Answer the question
In order to leave comments, you need to log in
The developer is undisciplined in tracking time. What to do?
Hello!
When working as a project manager, I came across the fact that developers are often too lazy to track time on their tasks. According to my observations, according to this indicator, developers can be divided 50 by 50 - the first half tracks time regularly, the second half does it only when they are constantly reminded of it, or does it retroactively, which leads to unreliable results.
As a manager, this state of affairs does not suit me - I am interested in truthful metrics and I do not want to spend time on manual control.
To solve this problem, I considered its possible causes:
Answer the question
In order to leave comments, you need to log in
Why track time at all? Met the deadline - well done. Did not meet - debriefing. Chronically does not fit - downgrading or dismissal.
Do not take out the brains of the developer with your tracking and let him work in peace.
Doesn't work - get fired.
God forbid working in such a toxic environment.
Sit and account for every minute/hour.
Ideal information concentration camp.
You considered the possible reasons "from your own bell tower" and gave the answer to them yourself. What shows your authoritarian (directive) management style.
Try to be closer to your subordinates, talk to them, delve into their problems. Then they themselves will tell you why they do not track, and you will understand how to solve it.
You don't have any feedback right now.
Bullshit all these trackers. If you do not trust each other, then nothing will help.
The developer also did not fall to constantly suffer with these trackers. It is easier for him to write down the approximate time spent at the end of the day.
It rarely happens when you work linearly. Constantly twitching from one task to another. If this is not the case, then the tracker is not needed.
the main questions that need to be answered:
1. What do the developers themselves think about this? Did you communicate with them? What are the conclusions - what has changed at the heart of this communication?
2. Why do the developers themselves track this time? you are shifting some of your managerial troubles to them, they do not need and should not need all this. Why do they need your "assessment of project marginality"? They have enough work to do.
If you want everyone to track time - give them a clear reason why they should do it and the perceived benefit. For example, pay by the hour.
"I need you to do this" - so-so motivation. Even if you add a whip.
You don't need to work 8 hours, but
maybe it's your head that you are a bad manager if you are not confident in your coders, and aren't coders so bad and imitating violent activity?
Do you need a coder or don't-get-what-the-button-poking "I did" Hindu?
In general, I think that your approach is not correct at the root. Your own unwillingness to study in detail people when applying for a job and led to distrust at the time of work.
Therefore, I rejected this reason.
I believe that such conditions should remove these phobias, so I rejected these reasons.
so I rejected this reason too
I don't know how to get all the developers to track their time without unnecessary reminders
As a result, I realized that I don’t know how to build this process and I don’t know how to ensure that all developers track their time without unnecessary reminders
Oooooh, yes, we have a manager here with claims))) you need to accept the result, and not control the process and engage in other brainwashing, and if you don’t know how much time you need to allocate a developer for this or that task, so as not to shake with a stopwatch behind his back - it's not his problem, it's yours
If the estimate exceeds the planned one by more than 30 percent
How about removing all the nonsense from the workflow that does not contribute to the creation of the product?
But if you really want to, then why not track yourself, upon the completion of tasks. Track just the performance of the developer as a whole (how many tasks he passed in a week/month).
Developers do not objectively need this tracking, it does not say anything and does not affect anything.
you, if you are a good manager, you should probably understand that distrust and excessive control become too expensive in terms of labor costs. read why the practice of "tu-an-shih" works in China, it seems, when in a huge company there are only 3 levels of management - and each manager has 40 deputies, and why this is not in the American-European practice, where 1-2-3 deputies - this is the "norm", and as a result, the administrative apparatus swells into many layers.
with developers something similar. if you want accurate reports, you don't trust them in a sense; do not believe that they were doing something usefully meaningful - you will spend more and more time on these very reports and the mechanisms that provide them.
and be prepared for the appearance of the column in the labor costs "filled out the labor log, 1 hour per day"
moreover, you must offer this column . at least because a good project manager always allocates a separate column of work in the wbs under the heading "project management". and in your case, filling out detailed cost reports for tasks is just "management costs".
if such a column suits you, you are ready to spend your salary budget on this - then you just have to explain to people the meaning of why you need these most accurate detailed reports with a detailed breakdown by tasks.
because people will not do meaningless (at least from their point of view) work.
especially engineers.
you yourself understand what is the point, what is the practical use of the most detailed reports?
