This manifesto represents the private opinion of my co-founder & myself on how operations-heavy companies, and more specifically their operations teams should manage their business in order to maximise their performance, and thus the return on investment for shareholders.
We produced this based on our 20+ years of combined experience working at or consulting operations-heavy businesses at McKinsey, Google, Delivery Hero or OLX, and by interviewing more than 200 COOs and operations executives globally.
Operations management manifesto 2022
- Data availability - Companies shall collect & maintain clean data about their operations.
- Self-service - Operators shall be able to access that data & self-serve their needs, without having to wait for data or tech teams.
- Proactivity - Operators shall use that data to get proactively notified if something require their attention, and not browse zillions of dashboards or reports to figure out what to work on.
- Standardisation - Operators should have clear & up-to-date standard operating procedures (SOPs).
- Automation - Operators should automate as many processes as possible.
- Structured incident management - Whatever task or incident management cannot be automated, shall be managed via structured & collaborative processes.
- Ownership - Operational incidents can happen anytime, thus there shall be a clear ownership who is on duty to resolve them and when.
- Collaboration - Operations shall be closely integrated into other functions. Other areas shall inform operations in advance of all relevant projects, and operations shall escalate relevant & serious issues transparently throughout the company.
Data availability - Most companies already collect data about their operations, however, many encounter challenges keeping the data clean. By cleanliness we mean not only transformation to the right format, but ensuring completeness (i.e., no missing data points) and coherence (i.e., aligned definition or calculation method for the same metric). The modern data stack & its associated products are solving for the above, however access to clean data is still a significant challenge for most of the companies we spoke with.
Self-service - Operators work with data day and night, however data is often accessible only by the data or engineering teams. While we fully understand the reasoning (being security, coherence of data or lack of required skills), this limits the agility of operations managers to get access to the necessary data in a flexible and timely manner. 80% of the operations teams argued that they were too reliant on the tech teams and thus were blocked in their day-to-day work. Fortunately, adoption of self-service analytics is growing, but we found that is true mostly for startups.
Proactivity - We found that many operations managers spend 2-3 hours per day just by looking at various dashboards to spot issues or opportunities. We find this a reactive and error-prone approach, not to mention that it can quickly get repetitive and disengaging. Those companies that automated anomaly or event detection either used the in-built alerting functions of their existing tools (e.g., in Salesforce) or requested custom SQL alerts from data or engineering teams. While this is a good first step, we believe that proactive monitors and alerts should be similarly self-supporting and modifiable as dashboards or reports.
Standardisation - Fortunately, most operations teams have standard operating procedures (SOPs), documented mostly in Google Docs or Confluence. We received, however, many complaints that these playbooks are not continuously updated or not sufficiently adopted by the employees. In order to utilise these documents, we reckon that they should be closely integrated into the everyday task management tools of operators, so that they could easily access & if needed, update the knowledge.
Automation - Automation is the buzzword of the last decade, thus not surprisingly almost all companies have implemented some level of automation. There are some clear picks, such as digitalising the core workflows or automating the simple but highly repetitive processes with RPA. However, 9 out of 10 operations leaders stated they would like to implement more automated processes, if engineering teams had the capacity to prioritise them over some customer facing features. As such, there is still significant unlocked value in automating operations.
Structured incident management - We met companies that established detailed incident management SOPs with integrated task management tools (e.g., JIRA or Asana), but that is not the common practice yet. Most teams use their existing communication channels, such as email, Slack or Teams to investigate & resolve operational incidents that often results in lack of transparency, accountability & context. We believe that operational incidents shall have clear owners, deadlines, playbooks and given the interdependencies, they should be resolved via a transparent & efficient cross-functional task management tools.
Ownership - IT support teams have long resolved the question of ownership by implementing on-duty systems to resolve IT incidents. This is not a common practice yet in operations, and even where this exist, the allocation of ticket is rarely automated. As serious operational issues can happen anytime, we reckon there should be a clear process who is on duty and when, and this shall be supported by scheduling & automated ticket escalation tools.
Collaboration - Operations is closely intertwined with all functions of a company, thus some form of collaboration naturally exists everywhere. Most commonly this is done via the usual communication channels (e.g., email, Slack) or via regular cross-functional meetings. What we're seeing less frequently is the escalation of relevant & serious operational issues transparently throughout the organisation that could quickly inform other functions to take necessary actions (e.g., stop a marketing campaign). Furthermore, operators should strive for a structured, documented & transparent ticket-based collaboration approach whenever possible.
How to achieve
Achieving excellence on all points of the manifesto is a challenging task, however getting more and more critical to stay competitive.
Those companies that excel in all points needed to implement not only the right management approach, but the appropriate technologies, too.
The above mentioned is especially important for companies that make performance, availability or privacy guarantees to their users, and thus action is required immediately.
We're building Flawless with the mission to get operations teams perform as per the above-mentioned manifesto as quickly and easily as possible.
We would be curious to learn about your current status and most prevailing challenges and discuss if we can help based on our experience. Feel free to reach out either if you're just getting started, or if you are already more advanced in the journey.