Create an Accurate Test Estimation Model: An Interview with Shyam Sunder

In this interview, Shyam Sunder, a manager for operations, quality, and testing at Sidra Medicine, explains how you can take different test estimation models and use them for your own real-world scenarios to greatly decrease the variation on your everyday testing.

Jennifer Bonine: All right. We are back with our last couple of interviews for the day. I'm here today with Shyam. Shyam, nice to have you with us.

Shyam Sunder: Thanks a lot. I appreciate it.

Jennifer Bonine: First time I get to interview Shyam. I'm always excited when I have new folks I get to chat with about their experiences. Shyam, for the folks out there that maybe haven't seen you before in our virtual conference, can you tell us a little bit about your journey and how you made it to STAREAST and are a speaker? Obviously, you had a session you gave.

Shyam Sunder: Sure, sure, sure, sure. I've been into IT testing for almost nineteen years now.

Jennifer Bonine: Perfect.

Shyam Sunder: I have been giving presentations in various test conferences like software testing professionals ... Phoenix, Las Vegas.

Jennifer Bonine: Wow.

Shyam Sunder: In EuroSTAR, QA Tag, and in Germany, Spain.

Jennifer Bonine: Nice. You've been all over.

Shyam Sunder: I sent a couple of papers last year to Lee, actually.

Jennifer Bonine: Yeah, we just saw him run by.

Shyam Sunder: Yeah, I said hi to him.

Jennifer Bonine: I know, the real Lee, live and in person, that we all hear from via email.

Shyam Sunder: Correct, correct, correct, correct, correct, correct. Yesterday, I was just surprised by him, the first time I'm seeing him face to ...

Jennifer Bonine: Yeah, right, a real person that's attached to that.

Shyam Sunder: It was just photos till now. I submitted five papers.

Jennifer Bonine: Oh, wow.

Shyam Sunder: And he's really ... Lee replied back this test estimation is a topic, which he would like me to present in the—

Jennifer Bonine: Yeah.

Shyam Sunder: And that's how this journey started, actually.

Jennifer Bonine: Wonderful.

Shyam Sunder: I presented a paper once more. It's with some changes ... and then the presentations, and then I'm here eventually.

Jennifer Bonine: Amazing! Well, it's great. See, it works. For all of you out there that want to be here next year with me, and hopefully be the next Shyam that's sitting next to me that I get to chat with, and see how things are going.

Shyam Sunder: Correct.

Jennifer Bonine: It's possible.

Shyam Sunder: Correct. Definitely. Definitely. I'm looking forward to more STAREAST presentations, actually.

Jennifer Bonine: Yeah.

Shyam Sunder: Because I'm submitting more papers and looking forward to it, actually.

Jennifer Bonine: Wonderful.

Shyam Sunder: Next year.

Jennifer Bonine: So, for those out there who are wondering about test estimation, all of us want to know how long is it going to take, right? All the executives, the leadership, everyone wants to know how long. Right? That's a common question.

Shyam Sunder: Correct. Correct. Correct. Normally I found in a software lifecycle, testing is at the very end.

Jennifer Bonine: Yeah.

Shyam Sunder: The majority of the time is consumed by the requirements, development, execution and then the testing comes. It's always getting squeezed out.

Jennifer Bonine: Yeah. Condensed at the very end.

Shyam Sunder: Condensed. One week changes to two weeks, one week. So unless you have a robust model in place, you'd be always squeezed up. That's what I've been navigating in the tester's admission sessions in my various forums so far.

Jennifer Bonine: Is a robust model the help with that?

Shyam Sunder: It's a robust model.

Jennifer Bonine: Okay.

Shyam Sunder: You have ... I've come up with templates, which depends upon the project needs, the organization need, you need to pilot it, get it into a mature state, run it for two, three months, and then it will turn into maturity, actually.

Jennifer Bonine: Yeah.

Shyam Sunder: That's it. That's what I found, actually.

Jennifer Bonine: Very good. So, for those folks out there who are saying, "I struggle with test estimation," you actually have this model where they could take the templates, bring them in-house, pilot them, so to say. Okay, now when I apply it to my real-world scenario I'm dealing with, how long does it actually take me versus take someone else, for example?

Shyam Sunder: Definitely.

Jennifer Bonine: And then, kind of refine it for your environment.

Shyam Sunder: Got it.

Jennifer Bonine: Get pretty accurate and then use that for your projects.

Shyam Sunder: Got it. That's what I've been doing, actually, for the last ten years.

Jennifer Bonine: Wow.

Shyam Sunder: And the effort schedule variation has changed from 60 percent to almost two to three percent. So, that's what I'm advocating actually.

Jennifer Bonine: Wow! That's amazing. So you're saying, you know, instead of being 60 percent off on your test estimate, you're getting as close as 2 to 3.

Shyam Sunder: Two to 3 percent. Yeah.

