaS environment is to ship and communicate simultaneously, or to communicate after. We moved to communicating before, consistently, as a condition of the deployment going ahead.
Third: what is the rollback plan? If this change breaks something in production, what is the path back? How long does it take? Who executes it? A change without a rollback plan isn’t a change — it’s a gamble. This question, more than any other, changed the quality of conversations the development and support teams were having before deployments.
The Incident Manager role — and why it exists separately from the support manager.
Alongside the CAB process, I created a dedicated Incident Manager role at AudienceView. In the early months, I was running major incidents directly — on the bridge calls, managing the internal communication, updating the status page, coordinating with engineering on resolution timelines. That’s not sustainable, and it’s not the right use of a Director’s time.
The Incident Manager owns the P1 process end-to-end. Not the technical resolution — that belongs to engineering. The Incident Manager owns the management of the incident: the moment a P1 is confirmed, they log it in the system (which triggers notifications to all relevant internal stakeholders), update the customer-facing status page so clients know we’re aware and working before they start calling, open the Slack bridge channel for internal coordination, establish the bridge call with engineering, and drive the regular cadence of updates throughout the incident until resolution.
That cadence matters more than most people appreciate. During a live incident, customers are not primarily frustrated about the outage — they’re frustrated about the silence. A status page that updates every 30 minutes with a clear, honest description of where things stand keeps most clients from escalating, because they know they’re not being ignored. The Incident Manager’s job, in a very real sense, is to manage the information as carefully as engineering is managing the fix.
After resolution comes the post-mortem. What happened. What was done to resolve it. What the corrective actions are to prevent recurrence. And — where the change management connection closes the loop — whether the incident was caused by a change that went through the CAB process or bypassed it. That data, accumulated over time, becomes the argument for the process. You can show the before and after. You can demonstrate that incidents caused by internal changes dropped after the CAB was implemented. The data does the advocacy work that no policy document can.
SaaS moves fast. ITIL was designed for slower environments. Here’s how you reconcile that.
This is the real tension in any ITIL implementation at a software company, and it’s worth addressing honestly. ITIL was originally designed in an era of much slower release cycles — quarterly or annual deployments to on-premise environments where every change was a significant event. Modern SaaS companies ship continuously. Some deploy multiple times a day. Applying the full ITIL Change Management process to every deployment in that environment would genuinely kill development velocity, and the engineers would be right to resist it.
The reconciliation is in the categorisation. A routine deployment that follows the same tested process every time — same checks, same rollback procedure, same notification template — is a standard change. It’s pre-approved. It doesn’t go to the CAB. It moves on its own timeline according to the established procedure. The CAB review is reserved for changes that introduce genuine new risk: new integrations, significant architectural changes, modifications to core workflows that haven’t been deployed this way before.
Getting that categorisation right requires a conversation between support and engineering that most SaaS companies haven’t had. What do we agree constitutes a standard change? What triggers a normal change review? Under what circumstances do we invoke the emergency change path? Those definitions, negotiated once and documented clearly, make the process lightweight enough to coexist with fast development cycles while still providing the protection that a formal review offers.
At AudienceView, P2 response times moved from 13–14 days to 2–3 days over the two years I was there. The CAB process was one part of that — it reduced the volume of unplanned incidents that were generating P2 tickets in the first place. But it worked because it was designed for the environment rather than imposed on it.
The COVID test — and what it proved about documented process.
In early 2020, the entertainment industry effectively went to zero. AudienceView’s clients — theatres, arenas, festivals — closed overnight. The support team was reduced by 60% through furlough. Case volumes dropped, but they didn’t drop proportionally — because the clients who remained active had more questions, not fewer, about how to manage their systems through an unprecedented situation.
The processes we’d built — CAB, Incident Management, post-mortems, the Incident Manager role — meant that the skeleton team operating through the worst of the pandemic wasn’t improvising. The playbooks existed. The escalation paths were documented. The knowledge wasn’t locked in the heads of the people who’d been furloughed. A team that had previously depended on institutional knowledge and personal relationships to function would have struggled badly in that situation. A team with documented, repeatable processes was able to maintain service levels that I was genuinely proud of given the circumstances.
That’s an argument for process that no normal period makes as clearly as a crisis does. When everything works, you can debate whether the structure was worth the investment. When you’re running at 40% capacity through a global pandemic, you stop debating and start being grateful you built the thing when you had time.
Hutch Morzaria is a Director-level CX and Support Leadership professional with 19 years of experience building global support organizations across SaaS, Fintech, and enterprise technology. He has hired dozens of support and CX leaders across his career and holds ITIL Expert certification across V3 and V4.