why are you not satisfied with, say, a 10% error in these values?
you need to track tasks, you need reporting (at least to understand where the budget goes), you need to remind people that reporting needs to be done, but you don’t need to go too far with it .
that is, reminding a freelancer of the need to report on expenses for yesterday is a must. alas, this seems to be a common practice, yes, this often happens. but it is not necessary to make reporting an end in itself, and one should not hope that it will work on its own and give true indicators.
and what's more - it's your job as a manager to keep an eye on it.
if all the people themselves did everything automatically and honestly - the same freelancer always filled everything truthfully, accurately and on schedule - you would become an unnecessary layer.
Our company is quite small, but what my boss did: He wrote a simple website where we write down the name of the company and the time spent + the work done. Completed the task - added to the site and the client receives an invoice for these hours. We take payment for units of 15 minutes, that is, 3 hours of work is 12 units at a certain price, in this application we enter the actual number of units.
Well, in principle, everything, here people say that this is all about managers, not coders and you need to give a reason to coders to track their time, but I was taught to do this from the very beginning and I took it for granted, as one of the duties of the job. But here all the same, you need to take everyone's word for it, if there is distrust, this method will not work. But it is the simplest and easiest, it will not be difficult to write such a site at all.
But let's say even an employee slightly increased the number of hours, entered instead of 4 hours - 5 hours, is that scary? You do not pay the developer hourly money for the work done, but on the contrary, clients pay you. They don’t understand how the matter is, and this practice is used everywhere, some have less, some have more, but basically everywhere. And so in IT it is very difficult to say exactly how long this or that task will take, because the skill level of each person is different, and you never know what kind of workspace, then no one will argue 4 hours there or 5 hours you have to pay. Well, this is nonsense.
It all depends on your company, this method works for us and is very successful.
I personally know what price is set for clients and when I do work for a certain number of units, I immediately understand that I have at least earned more for the company than they will pay me for that day and I can already safely perform other tasks without driving myself.
Everything comes down to money.
It would be nice to take it from the customer in 10 hours, and so that the proger closes the task in 1 hour, well, and says that’s it, hare me to pay. We live on these 2%. Correctly?
You don't need time. You need money! And what is most interesting, performers also need money. Piecework for tasks, not hours.
Yes, you'll end up with heavily maintained code, but that's what you wanted!
The task is evaluated by the one who will perform it, with a margin, as he feels comfortable, the manager additionally adds a percentage of the risks (based on history and experience), and reserves time and budget with the customer.
If it does not work out that the task was taken by the one who assessed, then it is especially important that the assessment be with a margin. Especially if the task is left to an inexperienced developer - he can complete it twice or three times longer.
Developers write down all the time for the task, as it suits them, in free form, but with reference to the task. It can be directly development, support, analytics, consultations - any useful activity must be paid.
They write down how and when it is convenient for them, but they are warned in advance about the closing date of the sprint / deal, so that they have time to put down the time.
Their (and not only their) income directly depends on the closed time, so if something was not recorded, it means they worked for free, the motivation to track more often. Again, transparency is needed so that developers understand that their income and firms depend on their records.
There are many tracking systems, all sorts.
But these are developers - so there are many custom, internal, even private trackers that merge records via api into a custom corporate tracker.
There are all kinds of clients for the corporate tracker, incl. web muzzle and mobile apps. We develop it ourselves, as it suits us, something takes root, something dies off - a living ecosystem.
The corporate tracker, in turn, can be synchronized with redmines. Some clients are able to upload both to the tracker and to redmines, tk. store data in a universal format.
Estimating by comfort time allows developers to meet the estimate more often than not. Therefore, almost every task has some percentage of unused time. Plus a manager's risk assessment.
At the end of the sprint, all unused hours are added up to cover overtime.
Overtime occurs on a small percentage of tasks, but sometimes significant: 200-300%. Not everything can be foreseen.
Almost always, the total supply of hours is enough: for example, there are 10 tasks in the sprint, the developers estimated them at 5 hours each, the manager laid down another 5 hours for risks, in total 10 hours each - 100 hours were reserved for the sprint, as a result, 2 tasks were in big overtime, and ate 30 hours, 3 tasks went into overtime, but met the risks - another 30 hours, but the remaining 5 tasks fit into the assessment (without risks), and they took 5 hours each - a total of 100 hours were reserved, 85 hours spent (30 + 30 + 5 * 5), in the assessment fit, the customer is happy.
