Notes on one of the most-recommended old-school management books for tech startups.
The output of a manager is the output of the organizational units under his supervision or influence.
What is that output? How is it measured?
Effectiveness
Managerial LeverageDo high leverage activities, those where your investment of time contributes most to increase output. Forget the rest. You also can have negative leverage, when you de-motivate your employees by badmouthing the company, when you do not provide them with clear direction, or when you do not make a decision. No green light is a red light, and can freeze the organization. Meddling has negative leverage, when you involve yourself too deeply in your subordinate's work, taking their ownership away (even if you are better at it than they are).
Take control of your time
The most important resource you allocate for leverage is your own time. Capital and manpower can be obtained. More personal time can not. Your calendar is your most important management tool. Actively manage it. Fill holes between time critical events with non-time critical but necessary activities. Do not allow others to throw in orders for your time. This is mindless passivity, defense, not offense. Say NO at the outset to work beyond your capacity to handle. Allow slack, a bit of looseness in your scheduling (like 3/4 h meetings instead of 1h meetings end to end). Maintain a backlog of projects that you can take on, and at any time only focus on a few active ones -- best one. Strive towards regularity, routine -- meetings always on the same day etc. (This only works if others do so too).
Delegation
Avoid the charade of insincere delegation, which has massive negative leverage. Delegation without follow-through is abdication. You can never wash your hands of a task. After you delegate it, you are still responsible for accomplishment, and monitoring is the only practical way to ensure a result. Adapt your monitoring to the amount of experience the subordinate has with the task, less if they are very task-mature, more if they are new to it. Only go into details randomly, and just enough to ensure the subordinate is moving ahead satisfactorily. A manager should have six to eight subordinates, so he can allocate about half a day to each.
Batching
To make most of your time, group similar tasks, such as email reading, trade-news reading. Meetings and production time. Ask people to batch questions into scheduled times, like one-on-ones, instead of interrupting whenever they want. Make things regular, that were irregular.
Meetings
There are two kinds of meetings: process-oriented meetings, where information is shared, that are scheduled regularly, and mission-oriented meetings, to solve a specific problem, often to produce a decision, and that are called ad-hoc.Timeliness is critical, It is criminal not to enforce it and allow latecomers to waste everyone's time. Being prepared is critical: if someone has not read up, stop the meeting and reschedule when they have. It is wasting everyone's costly time. Put discussion points that derail but need followup into a tickler file/list for future dates.
The One-on-one
Between manager, and direct report. Should at least last an hour, so the subordinate does not confine himself to simple things that can be handled quickly. It should be the subordinates meeting, with the agenda and tone set by him.
The subordinate should prepare an outline, which is important because it forces him to think through the points he plans to rise. He also should prepare supplementary material for that outline, and walk the supervisor through all the material.
Covered should be key indicators, especially those that signal trouble. Anything important that has happened since the last meeting, hiring problems, people problems, future plans, and very important - potential problems, issues that preoccupy and nag the subordinate. These are often obscure and take time to surface.
The supervisor should facilitate the subordinates expression of what is going on and bothering him. Do not talk with your subordinates about your problems, make them talk about theirs. Ask one more question, until you both feel you have gotten to the bottom of the problem. Be wary of a heart-to-heart issue brought up at last minute, though, when you can only with difficulty give it justice.
Both should take notes on the agenda, Take notes in outline form, which helps you categorize and sort the information. Follow up by checking last times notes in the next meeting. Create a "hold" file with important but non-urgent items for the next or future meetings. Exchange notes after the meeting, to make sure you have agreement.
One one ones foster development of a common base of information, handling of issues, enable effective delegation, and are essential if the supervisor is to make good decisions.
The Staff Meeting
A meeting of the supervisor with all his subordinates, to enable peer interaction. You get a much better understanding of an issue by listening to people with opposing views discussing it, instead of just listening to one person. Discussion is for anything affecting more than two persons present (otherwise, they can take it offline). If the meeting degenerates to a dialogue, suggest the two continue their discussion later.
An agenda should be prepared, but there also should be time at the end for open discussion of unscheduled items. Role of the supervisor is that of observer, expediter, questioner, decision-maker. Not lecturer.
The Operations review
An agenda should be prepared, but there also should be time at the end for open discussion of unscheduled items. Role of the supervisor is that of observer, expediter, questioner, decision-maker. Not lecturer.
