Research Computing Teams #144, 5 Nov 2022
Hi!
I keep hearing from leaders and members of teams where retention is a big problem. It starts small, with a single person leaving, and then a steady cascade of other people follow.
Worse still, there’s a worrying asymmetry between what I hear from the team members at these institutions, and the leaders and managers.
The managers seem convinced that industry’s salaries can’t be competed with, so there’s nothing that can be done.
But when I talk to the team members, salary is only one of a list of problems that come up. Team members talk about:
- No regular communication with their managers
- Having no growth in their jobs, or professional development
- Doing repetitive or disconnected work that doesn’t add up to anything that feels meaningful, and
- Not having a say
Retaining our team members is something we can do a lot about.
It’s easy to focus too much on things we can’t control (like salary bands). That leads us to the very convenient hypothesis that “there’s nothing we can do”. As people of science, we should reflexively be skeptical of our hypotheses, especially those that don’t require anything of us. There are teams in very similar situations that do retain their team members: indeed, team members enthusiastically recruit others to join. It’s in our power — it’s our responsibility — to create the environment that leads to such teams.
Most people who want to work in research are more interested in the work itself than the salary. Just because people leave and the job they got has a higher salary doesn’t mean they left for the higher salary. People who are interested in large salaries have typically self-selected out of research some time ago.
We can improve retention by focusing on things that matter to team members, like meaningful work and career growth
We know what people are looking for in jobs! Take a look at Understanding Factors that Influence Research Computing and Data Careers (#133) or the charts in How To Drive Your Own Career Growth In These 6 Easy Steps (#141). Regardless of field, people want:
- Opportunities for career progression and professional development
- Good work life balance
- Good enough compensation to feel respected
- Impact
- Flexibility in work arrangements
- Visibility
- Influence
We can improve retention on our teams by focusing on things that matter to team members. Improving retention helps us, but more importantly it gives the team members the work environment they deserve.
This Means Focussing on the Basics
How do we know what matters to our individual team members? How do we make sure they’re given stretch opportunities that are meaningful to and feasible for them? We start with the absolute basics - one-on-one meetings. That’s the data collection pipeline which will build relationships, alert you to issues, and help you know what each individual team member wants.
Address the issues they raise
Your ability to change things isn’t limitless, but you’ll (re)build trust with team members faster when you show that you take their input seriously and address what you can.
If one of the issues raised is impending burnout, do something about that now. If your team members are close to burning out, that’s entirely on you. Managing your team member’s workload is a key managerial responsibility. This may mean your team will have to stop doing some things, and focus solely on the areas where they have the highest impact or are otherwise highest priority. So be it. You’ll have to do the hard work of pushing back on community members who are demanding that other work to be done. No one said our job was easy.
Help your team members grow in directions that matter to them
With some of the immediate fires put out, we can use what we hear in our one-on-one conversations to start proposing some work for team members that lines up with directions in which they want to grow. Give them stretch opportunities that are feasible, that line up with work that needs to be done, and give them the support in the stretch opportunity to help them succeed.
The right amount of attrition of team members is not zero.
None of this is to say we don’t want anyone to ever leave our team.
It is good and healthy for team members to leave when the time is right for them. We want to see our team members move on to new opportunities that are right for them, even if that means leaving us.
Some turnover is actually an opportunity in fairly stable teams like ours; it allows for fresh perspectives to come in. But when people leaving becomes a cascade, that’s a symptom of management debt, of issues that have been left to fester in the team for too long. The departures cause further problems, hurting morale and leading to overwork for the people that stay.
It takes a lot to make people leave a stable job doing work they care about; once that barrier has been crossed, it takes a lot of work to rebuild relationships back to the point where people are staying not out of inertia but because they want to be there.
But the work is the usual work we should be doing anyway. Having proper one-on-ones to build individual relationships; giving clear goals and objectives; providing adequate resources; giving regular feedback; and finding work that matters to your team members. This is bread and butter management 101 stuff.
Resources I like for this
- The Stay Interview as a way of diagnosing immediate problems, and Stay Conversations to try to damp out some fires: those would probably be part of one-on-ones;
- “Why Are My Employees Quitting?”: 5 Ways to Stop An Employee Turnover Wave
- I’m biased of course, but I like my Starting One-on-ones QuickStart Guide and references therein, and my one-on-one meeting template: letter, A4;
- Having Career Conversations and An Engineering Team where Everyone is a Leader about arranging for growth opportunities
- Getting your team to speak up - How to Get Your Team to Challenge Your Ideas and You Might Not Be Hearing Your Team’s Best Ideas
- Let’s Discuss Attrition
Any thoughts on putting out attrition fires? Do you have any questions you’d like to ask, or would you like to be one of the first team managers or leads interviewed for RCT? Email me at jonathan@researchcomputingteams.org, or arrange a quick call.