But it rarely happens so badly, mainly because. developers evaluate comfortable time, in reality it takes less than an estimate, i.e. Estimated in 5 hours, done in 3-4 hours.
And those hours that are left after covering the overtime are not billed, so the work is usually cheaper than the agreements. A small nice bonus for the customer. You can safely appropriate these extra hours, but long-term relationships cannot be built on deceit, so what the hell.
In the case of absolutely force majeure, when we go beyond all the risks, we discuss possible options with the customer. Usually customers go to a meeting and allocate additional resources.
Advantages of the approach: we almost always fit into the budget, often cost less than agreements, good relations with customers.
The downside of this approach is that you have to reserve part of the budget for risks, this part of the budget is frozen for two sprints, and can be poured into the 3rd.
As a result, the customer has to budget a little more than necessary, but due to this we get much more predictable and stable planning, moreover, in most cases this small excess of the budget remains unused, and on average it is free: what is frozen in the current sprint will unfrozen in the next, and through the sprint can be put into action.
As a person who constantly forgets to log the time in the jira or does it retroactively and from a flashlight, I can name the following reasons:
- I don’t see the point, they never figured out how much the estimate was and how long it took, but I can’t find more use
- it’s difficult to do it reliably , it’s difficult to even reliably maintain the status of tasks, (I actually focus on it, filling it in retroactively, but sometimes it turns out to be outright crap and then you just “paint” a plausible picture)
- jira is an inconvenient tool for this, I didn’t find it convenient at all
Tips, if it’s real for you necessary:
- automate as much as possible, best of all focusing on the statuses of tasks in the jira, ensuring their reliability (do not allow the absence of active tasks, do not allow several active tasks at the same time, do not allow work on "foreign" tasks, etc.). You can start some kind of chat bot, which once an hour will ask "what task are you working on?"
- do not just explain why you need it, but regularly show that you use it for the stated purposes and bring the declared benefits, for example, at daily rallies analyze how many hours the team worked over the past day and identify the reasons for deviations from the plan
At a distance, forget about time tracking. It is often inconvenient for people to work according to the usual “working” work schedule and therefore they choose the hours that are most convenient for themselves: it is convenient for someone during the day, and for someone in the evening, at night, in the morning, when the children and wife are sleeping and distractions are at a minimum. Then they can work more than 12 hours. Do you need these revisions?
On the other hand, some developers neglect remote work and openly freebie, procrastinating to the maximum. So you need to set the deadline for completing the task - 2-3-5 days, depending on the complexity. Failed - minus. Scored a critical mass of displeasure - a candidate for elimination. Fortunately, there are a lot of unemployed now and you will find working hands quickly.
The task of the deadline ( and not the developer! ) is to track the performance and diligence of the developer.
The task of the PM is to track percentages of completion (stages) for current work in the current milestone of the project (and plan for the next one). In case of any delays - quickly identify a specific problem and solve it without disrupting the deadlines for the implementation of the entire project (or milestone).
Asking someone to track the time every hour manually (or install software) is extremely stupid!
lahomie93 , counter question: what exactly do I get money for:
1. For hours spent?
2. For completed tasks?
3. And how and who exactly evaluates their complexity?
UPD: on my blog(link under the nickname) this question and the answer to it are covered in detail.
I have a program that writes a log of my time, I take into account what I worked on and how long. with some clients I work on an hourly basis - this is the basis for me to bill them. Well, in general, it is interesting to know what is "earned", what is lost. to one incredulous client, I make a "video" of what and how much I did on his project. Although, of course, if I wished to cheat, I would have cheated without problems.
now advice to you
any action will not be performed if it is not necessary to achieve the result. that is, they will not be tracked if they do not receive a salary for this.
well, and motivation - if you give salaries a bonus for submitting the result before the deadline - they will track faster and much more willingly.
ps if you really want to take into account the time - automate this process.
You throw everyone into one channel for tracking time. From time to time write to the channel so that everyone can track the time. Those who do well can clean up so as not to interfere with them, while others are forcibly forced - there is no fine, and so on.
This is not a joke, you pay money for work, not for time from the air.