The Operations review
Formal presentation to other managers, peers, and parts of the company. The organizing manager is the direct supervisor of the presenters, responsible for timekeeping, disseminating materials, organizing. The reviewing manager it a higher-up manager, who should ask questions, make comments, set the spirit of the meeting, and encourage participation. The presenters should use visual aids. The audience should participate, make comments and ask questions, if something is not clear. Speak up if you disagree with an approach that is recommended. Go on record to correct factual errors.
The Mission-oriented meeting
The chairman, the person who has most at stake has to drive this. He calls it, and he should prepare it. Imagine you are the chairman. Define a clear objective -- what needs to be resolved, what needs to be decided. (If you do not know what you want, you will not get it). Is the meeting necessary, defensible, justified? Only call it if it is, and call off it if can be resolved otherwise. The cost of meetings is about $100 per participant and hour. Identify who should attend and get them to come, or send someone else who is empowered to speak for him. Try and limit the meeting to six, maximally eight people, if you want a decision to be made. Maintain discipline. Call out latecomers. Make sure the needed equipment is present and in working order. Send out an agenda that clearly states the time, location, purpose, and everyone's role (presenter with duration, or not).
Once the meeting is over, nail down what happened, and send around minutes summarizing the discussion, decisions made, actions to be taken, by when and by whom. Send this out soon after the meeting, when it still is fresh on peoples mind.
Decisions
Make or facilitate the making of good decisions
Decision-making produces a better decision if we are clear about: what decision needs to be made, by when, who will decide and who will be heard, who will ratify or veto it, and who then needs to be informed of it. Attendees to the decision meetings should have consulted their staff prior and gathered all relevant knowledge and views on the subject.
There is often a divergence between power of position, and power of knowledge. Junior people are closer to the issues and technology and understand them better. Make sure they are heard. Everybody should voice opinions as equals, ignoring status. The process for decision making should be 1. Free discussion, looking at the problem from all perspectives, 2. A clear decision, even if not everyone can agree it is the right one, 3. Full support by all, also those that did not agree to implement it. If it turns out it was wrong, repeat. Take pains to frame the decision with utter clarity. Do not fudge to try and keep everyone happy by formulating it ambiguously.
Free discussion is the most difficult to achieve. People should express their view forcefully, but tend to hang back until they see a view winning, and then pile in to support it, to avoid being associated with a losing position. People are full of ambition, fear, insecurity. People are afraid to stick their necks out. People are afraid of sounding dumb, and do not ask, when they do not understand. They are afraid to be vetoed or overruled and thus to lose face in front of their peers. (I might add, they are also afraid of coming across as spoilsports, of being accused of undermining a can-do spirit). If sensitivities of two interest groups are involved, give both sides roughly equal representation in meetings to foster an even-handed decision.
Do not push for a decision prematurely, make sure you have heard and considered the real issues, rather than the superficial comments that often dominate the early stages of a meeting. Do not use authority to influence the exchange of views in any direction.
When all views and arguments have been brought out, when everything is heard, it is time to push for consensus. A decision has to be taken, even if no consensus can be found. A senior person with position authority must make a clear decision at that point.
To ratify a decision, listen to the alternatives and background, and reasons for the choice, and ask questions to probe the depth of information and thinking. If the final outcome is dramatically different from what people expect, make the announcement, adjourn so people have a chance to recover, then reconvene and solicit views, to help people accept and live with this outcome.
Approvals
For projects that require funding or capital, ask the subordinate to think through the entire matter carefully, before representing a request for approval. To test their thinking and decision making process, ask them specific questions during a review meeting. If they answer convincingly, approve.
Visit plants, observe what is going. Visit customers. Read the trade press. Information gathering is the basis for all other managerial work. Gathering information is the basis for sharing and for making correct decisions. Customer complaints are a key input. Identify and defuse ticking time bombs before they go off.
- Detect and fix errors in a process at the earliest, lowest value stage possible. Reject before investing further value. (It is much more costly later, requiring more rework, discarding more work already done). Review rough drafts of reports you have delegated.
- Use performance indicators to measure output (not: activity indicators, which measure only busy-ness, but no results, pair opposing ones to combat bias). Use credible leading indicators to look into the black box and get a feel for what the future may look like, and act on them. Use trend indicators, like a Stagger chart, that plots the development of forecasts over time. Collect these indicators ongoing and systematically.
