Tristan Pollock

Waydev - Git analytics (2018) version 1.0

V1 - The fast & visual way to understand your developers. Waydev analyze your codebase from Github, Gitlab & Bitbucket to help you bring out the best in your engineers work.

Check out our V2 - https://www.producthunt.com/posts/waydev-2-0-1

Add a comment

Replies

Best
Kirk Gray

LOC was a bad way to measure productivity in the 90s. It's gotten no better. This creates all kind of bad incentives for engineers. Number of commits has nothing to do with number of quality commits that delivered customer value.

Pros:

None

Cons:

Horrible way to measure effectiveness

Hamilton Greene
@alex_circei So at first glance, I assumed this was done via LoC as well which I always have a bad reaction to (it's like measuring a writer's performance based on the number of words they've written as opposed to how well those words do their job). Could you break down how you are coming to this "impact" score if it's not LoC?
Vlad Nicula
@alex_circei "where you don't know anything about them" . That's the main issue right there. You don't have to be technical to know how to pick team players, or how to spot dedicated developers.
Peter Bell

It's super important to find ways for non-technical founders to connect with and manage engineers. This isn't one of them. Tracking an engineer by lines of code is like tracking a marketing person by the number of words on the home page. You don't want the most words, you want the most concise way to tell your story.

And number of git commits is not much better as it's super easy to game by just committing more often. Plus, if you want to get a high level sense of this, just look at the list of commits in GitHub.

Sorry - don't like to be negative about a startup or new product as I know the blood sweat and tears that go into making it, but this makes the world a worse place - sorry :(

Pros:

None

Cons:

It's a really bad idea :(

James Barry
This tool is a throwback to the 80's and 90's. Measuring lines of code and commits does not work. It usually leads to bloated code . Usually in the 90's it meant that instead of writing some neat quick routine, you would cut and paste some crap that you found that did the same thing, but had 3x the lines of code. So you looked better. As a developer I would write extra long code (usually slower and less effective) to get my code lines up. As far as commits, my best developer commits a couple of times a week as he wants to noodle on it and make it better. depending on your coding culture this could stifle good code writing. Frequent commits is a way of getting code out quickly, not necessarily getting good code out. Think of the paper at school where it had to be 10 pages. You wrote a great 8 page paper that hit the mark. You would then fill in crap to make it the 10 pages. Same here. Horrible metrics for measuring developers. IMHO
James Barry
@alex_circei I understand why you think its a great idea... But Github does not change things IMHO. We had CVS and tools to see about commits, and I was a co-founder of Collabnet where we launched Subversion. All had tools to measure lines of code by developer and "commits" atomic or otherwise. I think Dilbert unfortunately sums is up well http://dilbert.com/strip/2003-08-26 or http://dilbert.com/strip/1995-11-13 where the incentive for the developer changes (lines/commits/bugs) and that is where they focus, not good code.
Tristan Pollock
Waydev is designed to help non-technical founders understand and manage the work of your developers. You can access the detailed reports on your tablet or smartphone and at a glance you can see how much progress is being made and where the major roadblocks remain. Because it is designed for non-technical founders (like myself), you can still keep your project on track, even if you don’t know the details behind each specific line of code.
Denis Todirica
Congrats for the launch @alex_circei and @buzea_valentin! I saw the Slack and Google Suites logos on your homepage, how are you integrated with them?
Denis Todirica
@alex_circei sounds great! Looking forward for the coming updates and releases.
Alin B

In the entrepreneurial age, devs are the lifeblood of all projects.

Problem is that non-technical founders or managers have little to no control over them.

Having met 10s of founders with failed projects due to complicated circumstances with their outsourced dev teams, this simple tool gives an overview of their work so you can basically sleep better at night and focus on growing the business.

Pros:

For a non-tech manager, it gives an overview of your development teams' work

Cons:

Might be seen as micro-management by devs

Savelii Kovalenko
It's really a pain to track development efficiency sometimes so I see a lot of potential in this product. Great work!
Moritz Plassnig
Interesting product. How do you compare to other tools like GitPrime or Velocity from CodeClimate?
Jonathan Woodard

If you don't trust your technical team to execute the product vision and present you ongoing results, it's not time for a new tool, it's time for a new technical team.

Pros:

Unclear

Cons:

Utilizes specious metrics for "Impact"

Anna Martirosyan
Congrats on the launch! It seems interesting and well executed!
Anna Martirosyan
@alex_circei You are more than welcome, Alex!
Jeremy Thomas
I agree with what folks like Peter and James are saying. Lines of code and commit count are not accurate ways to measure "impact." Usually a high frequency of commits is an indication of poor code quality -- developers are introducing new changes to fix errors introduced by previous ones. Instead, you should expect a steady cadence of incremental changes, where proper time and thought is put into each. Test coverage is another important metric, which is not covered here. Further, most product-development organizations celebrate when developers remove lines of code. We, for example, have a bot that adds a "hand clapping" emoji to a commit when the developer has removed code. In short, a high commit volume and more lines of code are often anathema to what good software development is.
Hendrik Haandrikman

I'm usually not one to go for negative reviews, but:

What is the "Impact" metric?