Hm. How fiercely people defend their desire not to sit on a leash :)
I will describe the work on the galleys:
there is an upwork-type counter (there are enough of them), it keeps track of time with screenshots. The developer changes tasks in the tracker,
changes the memo (comment in the tracker) every 2 hours, changes the active task in Trello/Fat as it is completed. At the same time, there is no more than one in the work, inside the task there are checkpoints required for implementation for a clearer understanding.
Agile work, Scrum. Requires planning tasks for the sprint, an approximate estimate in hours or arbitrary units. Sprint - a week or two. Every day stand-up for 5 minutes following the results of the previous day and planning for the next one.
Every evening a written report on the completed tasks and a plan for the next day.
The project manager monitors the progress, if there is a speed drawdown, he pays attention to it in time, with the team he tries to find a solution to the problem or notify the customer about the difficulties that have arisen.
Payment - fixed salary for work at the rate of 8 hours per day according to the counter. Overtime is a separate fee.
For what - the office works on the implementation of custom development. Preliminary, there is an assessment of labor intensity, based on the planned hours, the cost is rolled out to the customer. Accordingly, the control of the time used for tasks is important and necessary for the manager and accounting department to evaluate whether the work is going according to plan or we start working at a loss.
Efficiency of the methodology - for 5 years the office got into the top 50 IT rating in the Russian Federation, the constant growth of both projects and staff.
The work is not easy, either you become an effective specialist, or you go for a walk.
In the company, team leaders look after the wards, if problems arise, they look for a possible solution, they help.
Strict control of processes, labor productivity. The growth of knowledge is welcomed, the projects are very diverse.
You can work at any time, estimate the hours spent at the end of the week. They won't pay money for free.
Here they write 'I managed to meet the deadline' and other nonsense. Question - who sets the deadline in your offices? Agile implies self-management of development teams. The numbers are not taken from the ceiling. Techniques allow you to evaluate the speed of the team’s work, review past sprints in retrospect, find solutions to any problems, find out whether we are on time or not. Everything has been invented for a long time, take it and use it.
If an employee does not see the need for reporting, it does not matter to him whether the next customer comes, works hard - what is he doing here then? Reporting is important for predictable outcome. For normal work, it is necessary to formulate the basic values of the company, which will allow it to survive and be effective. There is no goal - four blacks will dance a funeral.
Learned a lot of interesting things from this thread.
It would seem that the author asks one simple question "why do people not note the time spent?". And so much empty chatter, unnecessary advice, not related to this issue. Razmusolivaniya some situations that they have at work, but, again, absolutely not related to this issue. Or a transition to another topic, for example, there was a lot about task assessment, yes, these are all interesting topics, but again this does not apply to the topic of this post.
The author pleasantly surprised me, trying at first to answer everyone and to the point, but was already tired by the end, although the correct answers are not on the most liked answers.
At one time I worked in a company where after each task it was necessary to write a report, attaching a code listing! since in reality, another department was involved in the submission of reports (according to the hours written off from such reports, an invoice was issued to customers) and they were not even aware of what the programmers were doing. I have never seen more insanity in my life. And when I started working in other companies where you just need to write off the time spent, I realized that it’s much easier and I understand why management needs it.
But then I met people who do not understand this. That is, in the current office there was no such thing that it was necessary to write off time. Many (of the old people) frankly sabotaged this process, the resistance is simply enormous, one of the leading opinions was, approximately, the following: "I did not write off time before and now I'm not going to, because I don't need it." It is clear that in such a situation, all management and planning is in full ass. Well, I understand. And here comes the main snag and the answer to this thread. You need, firstly, to correctly explain why this is necessary. Second, make time tracking an integral part of the process.
There have already been many arguments on the first point, and the options for a particular person can be quite specific. What if an employee simply writes off time, then his load will already be visible and fewer tasks can be set in a sprint. Or another example, one employee is working on some project and it is neither cold nor hot for other people whether he writes off time or not, but some task appears, he needs to make an api for another employee. And this task is blocking, but he does not do it. Making excuses at rallies that he was doing something else. "Dude, I don't see what you are doing, if there is no deducted time, I can assume that you are not doing anything, but at the same time the blocking task has been moved to another sprint" - debriefing and rationale for writing off time.