- Reports and Plans. The most useful information comes from brief, direct discussion. But written reports are necessary, less for the reader, for the writer: they force him to think things through, force self-discipline on the thinking. Likewise, the resulting plan is not important, but the process of writing the plan is. A capital authorization process is a must, because of the soul-searching analysis that it forces to support the spending request.
Leadership
Lead and Train your employees.Communicate your objectives, priorities, preferences. Transmitting objectives and preferences is key to successful delegation. Act as a role model -- nothing leads as well as example. Values and behavioral norms are not transmitted by talk or memo or wall poster, but by visible action. Share and disseminate information. Most of the time you do not issue direct commands, you "nudge".
When a person is not doing his job, this can have two reasons: he either can't do it, or won't do it. (Aside: either train him if he would but can't, find another job for him, or, failing that, let him go. And you need to improve your hiring process). Fear and punishment may work to drive galley slaves, but won't work on computer engineers.
You cannot motivate someone, real motivation comes from within. All other forms, like safety through compensation, belonging and social affiliation, esteem and recognition are self-limiting. The only source of motivation that has no limit is self-actualization. Being the best you can be. The need for achievement is boundless. At that level, money is not the objective, but it is a way to keep score, a measure of achievement. You cannot stay in this mode, if you are always worried about failure (Note: this from a guy who was an outstanding achiever and who wrote a book called Only the Paranoid Survive. Does not parse. I think what he means is that you must be willing to take risks, or you will never stretch to your limits.) For achievers, the most important is feedback on his performance. They run to beat other people's bests, or in the end, their bests. They do not need incentives, they need an arena. A good manager must be a coach: take no personal credit for the success of the team. With achievers, be tough on the team, be critical, expect more. To get the best performance out of them they can achieve. And it helps if he used to be a competitor himself, so he knows and understands.
Manage people based on their task-relevant maturity. An experienced person can be a newbie with a new task. If they are, be task oriented, offer detailed, precise instructions, what needs doing when and how. As they mature, move more to encouragement, emotional support, listening and mutual reasoning. As they become very mature, minimize your involvement, just make sure the objectives are agreed upon. But always monitor progress to avoid surprises. It's the difference between delegating and abdicating. In all cases, you are responsible to transmit common values, which are a must for effective delegation. And for pragmatic reasons, try to raise the task maturity of your employees as quickly as possible.
Should you make friends with employees? This is a personal decision. But you will need to give tough feedback, give orders to or maybe even lay off an employee; this is very hard to do to a friend. (At the same time, if you deeply care about people, you will give that feedback, and will be mindful and respectful when you have to take painful steps).
Performance Review
These Reviews are the most important form of task-relevant feedback you can give. They should be done in any size of organization. The desired output is to improve the subordinates performance. You must first judge fairly a fellow worker, and then deliver this judgment to him, face to face. Review is dedicated to first assess the subordinate's skill level, determine missing skills, and find ways to remedy the lack; second to intensify the subordinates motivation.
First, review material such as progress reviews, performance vs quarterly objectives, one-on-one meeting notes (take good notes, so you have them). Sit down with a blank sheet. Dump everything on it, do not edit in your head. Get it all down. Major, minor. Don't matter. Once you are done, put the supporting documentation away.
Now, organize the things in groups, for positives and negatives; identify common themes, shared strengths, weaknesses, and turn them into messages for the review. Support them with specific examples. Think about what they will be able to remember, and cull your list to the most important ones. Just one or two major ones.
Even if you did regular one-one-ones throughout the year, you may be surprised of your findings. And even if they are uncomfortable, they need to be delivered. If you discover a surprise, swallow hard, and deliver it.
Performance Review: Assessing and Judging
For useful review, define the goals up front. Output measures (stuff that could be plotted on charts, like bookings, designs delivered, increased production yield etc), but also internal measures that will define output for future periods: are we achieving our present results by sacrificing the future, disgruntling employees etc, or are we doing it in a way our business can handle its tasks in the future?We must weigh long-term vs short-term oriented performance. How much will the future oriented work pay back over time? How much is it worth today? Look at the time offset between activity and resulting output -- often results this year are really the fruit of a previous year's work. Most jobs involve work that is not producing output in the period of review, still, it has to be assessed.