Jennifer Bonine: That's amazing.

Shyam Sunder: That's what ...

Jennifer Bonine: That's a phenomenal number. No wonder people are excited about this.

Shyam Sunder: People have been asking me whether it's applicable for long projects or short projects. Depends what other projects you did.

Jennifer Bonine: Okay.

Shyam Sunder: Implement this template. Then let it run for two, three projects, and then it'll mature, that's all.

Jennifer Bonine: Where do people get this fabulous templates and how do we-

Shyam Sunder: A couple of peoples who attended my session today, they have emailed me already that they want this template, so I'll be sharing the templates with them.

Jennifer Bonine: Perfect.

Shyam Sunder: All those ... can you email me directly, in case they didn't get the chance.

Jennifer Bonine: Yes. So these guys didn't get to attend and now they are feeling bad that they missed out.

Shyam Sunder: Send email to Shyam, [email protected], and I will replay back with the template and the presentation, which I did.

Jennifer Bonine: Okay. So hopefully everyone got that out there. If you didn't, there will be a replay and you can replay it and get Shyam's email address.

Shyam Sunder: Sure. Definitely.

Jennifer Bonine: So that you can get the templates.

Now, what was the reaction to the session, then, when you were giving these to people? Were they pretty excited?

Shyam Sunder: They were pretty excited. They are looking forward to it, implementing it in their projects, I believe.

Jennifer Bonine: Right.

Shyam Sunder: Because I started off with long projects and some people came like they'll run for one or two months projects. It is valid for one to two years on five years projects. You just have to implement it, get it to a mature state, that's all.

Jennifer Bonine: Wow. Amazing. So this is great. And how did you come about this? Was it just through experience and the projects you had estimated before gave you the knowledge to build up these templates?

Shyam Sunder: Initially I've been ... I've been to testing for almost nineteen years now. I have been pretty much squeezed up in initial five to six years.

Jennifer Bonine: Right.

Shyam Sunder: So that learning, that experience, made me come up with a test estimation. And that's how I came up with it and I've been implementing it vigorously in my projects.

Jennifer Bonine: That's amazing. That's awesome. Now any other ... so this is your first time at this conference?

Shyam Sunder: STAREAST, first time, yeah.

Jennifer Bonine: First time at this conference. So for those that aren't here, some of the things as a first time participant, right. So we talk to a lot of people who have been at the conference before many years. Any observations for your first time and things you really like about this event or that you've seen?

Shyam Sunder: At audience level, the participation in STAREAST has been phenomenal.

Jennifer Bonine: Yeah.

Shyam Sunder: Normally the sessions, which I've given for testers in other forum, pretty much twenty or twenty-one, but here it was almost like forty, forty-five.

Jennifer Bonine: Wow.

Shyam Sunder: And the way the people were asking questions, the way they were interpreting it, the way they were assimilating it from ... I really liked it. And then already ... it's just one hour. I already received fifteen emails in my email box.

Jennifer Bonine: See, there you go. Out of ... that's a really good stat too. So you said about forty, forty-five people in the room.

Shyam Sunder: Yes, yes, yes.

Jennifer Bonine: And already fifteen in the last hour have emailed. So it tells you, right? I mean, this is a topic that probably a lot of you are maybe struggling with currently, or have struggled with. And getting accurate, right? Because the more accurate you can get, then it builds that trust, right. With the leadership that when you give them an estimate, that it's reliable, that they can count on it, use that for planning purposes, budgeting, all of that.

Shyam Sunder: Definitely. And I'm from testing fraternity, testing community. I don't want to get squeezed all the time. And that's what I'm telling other actually. "Don't get squeezed up. Come up with a robust test estimation technique, and let it work.”

Jennifer Bonine: Yeah, and then they know, right? So they're going to be able to depend on it. Because that's the other thing is ... I think a lot of times when our estimates are off or we're not sure and we just pad estimates, right? Where we put in a lot of extra stuff, because we feel like if we don't, we're going to get squeezed.

Shyam Sunder: Correct, correct.

Jennifer Bonine: That you can actually be more accurate and not have to put that padding in and then have better results.

Shyam Sunder: Definitely, definitely, definitely.

Jennifer Bonine: So, that seems like a really good message. Are you here all week at the conference, or ...?

Shyam Sunder: No, I'm flying back on Friday. I'm coming all the way from Qatar, actually, for this conference.

Jennifer Bonine: Wow!

Shyam Sunder: So I'm flying back on Friday.

Jennifer Bonine: Wow, so you had a long trip over here.

Shyam Sunder: Sixteen hours.

Jennifer Bonine: sixteen hours to be with us. That's amazing.

Shyam Sunder: But I looked forward to these conferences. I want to share my knowledge is the thing ... that's what drives me on.

Jennifer Bonine: That's what's great about this too I think is you see people from many different countries, different experiences and you come together in a forum to share that knowledge.