There is such a type of people, somewhere here, too, flashed "I score on the time written off and at the end of the week I just draw numbers from the ceiling to fall behind." This is also present. This is low culture and low discipline. And you can work with this too (this applies to the second point). Solutions were also offered in this thread, someone said that the team leader said to write off time at the end of the day, as he finished working. Well, forbid at the program level to write off time for previous days. In fairness, it should be noted that sometimes in the evening you don’t always have time to do this, that is, you can make it so that you can write off the time of the previous day before the daily rally (until 10 am, for example). If this does not happen, then in the presence of the whole team, clarify what the person did yesterday and stress that instead of
Summarizing, firstly, it is necessary to clearly explain why the employee (not you, but him) needs it, secondly, it is necessary to make writing off time part of the workflow
The same problem was, but the team leader controlled and asked every day and everyone got used to it. We do this not because of discipline, but to improve planning, adjust coefficients and calculate the cost of features, well, we cut unproductive tasks like rallies
I had problems that the developers do not check the time for tasks. Prohibited check backdating, and the problem is solved. Everything is on time. No time - no money.
So far, I have come to the conclusion that the most effective, although not 100% solution, is to set up the tracker in such a way that when you transfer to certain statuses, the tracker requires you to mark the time.
You also need to convey the importance of this at stand-ups / rallies / planning meetings, if it is really important, and remind those who forget in front of the whole team.
ps There was one incurable case, and this person was transferred to hourly pay (although there were other reasons for this). There are no problems with tracking time now.
it has already been said here more than once - the distribution of tasks and workload is a direct task of management, I see no reason to paint the rest of the things
a person who comes to work subscribes to the RFP in most cases and it doesn’t matter to him what he does, but only motivation is important - which is also the task of management, if not strange sounds (well, or the developer himself if he wants more)
the answer is simple - engineers do not need this - this is intellectual work, but you can a) motivate / sell them and b) automate the process (this is the best solution)
and it does not mean "I WANT I DECIDED, I..."
There was such a problem. First option. There was a tracker TargetProcess. Probably the worst tracker in the world. Hook written. Scheduled sprint tasks by the hour. The PM controlled, once a week, that everyone filled in the time.
You can divide the team into two squads
1) They get more for tracking
2) They get as usual for not tracking
That's the motivation
we have the same trouble in the outsourcer. Moreover, there is a tracker of its own, and there is a Porkovsky one. At the same time, the programmer has his own virtual machine for each project (oh, don’t ask why), and the programmer tracks two trackers. The manager at the end of the week also reduces the time, because there may be a difference somewhere. Programmers receive money at two rates - one for the tracker, the second for the bench. In total, if he did not track, then his RFP will be calculated at this time at the benchmark rate.
But there is another crap here, when programmers just spin the tracker.
And then the softserver invites them, and shows that they don’t have trackers - then programmers are happy to go there, because you don’t need to write reports at the end of each day, you don’t need to turn the tracker, sit and do as you like and when it’s convenient, the main thing is to have a result .
In short, pay them two rates and they'll have an incentive to spin the tracker.
By the way, I like Toggl. But we also have our own, our own internal corporate development.
There are challenges in development. Estimate them, set a time. When you close a task, for example, when you commit, you automatically receive the data. If the developer does not meet, he edits the task, adding time.
You have a list of tasks associated with the TK and the time to implement it. Plus, you can evaluate estimates and actual time.
Control time from a week to 2 weeks. On Monday, hold a meeting, summarize and distribute tasks. For people on a salary ticket system. There, control the average load in the region of 60%.
In general, the dealmaker is interested in this list, they pay him exactly by the hour. And the customer himself wants to see not the ephemeral track "Worked on the search subsystem" for 2 weeks, but a whole bunch of subtasks that can be seen from the log.
Calculating how much the developer has earned the company gives the author a great experience in sales, where it is reasonable to evaluate the seller in terms of sales and the implementation of the plan in money. With the excellent work of developers and the shortcomings of analysts, sales department, support, designers, marketers, a company can easily earn 0.
We have all developers, analysts and testers regularly track time: either at the end of the working day, or at the beginning of the next working day. If you were told to track, then why not track? Will you break from this or what? The developer wants to work so that no one interferes with him. So why not meet another participant in the workflow and, in this case, start regularly writing off time.
Didn't find what you were looking for?
Ask your questionAsk a Question
731 491 924 answers to any question