With that, on to the roundup!
Managing Teams
Hack Your Bureaucracy - Marina Nitze and Nick Sinai
This book is definitely going on the very short Official Research Computing Team Book List™ that I recommend to people new to our line of work.
Nitze and Sinai have assembled a compendium of techniques for getting things done in large organizations. It is super tactical, almost to a fault, full of specific tips for getting things moving and enacting lasting change.
I’m not going to be able to summarize everything that’s in here, but there are several common themes. I’d sketch them out something like this:
- The Organization’s job to deliver stability and repeatable outputs.
- The Organization is designed to withstand and repell incoming new ideas and change
- The Organization does not, can not, care about your new idea or your important goals.
- However, the Organization is composed of people, of individuals
- The individuals can care about your ideas and goals
- Individuals have their own problems they’re wrestling with, and their own incentives driving them
- Individuals may have power and influence which will never show up on any org chart or intranet document but are nonetheless very real
- Because of people, and the rules of the organization itself, you can in fact effect change if you focus on specific problems to be solved and don’t get too attached to any particular solution
- The rules themselves have gaps
- What isn’t explicitly forbidden is likely allowed, even if reluctantly ..
- … and be sure to check the receipts on what “everyone knows” is forbidden
- The gaps and handoffs between fiefdoms are uncontested (neither fiefdom cares), can often admit change, and often have real problems that can be fixed
- Sometimes you can make the policy yourself
- There are often classes of activities (“pilots”, “tests”, or other bounded, temporary measures) that don’t trigger the full weight of the rules and can be used as testbeds and to demonstrate success
- Get something small done
- Then another
- But understand and be sensitive to the history and vocabulary of your organization. There are booby traps everywhere. People have been burned before, usually by someone else’s well-intentioned idea for change.
- The rules themselves have gaps
- You can build on successes to achieve bigger wins
- There is always people-time slack in the system
- Find the doers
- Understand the intricate web of relationships, influence, and “guilds” in your organization
- Don’t try to get it all done in one go
- Make it easy for supporters or marginal supporters to endorse your efforts
- Make it easy for opponents to oppose your efforts a few notches less fiercely
- Maintain relationships with individuals, and not just your supporters
- Enlist external community members
- For lasting impact, get the rules or incentives changed
These are people who have effected real change repeatedly in large complex organizations, and it shows. It’s a terrific book. It’s also very aligned with this newsletter’s point of view, which is that doing important work (like managing our teams) is doable, and doesn’t take a particular personality type, off-the-charts charisma, or particularly profound insights into the nature of humanity — but it does take some thinking ahead, attention to people’s needs, and sustained, persistent effort.
I hope a second edition expands on a couple of sections. In a section about knowing how your organization really works, mentioned for a scant couple of pages is the idea of building a personal “stakeholder map”, documenting the undocumented web of power and influence and priorities in your organization. Putting together one of those is the sort of thing one could profitably write (or read) an entire book about.
Why Feedback Hurts (and that’s optional) - Roy Rapoport
A good reminder for those receiving and giving feedback: one way feedback can unnecessarily hurt is when it sounds like it’s about the person itself, rather than some behaviour they demonstrated.
When we get feedback that hurts, one way we can take it better is depersonalizing it, and making it about some action performed rather than us as a human. And when we give feedback, we can usefully remember that focussing on behaviour and impact, regardless of feedback model, can avoid unnecessary hurt.
Navigating power dynamics as a manager - Pat Kua
Something new managers frequently have a tough time understanding is that they are now The Powerful One in their relationship with their team members. It’s especially hard to wrap your head around this when you’ve been promoted within your old team; you’re still just that same person, right?
Well, wrong. You’re now the person who can fire others, or at least write poor performance reviews. Everyone else is exquisitely aware of this, even if the new manager is kind of oblivious to it for the first while. That means others start behaving differently - deferring to your opinions, less willing to contradict you.
Navigating this means changing your own behaviours. Kua’s article gives five important ways to do this, and why:
- Be the last to contribute
- Explicitly invite others to contribute
- Be more deliberate about your actions and comments
- Build trust with each individual
- Explicitly delegate
Technical Leadership
How to run an effective retrospective - Gregory Witek
As I’ve mentioned before, retrospectives are to Management 201 (managing the team as a whole) as one-on-ones are to Management 101 (managing individuals). They are the fundamental tool of having a technical team improve how it works together.