You need to be as objective as possible, but in the end you must judge performance, which naturally has to be subjective. It's not just recording what is measurable in plain sight. Judge his personal performance, as well as the overall performance in his group. In the end, what counts is the overall performance of the group and the manager must add value in some way. Hi performance rating cannot be higher than that of his organization. What counts is results, not good form. At all times, assess performance, not potential. It's a trap.
Level: be frank. The credibility of the entire system depends on it.
Listen: make sure they understand the message. The goal is not to even document you delivered it. It is not delivered, unless you have been heard and understood. Watch them, look for signs with all your senses. Listen with all your might to make sure they get the message, and do not stop delivering until they have.
Leave yourself out: this is not about you. Forget your anxieties, insecurities, guilt. At issue are the subordinates problems, not the supervisors.
Limit: The goal is not to clean your system out. They may have a finite capacity to accept and process facts. Your goal is to deliver the most important ones, to improve their performance.
Best deliver a written review to your subordinate some time before the face-to-face discussion. He can digest it and react or overreact to the messages, and by the time you meet will be much more prepared, emotionally and rationally.
Reviewing poor performers: poor performers have a strong tendency to ignore their problem, passively ignoring or actively denying it. If they cannot, because of overwhelming supporting facts (which you need to collect), they might justify it, by blaming others. Things tend to get stuck at the blame-others stage. They must take the step to assume responsibility, to accept it is their problem. This is fateful, as it means work. Once responsibility has been taken, finding a solution may be relatively easy. Assuming responsibility is an emotional step, which is much harder than the intellectual one of problem-solving. You as the supervisor must move them through all the stages of this process. You cannot try to go find a solution with him, if he still is stuck at blaming others or denying the issue. Knowing where you are will help you to move through together.
It is sufficient, if they commit to do it, even if they do not agree with your view. They do not need to side with you. They just need to commit themselves to pursue a course of action. On the job, we are after a person's performance, not our psychological comfort. Say "This is what I as your boss, am instructing you to do. I understand you do not see it my way. You may be right, or I may be right. But I am not only empowered, but also required to give you instructions, and this is what I want you to do..."
Reviewing the ace: concentrating on the stars is a high-leverage activity. If they get better, the impact on group output is great. No matter how stellar a persons performance, there is always room for improvment (?)
Review template: Name and role, review period, description of assigment (full paragraphs); accomplishments during the period (full paragraphs); evaluation with areas of strength and for improvment (full paragraphs, messages with supporting examples); outlook and recommendations (full paragraphs); overall performance score. Signed by manager, employee, HR with date
Read all the evaluations written by your own reports, and a random sampling from below that level in the hierarchy. Send them back with comments, and high visibility of this activity, if you want to impress the importance of this process on your organization.
Hiring and Interviewing
Interviews serve to select a good performer, educate him about your company, identify if there is a match, and sell him on the job. Reference checks do not exempt you from getting as much as possible out of the interview. Try to have a bit of a longer conversation with the reference giver, to build a relationship. Often you learn the most valuable information towards the end.
The applicant should do 80% of the talking, and it is up to you to steer what about. If they drone on, interrupt or stop them to conserve the valuable time you have with them. Apologize if you like and say, "I'd like to change the subject to ...". Try to talk about areas where both of you have expertise. Try to assess their technical expertise, their skill level ("describe some projects"), as well as how he performed in previous jobs using what he knows ("past achievements, past failures". Try to understand why there are discrepancies ("what did you learn from failures, what are problems in the current position"). Try to understand their operational values ("why do you want to change"). Don't worry to be blunt and direct. You can ask "How good are you technically?", and learn something from the response. Ask them how they would handle a hypothetical situation. Ask him what he would like to know, to understand how he thinks and how well prepared he is.
Show yourself and your environment how they really are. No point to hire someone, and then he hates it here.
Retaining Quitters
When a key employee quits, deal with it right away. Often this is about recognition, he feels unappreciated and unimportant to you. Don't confirm these feelings. Drop what you are doing. Sit him down and talk with him, why he is quitting. Let him talk, do not argue -- he has rehearsed his reasons many times. Let him talk, then ask him more questions. Because after the prepared points have delivered, the real issues will surface. You have to convince him by what you do, that he is important. Do not try to change his mind at this point. Then go to your supervisor for help and advice. Make him participate in the solution to your problem. If you cannot keep him with you, try if you can save him at least for the company. If all managers take this position, they will all win in the long run.
