How to conduct a performance review
This is a tricky topic. To conduct a performance review you need to know a) what is the performance you are measuring and b) what will you change as a result of that measurement, and c) how will that affect your employee. Is your review process about coaching? Or is it about salary? Is it about measurement? Perhaps ironically, those are all topics outside the scope of today’s update, which is just about how to conduct a performance review.
Performance reviews are stressful. They are threatening. It shouldn’t be that way, but most people rely on their job to make a living, so anything that places that at risk will be viewed with a gimlet eye. Plus, common wisdom suggests you should “let go” of poor performers, which explicitly places a severe consequence in front of people who, on average, need a job.
Taking that into account, good performance management is about measurement. It is about fairness. The conclusions drawn in performance reviews should be extremely obvious. If someone doesn’t know they’re doing a bad job, and they don’t find out until a performance review, your whole performance measurement process has failed. This is because you have given that person no opportunity to self-correct.
What you want is a process of constant feedback, realignment, and encouragement throughout the year. This allows people to redirect in a low energy way without feeling threatened. Through this feedback they get an understanding of what their weaknesses and strengths are. This makes performance reviews easier because both parties are able to agree on most of the main points.
People also grow faster with more frequent input and encouragement (assuming it is constructive and they are able to process it without feeling threatened), creating a positive feedback loop. The trouble is that many businesses do not really understand how to promote and measure the behaviours they are looking for, so ideal performance from a manager’s perspective is “just do your job please and don’t make my life harder”, and from the employee’s it’s “what the fuck do they want me to do now?“.
There is a better way.
What the ideal performance review looks like
In a perfect world, manager and employee should walk into a performance review and agree on everything that gets discussed. How could it be otherwise? If someone walks into a performance review and disagrees majorly on several items – either they’re badly wrong, or you are. You’ve either failed to measure accurately what you are supposed to be measuring, or failed to communicate it.
The only way to reach agreement here is to have many small discussions and provide regular feedback throughout the year. You will need to:
- set up performance measurement standards
- inform people of what the standards are at least 6 months in advance of any review process
- find a way to (kindly) let them know where they are at relative to those performance standards (again, well in advance of any actual review)
- understand where each person is at and where they want to head with their career. Many jobs require more than 1 skill/domain and people may want to excel in one, or progress across multiple domains into another job entirely. You need to learn this and structure feedback and guidance accordingly.
I assume that you have a criteria on which to conduct reviews and ratings or milestones that let you discriminate between different levels of performance. You should be able to discriminate between skill levels along the lines of “yes you can do a backflip, but you need to be able to do a backflip and land on both feet simultaneously, without wobbling on landing, in order to reach the next level“. If you can quantify your requirements at this level of specificity, and give people tangible feedback on how to get there, it will make your review SO MUCH EASIER.
Plus, people will respect you more because you can provide them tangible advice and feedback on what to improve instead of being the pointy-haired wanker from Dilbert.
How to introduce a performance review process
Once you have your criteria, introduce the concept of performance reviews to your team in a very nonthreatening way, usually 6-9 months in advance. This acclimatises people to the process and lets you refine it before judging people for real. Explain to your team why the reviews matter and how they help the business. It is very important to phrase this discussion in a constructive way.
You must commit to delivering what you promise. If there’s no intention to use performance reviews as an input into salary discussions, do not say that there is. People want to work at a business that is constructive, helps them improve and gives them tangible feedback. Unsurprisingly, people don’t like to be brutally demoralised by a heavy-handed power trip or treated like a disposable minion. If you invest in people and support them – and this includes giving kind and constructive feedback about what is not working well and what they could do better – you will be pleased with the results.
With your team, hold a discussion on the process you intend to implement. Walk them through some slides, which should be very detailed and contain step by step instructions on how the review and measurement process will work. Clarify how the whole process will work and how often the reviews will happen. Detail builds security and psychological safety because people know exactly what to expect. You will probably field some questions around “how does this impact salary” or “why are we measuring this way”. It goes without saying you should have answers, but be flexible and incorporate feedback. The goal is a good process, not “I invented this thing and therefore that is what we are going to use”.
Conduct a practice performance review
Once you have established the criteria, conduct a practice performance review (I will talk about how to actually do this a bit later) with no consequences. Introduce a practice review 6-9 months before the first real review, giving people time to get used to it, and time for you to measure. Treat the practice review seriously and put the same amount of work in as you would if it was a real review and you knew that you had to make a decision on salary immediately after.
I explain to people that the review is me in my role as manager, forming the business’ view on them as an employee. For this reason, it is important they read the review carefully and clarify anything they disagree with or are uncertain about. It is critically important that you approach the reviews in a non-adversarial way. You are not “judging” them. You (as manager) are facilitating the forming of a view which will be used as an input into promotion and salary and what have you.
Remember; you are representing the business and must provide your judgement on how the employee measures against the independent criteria, but you are not the supreme arbiter of people and all that is good in the world. The review requires judgement but it is not actually *about* your judgement. It is about data and measurement. You must be well prepared with examples for each measurement that you provide.