Witek has thee key principles for running retrospectives that work:
- Always improve
- Stay realistic
- Everyone participates
and has very concrete, simple recommendations for how to run them, tools to use, agendas, and choosing what to work on improving.
Managing Your Own Career
Speak Up to Get Better Feedback in Your Next Performance Review - Karin Hurt
Hurt has some great advice if you get some infuriating feedback (or lack thereof) from your manager again around performance review time: some examples are
- “I don’t have much end-of-year feedback for you. You know you’re doing great.”
- “I rated you as meets expectations for your end-of-year feedback. Your performance really was an “exceeds” but I had to make the math work out.”
- “I know we haven’t talked about this before, but…”
- I don’t really have any specific examples, but it’s become a real issue.”
- “I’ve gotten a lot of feedback from other people about your performance in this arena. Who? I’m not at liberty to say.”
- “Just write up your accomplishments and I’ll sign it.”
In each case, she gives three or so specific scripts for how to proceed.
The underlying techniques are great worth highlighting. Some are about coming into the review with specific areas you want feedback in and being politely insistent on getting it (“Thank you! You know, one area I’m really working to improve on is ___. What is one suggestion you have for how I can be more effective in that arena?”). Others are about explicitly but politely stating your unhappiness and possibly making an ask (“I’m a bit surprised by this feedback and would like to take some time to digest it. Let’s set up a follow-up in a week to talk a bit more.” or “Oh, wow. That must have put you in a difficult situation. And, I’ve got to tell you, that makes me feel really _” or “I’d like a chance to better understand this issue. Who do you suggest I talk with to learn more?”)
And of course we can use this article as a checklist of things not to do in performance reviews with our team members.
Product Management and Working with Research Communities
The Helmholtz Association, which runs 18 research centres across Germany, has set a policy which makes open science a requirement:
The Helmholtz Assembly of Members adopted the Open Science Policy in September. This policy stipulates that scholarly articles, research data, and research software be published openly. Open science thus becomes the standard for publication practices.
Random
Functional programming is the future, just like it has been for the last thirty years.
How did Télétel, the French internet of the 1980s, work?
Decker is a retro-feeling HyperCard inspired no/low code tool, but with modern tools like easily deployable via static HTML, data query languages, and more.
Shell of a New Machine is a client-server dotfile manager.
Best practices for managing your BibTeX bibliographies.
It shames me how often I have to look up SSH tunneling options for non-trivial situations; maybe this visual SSH tunnel cheat sheet is just something I can print out and leave up somewhere.
That’s it…
And that’s it for another week. Let me know what you thought, or if you have anything you’d like to share about the newsletter or management. Just email me or reply to this newsletter if you get it in your inbox.
Have a great weekend, and good luck in the coming week with your research computing team,
Jonathan
(Cover Photo in twitter card by Nick Fewings on Unsplash)
About This Newsletter
Research computing - the intertwined streams of software development, systems, data management and analysis - is much more than technology. It’s teams, it’s communities, it’s product management - it’s people. It’s also one of the most important ways we can be supporting science, scholarship, and R&D today.
So research computing teams are too important to research to be managed poorly. But no one teaches us how to be effective managers and leaders in academia. We have an advantage, though - working in research collaborations have taught us the advanced management skills, but not the basics.
This newsletter focusses on providing new and experienced research computing and data managers the tools they need to be good managers without the stress, and to help their teams achieve great results and grow their careers.
Jobs Leading Research Computing Teams
This week’s new-listing highlights are below; the full listing of 188 jobs is, as ever, available on the job board.
Manager, Research Infrastructure Services - Colorado School of Mines, Golden CO USA
Reporting to the Sr. Director of Infrastructure, the Manager of Research Infrastructure Services is a key member of the Information and Technology Solutions (ITS) Department. The Manager of Research Infrastructure Services is responsible for driving a cloud-first strategy, leading the team that designs, implements, and monitors Mines’ research infrastructure and services, to include, managing, planning, and developing operations of research-related systems. The Manager of Research Infrastructure Services leads the support and development of HPC (High Performance Computing) data centers for on-prem and cloud solutions, new cloud implementations, research platforms, and data storage and transfer services. Key accountabilities will include the availability, scalability, security, and agility of our research support technologies at Mines along with customer facing and back-end systems.