Shyam Sunder: Correct, correct, correct.

Jennifer Bonine: Which is great. And it just makes it better, right? As people share their experiences and their knowledge. So we're really glad that you were able to be here and stuff. I'm glad that we had the chance to interview you. Now, you said, for the past nineteen years you've been in testing. So when you started in testing, was that always what you wanted to do?

Shyam Sunder: No. I started as a Java developer actually.

Jennifer Bonine: Really?

Shyam Sunder: I was one year into programming and to Java and then due to project I was put into something ... when I started my career there was no specific testing team. Developers used to do testing or just squeezed up. So I was put in a project for almost two weeks for doing some testing and that got squeezed to one week. And I never wanted to remain in testing, but I found it interesting and I carried on. That's what has happened.

Jennifer Bonine: Yeah. We ask a lot of times, folks that we interview, what their story is or how they ended up in testing. And more often than not you find people that maybe didn't consciously say, "this is what I'm going to do," but found their way to it. And fell in love with it, right? Like said, "this is what I want to do. I love this career. I like doing it."

Shyam Sunder: That's what happened, correct.

Jennifer Bonine: Yeah, and you had that Java development background and kind of maybe didn't say, "When I grow up, I want to be a tester." But then ended up there ...

Shyam Sunder: But that has given me inputs to more of automation also. So I do automation testing also.

Jennifer Bonine: Right, because you have that coding background.

Shyam Sunder: So, that ... coding background—

Jennifer Bonine: I mean, so it really helps you.

Shyam Sunder: Definitely.

Jennifer Bonine: In terms of advice for folks out there that maybe are new in their career or new to testing, any things you'd like to share from ... advice, or best practices, or what should they do as a new tester and they're struggling to get information or nowhere to go?

Shyam Sunder: Based up on my experience, when I started off, there was no specific testing areas actually to be centered on. But now there are dedicated areas of functional testing, non-functional testing, manual testing, automation testing. So depends, if you want to get into testing, just see what you want to do. Whether it's manual or automation or performance or something like that. And then there are specific areas of testing.

Jennifer Bonine: Right.

Shyam Sunder: So testing is bigger now. It's not like ...

Jennifer Bonine: It's so much bigger.

Shyam Sunder: It's bigger.

Jennifer Bonine: Yeah, and it has ... you know, when you look back nineteen years ago to today. I mean, now there's whole conferences dedicated to it, right? And careers around it that people have build over years, right? So it's evolved a lot in the last nineteen years. Now, looking ahead. So you're test estimation piece and the love of sharing knowledge, where do you see your career transitioning? Do you have any thoughts on where you want to go next or what your next adventure might be in terms of your testing career?

Shyam Sunder: There are some areas for ... if I say test estimation have still not gone into test estimation still. It's still fuzzy right now for me, because I've done estimation, but it doesn't reach that maturity, which I'm trying to. So there are some areas of testing in terms of estimation and other areas, which I'm still trying to evolve. I've swam for nineteen years in testing, but I want to swim more.

Jennifer Bonine: You want to swim more in there. So more around that estimation piece.

Shyam Sunder: And more into other areas of testing.

Jennifer Bonine: More into other areas. No, that makes total sense. And kind of some topics maybe that other sessions you've attended or anything else that's kind of interested you here at this event?

Shyam Sunder: No, I haven't still studied, I just came yesterday.

Jennifer Bonine: Oh, so you're just getting into everything.

Shyam Sunder: Tomorrow I'll be having a busy day, so I'll be attending more conferences.

Jennifer Bonine: So we'll check back in and see how your day went with everything.

Shyam Sunder: Definitely.

Jennifer Bonine: But we're very glad to have you here.

Shyam Sunder: I attended a speaker session yesterday. Speakers introduction dinner, it was passionate enough—

Jennifer Bonine: Attend that ...

Shyam Sunder: Other speakers, like Lee and others.

Jennifer Bonine: It's a great opportunity, right?

Shyam Sunder: And really, thanks, TechWell, for inviting me for this particular conference, also.

Jennifer Bonine: Well, we're glad you were here.

Shyam Sunder: Thanks a lot.

Jennifer Bonine: Thank you so much for taking time to sit with us and talk to us about test estimation. So, if you have more questions, you'll be able to send an email to Shyam and he'll be able to respond with that template, that hopefully will be useful to folks out there. So we will be back in just a few minutes with our next interview. So hang tight with us and we'll be doing another interview shortly. Thanks, everyone.`

Shyam SunderShyam Sunder is a manager for operations, quality, and testing at Sidra Medicine, where he help manage clinical applications, including Change Control, Service Requests, Incident management, Package Install, Cerner Millennium Domain strategy, Code upgrades etc. He’s responsible for service delivery across multiple clinical solutions and for the delivery and maintaining of Operational Clinical System.

About the author

Upcoming Events

Jun 02
Sep 22
Oct 13
Apr 27