They will already have accepted, so you have to make them quit again. Say he really has made two commitments -- one to the new employer whom he barely knows, and one to you, his present employer.
Keeping them is even more important for the overall company than just his work. Other good people will respect him, and see this as an example to follow.
Compensation and Promotions
Once the absolute raise in compensation is not important any more, but the relative raise, you are out of the area of fulfilling basic security needs, and into the area, where money really just is a measure of success, for self-actualization. You want to allocate money, like promotions, as task relevant feedback. It always should be tied to performance. If you practice a pure tenure based policy, your message is that performance does not matter. A performance based system is obviously more work, as it requires a competitive, comparative evaluation of employees.
No action communicates values more clearly and loudly to the organization than a promotion. By elevating someone, we effectively create role models for the organization. They must be based on performance, to keep the idea of performance highlighted and alive. The Peter Principle unfortunately cannot be avoided, as you only learn if someone is up to the new task after they have been promoted, and not promoting them to keep them at their current job where they perform great, will make them leave unless they are happy that way. The one solution is to agree with them to go back to their old job -- obviously very difficult to do, but if done, often works wonderfully for both sides.
Training is the Boss's job
It is the highest leverage activity you can perform. To be effective it needs to be reliable and regular, not an ad-hoc patching up of issues, a process, not an event. The managers themselves should do it, as they are believable role models. 2-4% of the work time should be spent on training (about a week per year). You can train managers skills such as strategic planning, communication style.
The first task is to train new employees on the ways and values of the company. The second one is training new skills to all. Ask people what they think what training they need, then make a list of possible trainings to give. Take inventory of the available managers and materials, and make a priority list to deliver. Start small - one short course on one subject. Define a course schedule, with deadlines, prepare just the outline and the first session and go. (To force yourself not to get lost in spending all your time to prepare this). To limit damage, first do a trial run to knowledgeable employees, who can help you refine it. After delivering it, ask for anonymous ratings, in score but also open ended comments. Be aware you'll never satisfy everyone -- what is too detailed for one person is to superficial for another. The person who will learn most from the course, of course, will be - you.
Find the most cost-effective way to deploy your resources.
Analyze processes for work simplification: chart out every single step. Then look for simplifications, cutting 30-50% of the steps. ("Complicated cases make bad laws"). If there is a set deadline, work backwards from there, identify the rate limiting steps, and arrange the rest around it.Plan
1. Establish projected need -- what will the changing environment demand of you, your business, your organization? (Don't consider at this stage what practical steps to take to get there, this will just confuse the issue). How do you know?
2. Establish your present status -- what will you produce if you do not do anything different than what you do now?
3. Reconcile -- what will you need to do to to close the gap and produce what the environment will demand? What can you do? Then decide what actions to take. This set of actions is essentially your strategy, and the details to implement them are the tactics. The strategy on one managerial level is often the tactics on the next higher level.
Today's gap represents a failure of recognizing and planning in the past. What do you have to do today to solve or better avoid tomorrows problems? In long term planning, look at the next five years. But what you really influence, is the next year. You will only implement one portion of the overall plan, before you review it. Be careful not to plan too frequently, so you have time to judge the impact of the decisions made and see if you are on the right track or not. (Annual seems reasonable, with monthly or quarterly progress review).
Involve the operating management in planning. Planners must be the very people that have to implement the plan.
Saying yes to a project means saying no to something else. Keep the number of projects small. If you focus on everything, you focus on nothing. Each objective must be specific and have a due date, so there is no room for ambiguity.
Organize
"All large organizations with a common business purpose end up in hybrid organizational form", that is, with functional company wide units that enable critical mass of know-how (sales, R&D, admin), and localized or product-specialized business units that enable the business to respond to the demands and needs of the market environment. Thus, many people will need to dual report -- to the local or business manager, and to the functional or corporate manager. There also can be peer group supervision. For this to work, you must have a strong and positive corporate culture.
The most important characteristic of culture is that the interest of the group to which an individual belongs takes precedence over the interest of the individual himself. For this to happen, all need to share a common set of values, a common set of objectives, a common set of methods, generated by common, shared experience. This makes it easier to work on tasks with high complexity, ambiguity and uncertainty (which are often the higher paid leadership roles). New employees cannot have this culture, need to get familiar with it, so will have a hard time coming into a leadership position from the outside in a company with strong company culture.