Project Manager - Industry - Alberta Machine Intelligence Institute, Edmonton AB CA
Reporting to a Product Owner, the Project Manager serves as a key member of a dedicated product stream and an active participant in facilitating excellent client relationships and industry partnerships within the organization. In this role, the Project Manager will bridge the gap between business and technical expertise while focusing on planning and executing client engagements. The Project Manager is responsible for clear communication with the team and stakeholders on the progress and performance of projects. Capitalizing on the knowledge and expertise of cross-functional teams at Amii, the Project Manager will empower clients with the opportunities and knowledge to build strong connections into the AI talent community.
PhD Research Software Developer/Senior Scientist, MAGPIE group - Simon Fraser University, Burnaby BC CA
The MAGPIE group (www.sfu.ca/magpie) at SFU is seeking a senior scientist/research software developer to lead and support the development and implementation of mathematical methods in infection and evolution. We anticipate that the candidate will contribute to the daily activities of the research group, including supervising trainees, pursuing research and software development aims, presenting work at meetings and workshops and participating in collaborative projects. We would facilitate regular interaction with teams of research software developers at SFU and beyond.
Senior Research Software Engineer - University of Exeter, Exeter, UK
We are seeking a talented and experienced software developer and manager to join the University of Exeter Research Software Engineering (RSE) team. As a Senior RSE, you will spolut your time between collaborating with world-leading researchers on a wide-range of cutting-edge research projects, and providing technical leadership and mentoring for other members of the team to drive skills development and embed best pracise. Working closely with the Head of Research Software Engineering, you will play an important role in helping shape the future direction of this dynamic and growing team, supporting personal development and ensuring positive working cultures.
Research Informatics Project Manager, Early Detection of Neurodegenerative Diseases - Alzheimer’s Research UK, Cambridge UK
The Research Informatics Project Manager will work in the Research Informatics team to lead elements of the Project Management team alongside another Research Informatics Project Manager, to support the EDoN initiative and its partners to achieve their goals. You will report to the Head of Research Informatics and, although there will be no immediate line management responsibilities, it is expected you will manage between one and two Research Informatics Officers in future.
Software Product Manager, ISIS Neutron and Muon Source - UKRI, Didcot UK
As a Product Manager within our ISIS facility, you support technical leads to deliver improvements to existing software and deliver new systems. We are looking for someone with experience working on software projects or projects with a significant IT focus. We have recently expanded our research software engineering team, and the successful product manager will help drive the improvements and new projects in this team’s development pipeline. This includes the development of software for our new ISIS Endeavour instruments, which will launch by the end of the decade. Our software development teams make use of modern agile techniques and tools. Depending on the requirements of the application they may run on computers ranging from desktops to mobiles and tablets. In addition to software projects, you may also sometimes be required to assist in the management of other more IT infrastructure-based projects.
Manager Insights and Statistics - Transport for NSW, Parramatta AU
Right now we have a permanent opportunity for a Manager Insights and Statistics within the Data Advice Research and Testing (DART) team in TSSEM.
Reporting to the Senior Manager Insights and Advanced Analytics, this role will be leading the development of complex qualitative and quantitative analysis as well as various reporting to support the delivery of the branch initiatives, including sourcing and managing data, presenting findings and leading and developing design strategies. The role acts as a subject matter expert when designing and recommending appropriate analytical approaches and methodologies addressing key issues within the business.
Director, Operations, Data Science Core - University of Utah, Salt Lake City UT USA
The Department of Internal Medicine ( DOIM )/Division of Epidemiology is seeking a Data Science Director to direct and oversee all operations within the Department’s newly launched Data Science Core ( DSC ). The Director identifies and fully understands key requirements for data processing and implements best practice tools, processes, and procedures to achieve exceptional service and collaboration aligned with research goals. This position supervises the development and implementation of policies, procedures, and auditing functions of all DSC services. For qualified candidates, the directorship is potentially associated with a faculty position.
Director, Health Data Management - Vector Institute, Toronto ON CA
The Vector Institute is seeking a Director, Health Data and Partnerships to join our growing team in Toronto as we continue the work of further establishing Canada as a centre of expertise for artificial intelligence (AI) and machine learning (ML) in the world. Reporting to the Chief Data Officer & Vice President, Health, the Director, Health Data and Partnerships will lead the implementation of Vector’s data strategy and its data partnerships with public and private sector stakeholders, with the goal of establishing Vector as a centre of excellence for data access for AI/ML research. The incumbent will drive modernization of data governance and infrastructure, prioritizing strategic stakeholder engagement, and collaborating closely with Vector’s Health Research Operations portfolio on data policy and process implementation, as well as execution of data-related agreements.