Once the initial presentation and “test review” (with no consequences) is done, tell people to think about it for a few days. Follow up 2-3 days later to see if there was any further feedback on the process itself. If not, proceed with implementing your process (including any adjustments that may have come up during your presentation). Congrats – you now have a performance review process. The next bit is a fair bit harder.
The way I conduct performance reviews is like this:
Before the review:
- Set a mutually convenient time for a review.
- 2-3 days before the review, I provide a “draft”. This is “final” on my part (nothing else to add) but give the person time to read over it and comment. The review usually has 3-5 criteria, and ratings from 1 – 5 in each criteria. Provide detailed comments on why that rating was chosen, with examples where relevant, and actionable ways to improve. I keep the language non-emotive and write in the third person. “John is satisfactory on criteria X. This work was good, on some other work he struggled with [thing]. John will improve by [action].” Writing in first person is also fine – whatever works for you.
- I make it clear that the draft is exactly that. We will talk through the rating and their comments in the meeting and (ideally) agree on the final version. There are really only three outcomes to this discussion:
- A) I am accurate and fair in all of my comments and they agree with everything.
- B) Some of my comments are inaccurate, wrong, poorly phrased or otherwise misunderstood. We discuss it and amend the comment in light of the ensuing discussion. It is perfectly okay to change or remove a criticism or score after reexamining the situation. However, if there are badly wrong scores (say you rate someone 2/5 and then change it to 4/5) this is a warning sign that you need to rethink your process. Excluding mistakes, which we all make, you should never repeatedly find that someone goes from “inadequate” to “adequate” after a discussion. If you do, it is a sign that there are problems with your process. Either you are too busy, not measuring correctly, not collecting enough evidence, allowing yourself to be too easily swayed by debate, or what have you.
- C) The employee disagrees with comments or scores, and I decide to keep the ratings and explain why. I make sure to record their comments on why they disagreed with the rating. This is very important for feeling like they are being taken seriously. With low ratings, it is extra-important to provide a clear path to improve. It is okay to be not of a good enough standard IF (IF!) there is an achieveable way to progress and the person trusts you to help them.
Holding the actual performance review
- Once the person has read the draft review and added comments/questions, proceed. I personally choose not to read their comments before the review meeting. This is for a few reasons:
- I want to avoid setting up a confrontational dynamic where I’m composing responses to their disagreement even before we talk.
- It equalises the power imbalance somewhat because the employee has had time to prepare, whereas I must respond on the spot. I take a very high-ownership approach to this. If I haven’t thought about my review in detail and can’t thoroughly justify a certain rating or comment, I have no business providing this sort of review.
- For new managers this may be intimidating; it is perfectly okay to prepare your responses in advance. Do what you can to reduce the power imbalance where possible.
- Once we are in the meeting, we talk through the document line by line. “I wrote here that we would like to see a little bit more initiative taken on [blah]. These are some examples [blah] and this is [what we would have liked to see you do instead]. Does that make sense? What are your thoughts?” In an ideal world, they will agree a lot of the time. Sometimes they have questions or comments. Sometimes I am wrong or have misunderstood something and we make changes. Much less frequently, I amend a rating as well. The goal is a fair and constructive discussion, as much as is possible in an environment of power imbalance.
- Following this, we discuss development – what they want to work on improving, or what we need them to improve. I ask for their ideas on how to improve, and record them at the bottom of the review.
- At the end of the meeting, I make it clear that the review is over and thank them for taking the time to sit down with me.
- I provide information on next steps. “I will email a copy of this final review to the HR team with your comments and cc you in the email as well.”
- If there’s a performance management or learning & development element indicated by the review, I set time to sit down and talk through this. Usually the onus is on the person to come up with some ideas on how to improve which we can iterate on.
- If there is a salary review upcoming I mention when that will be (I generally try to have a two week interval between them, but I’ll talk about that another time).
That’s it! All told, the review generally takes about 60-75 minutes.
Following the review and once I complete any follow-up actions around salary etc, I provide the person with a rating scale on me as a manager. I ask them to rate me, explain each rating, and suggest ways to improve. I make sure that this rating is separate from the review process. There should never be a question of the manager manipulating a person’s outcomes in response to a poor rating they received.
Thanks for reading. If you have some feedback, would like to share differences in the way that you conduct reviews, or suggest topics for future consideration, I’m all ears. Cheers.
Notes & Further Resources:
Indeed and Robert Half have some useful tips on the more high level considerations of performance reviews, like when in the year to schedule them.
Nb: There are a bunch of closely related concepts here that I didn’t have space to discuss in this article. The generally accepted practice is to separate the salary review from the performance review. Businesses should consider salary relevant to changes in role & responsibilities, peers and the market, not just performance. Additionally, larger companies often separate performance measurement from salary to avoid conflicts. Other things not addressed here but immediately relevant & worth thinking about are:
- How to get honest feedback from employees as a manager
- How to set up performance measurement criteria
- What to do with people that are failing at their job,
- How to manage differently if you are a manager as opposed to a combined manager + team lead
- Whether & how to use a peer rating of skills if you are not a technical manager