The metric you need to watch, week-by-week. At Waydev, we developed the "Impact" metric to analyze performance data. Impact takes the following into consideration: (1) The highest chunk of activity you did in the past, (2) The average activity of a developer based on our research and (3) The behavior of your activity in the last weeks.

"Activity" is not the same thing as impact. Teaching 'non-technical founders' that those two things are even remotely related sounds like a great way to undermine your team's actual impact. You don't want devs to blindly ram out and submit code. You want them to come up with smart, scaleable and effective ways to tackle issues and resolve challenges.

Pros:

I'm sure they mean well

Cons:

The core 'impact' metric doesn't signify actual impact and 'trains' non-devs on the wrong KPI's

Hendrik Haandrikman
@alex_circei That sounds a lot more sensible ;) Good job at listening to - and incorporating - feedback
Vlad Calus
What an amazing product, I'll make sure to give it a try! 😍
Gabriel

I've been using Waydev for a while now and I love it! It's easier when you have an overview on your team's activity

Pros:

easy to use, clear metrics

Cons:

still looking for them

Chris Turlica
Really cool product and product team. Good luck!
Ricardo Ghekiere
Amazing product! Just wondering how secure the data storage is?
Karl Karafiat
Hey guys, looks interesting. Let me ask a simple question to understand it a bit better - what exactly is the impact score and how is it calculated?
Marius
@kkarafiat @alex_circei (2) >> The average activity of a developer based on our research >> what does that mean, the average daily LoC and commit counts? What was your research based on?
Marius
@kkarafiat @alex_circei is the behaviour(LoC, commit frequency) of developers in opens source repos the same with the behaviour of a web agency freelancer or a developer working 9-5 in a startup ? (I assume those thousands repos you mentioned are open source repos on GH). Also, would love to read a more detailed blog post on this 'Impact' magic sauce, since it's the main selling point.
Ovi Negrean
Nice work, @alex_circei & team! I'm curious - How is Waydev free?
Anton Ushakov

Can't use superficial metrics like that. Means nothing. Needs to measure actual value created, e.g. by impact on the users, or feedback from customer interviews etc

Pros:

Seems like a good idea to non-tech founders

Cons:

Actually tells them nothing about progress made

Jacob Ablowitz
"Helping non-technical founders understand what's going on with their technical team" is a fantastic objective and well worth the effort - and this is why you're getting consistently positive early feedback from prospective non-technical founder customers. Unfortunately, counting SLOC and/or commits, *no matter how sophisticated your algorithm,* is a fundamentally flawed means of assessing either "what's going on" with a technical team or "how well are they doing," and this is why you're receiving consistently critical feedback from senior technical folks. The big problem you have is that your non-technical customers will eventually encounter the bad outcomes driven by a "how much stuff did I do" measurement approach, and when they do, they will be really angry with you because you will have effectively been lying to them for an extended period of time. *Non-technical founders: please don't ever measure software teams on how much stuff they did. It will make you very, very unhappy in the end.* Digging deeper, this is not a matter of "well what if we analyze the data better?" The problem is that the only real way to measure the impact of a technical team is in their ability to deliver value - and value is not encoded anywhere within any software repo, whether CVS, SVN, git, MTFS, BitBucket, or any other repo management tool. Period. If you want to measure value, you have no choice but to measure against requirements or user stories in some way.
Jacob Ablowitz
@alex_circei I understand all of that, and understood it before you replied. The problem here is that instead of getting "the right data" to analyze the problem, you are choosing instead to analyze "the data we have," which is activity (NOT PRODUCTIVITY) data. Activity data - how many lines of code, how many commits - tells you ZERO about the value of what was produced. Without knowing the value of what was produced you have zero information on productivity, and therefore cannot possibly inform decisions in any meaningful way. My response to your question "How does a non-technical founder take decisions without real data?" my response is "By getting real, relevant, meaningful data and analyzing it - NOT by analyzing whatever data I can easily get my hands upon and hoping there's some vague correlation to the thing I actually want to measure." Bottom line, speaking as an experienced startup CEO and CTO, I'm warning you that you are not only not providing value to users but actually providing them NEGATIVE VALUE by giving them "data-like substance" that isn't actually data!
Jacob Ablowitz
@alex_circei As for "there are developers that are not that good," the problem here is that a significant percentage of "not that good" developer types will actually look just fine according to your metrics, simply because they're creating lots and lots of code and commits, while many "good developers" who know how to do things efficiently will have fewer lines of code and fewer commits while producing more actual customer value. This isn't speculation, this has been thoroughly studied by the Computer Science academic community for literally decades. You clearly haven't studied this as well as you think you have - it's nice that you've received criticism and praise from people including senior devs while building it, but you're failing on a seriously basic level here. I can't emphasize this enough: reporting software development outcomes to non-technical users using lines of code and commits as the basis for your performance metrics is OUTRIGHT MALPRACTICE AS A TECHNOLOGY LEADER.