Research Director - Borealis AI, Montreal QC or Toronto or Waterloo ON or Vancouver BC CA
Borealis AI is looking for a Research Director. The role includes directing a team of machine learning researchers and managing other research team leads with the goal of delivering AI-based products for the financial services industry. In addition, you will provide leadership and deliver on initiatives that improve the effectiveness of the entire research organization.
Development Manager – Applied AI - Thompson Reuters, Toronto ON CA
Are you intrigued by the opportunity to drive the next generation of Legal innovation? If so, we are looking for a Development Manager who will be working on all aspects of software solutions being developed by the Legal Tech Innovation team. The successful candidate will collaborate with team members, other TR development teams, and TR Labs within a dynamic and innovative rapid development environment using the latest Cloud PaaS technologies. The successful applicant will be working with agile teams and stakeholders located locally and in remote locations.
Director, Cyber-Infrastructure Researcher Support - University of Texas at Dallas, Dallas TX USA
The Office of Information Technology (OIT) provides all central IT services to the institution. A critical success factor is having a technically advanced team that provides cyber-infrastructure services in support of our computational researchers, and their academic and industry partners. The Director, Cyber-Infrastructure Researcher Support, is a key member of the OIT senior leadership team. This position is responsible for reaching out and helping campus researchers: Determine how cyber-infrastructure services can best help them advance their research; Be adequately trained on using cyber-infrastructure services; Get maximum value out of the local High Performance Computing (HPC) cluster farm.
Quantum Readiness Delivery Lead - National Quantum Computing Centre, Didcot UK
The National Quantum Computing Centre (NQCC) is a new research institution, funded through UK Research and Innovation, which is dedicated to accelerating the development of quantum computing by addressing the challenges of scaling up the technology. The Centre will work with businesses, government and the research community to deliver quantum computing capabilities for the UK and support the growth of the emerging industry. The NQCC will give assured and direct access to early prototype machines for developers and promote the formation of a strong UK-based quantum computing supply chain.
Lead Bioinformatician - UK Health Security Agency, London UK
At the United Kingdom Health Security Agency (UKHSA) our mission is to provide health security for the nation by protecting from infectious disease and external hazards. This role will be responsible for leadership of the bioinformatics function within the HCAI, Fungal, AMR, AMU & Sepsis Division. The postholder will have responsibility for setting the strategic direction of the function within the wider Division as a member of the Leadership and Management Group.
Principal Research Product Manager - Privacy Preserving AI - Microsoft, Richmond WA USA
We are looking for a Researcher Product Manager to shape our long-term research strategy and enable product impact. You will work closely with research and engineering teams across Microsoft to identify research opportunities, drive organizational alignment, and communication our cutting-edge research agenda both internally and externally. You will play a critical role in shaping our cross-disciplinary research strategy spanning several critical technology areas such as privacy preserving machine learning, and NLP.
Director - AI Tech Ops - Novo Nordisk, Lexington MA USA
Recognized as a key leader in Novo Nordisk’s Data Science efforts, the Director of AI Tech Ops will drive activities around the comprehensive integration of AI and Data Science activities within the larger technical ecosystem at Novo Nordisk. These responsibilities will include ownership of MLOps activities, interaction and coordination with IT on behalf of the AI and Analytics Organization, and ensuring that the appropriate technology stack exists and is organized for the delivery of successful solutions. The individual will be a recognized expert in AI and ML systems, tools, approaches and technologies for a variety of applications, including cloud architectures, cloud engineering, software engineering, and AGILE development practices.
Manager, qLab Operations, Quantum City - University of Calgary, Calgary AB CA
The University of Calgary (UCalgary) is creating a vibrant quantum economic and scientific hub based in Calgary - Quantum City - leveraging the leading quantum science and technology programs of Campus Alberta institutions and a burgeoning UCalgary innovation ecosystem. Quantum City outputs will include specialized infrastructure, qLab, to support university research and company agendas. This is a key leadership position within the qLab.
Manager, Information System Security, HPC - University of Toronto, Toronto ON CA
The Manager, Information Systems Security provides strategic leadership and management for SciNet’s Information Security Program which includes the security of data centres, network perimeter, and enterprise systems, including on-prem and cloud systems, hardware and software architecture, as well as managing risk and privacy assessments, incident response and investigation, and outreach and awareness. The Manager provides strategic and tactical planning, evaluation, design, development, implementation, and overall management and support of SciNet’s Information Security Program, with the goal to protect and improve cybersecurity infrastructure, posture, and culture to minimize risk of compromise, and in managing security controls required, in the support of teaching, learning and research, and services to staff, faculty and students.