Note: You can also find this post on my LinkedIn.
I spend most of my time away from Singapore. I do try to participate in agile meetups whenever I can, but my schedule often forbids me. Nevertheless, I have a keen interest in what is happening in the agile scene in Singapore. I conducted this survey as a means to stay in touch with the local scene and perhaps to contribute to the nurturing of the software engineering community here in a small way. Of course, there are many ways to contribute. I chose doing a survey over other approaches because I believe that all nurture and progress must stand on empirical evidence and data. Having such data is the first step. I presented results of the survey in Agile Tour Singapore 2015, on 27 November 2015.
1. A Big Thank You to All You Respondents and Organizers!
I created the survey using survey monkey and posted it on the Agile Meetup and Agile Singapore mailing list, the Agile Tour Singapore website and my personal website. My initial fear was that no one would respond. Fortunately, my fears are unfounded. Many respondents filled the survey within a week with responses coming in now and then. I am elated to find responses even today, the first day of Agile Tour Singapore 2015 (see Figure 1). A big thank you to all you who responded. Thanks to Stanly for hosting these communities. Special thanks to the Agile Tour Singapore organizers! they really put in tremendous amount of effort into making this event possible.
Figure 1 – Responses received over time.
2. Respondents are Primarily Practitioners Working in Small Development Contexts
My target for this survey more towards practitioners like business analysis, developers, etc. not so much agile coaches whether internal or third-party coaches. After all, practitioners are the guys that do the real work. They are people with real problems, who need real solutions to their problems. They are less likely to be skilled towards ideology and what agile coaching companies are trying to promote. This is in fact the case. Of the 65 that responded (first row in Figure 2), 52 (80%) are practitioners. This is good. It validates that the survey channels I used are hitting the right audience.
Figure 2 – Who respondents are
Respondents come mostly from relatively small organizations and are involved in small development cases (see Figure 3). 54% comes from organizations having 50 or less people and 61% come from organizations having 100 or less.
Figure 3 – Scale of development organizations and scale of development projects
The scale of development the respondents are involved in is small. 61% (40) participate in development with 10 people or less. 92% of the development have 50 people or less. This is quite expected. Singapore being a small country does not boast huge development centers like what I see often in China, where 50 or more seem to be the norm. This also indicates to me that large scale agile practices and methods may not be popular here, which I will discuss in a minute below.
3. An Even Spread of Adoption Ranging From Those Who Are Just Beginning to Having Agile As Mainstream
So, how far has the respondents’ organization venture into agile? Are they just starting? Has agile gone mainstream? We have quite an even spread here among respondents. Some are evaluating, some have some teams trying. Some have organizational push towards agile. In some organizations, agile is mainstream, all development are agile by default (see Figure 4).
Figure 4 – How Mainstream is Agile in Respondents’ Organization
4. Most Agile Adopters Achieve Their Intent of Improving Time to Market, Quality, Customer Involvement and Team Collaboration
We don’t do agile for agile’s sake. To be useful, agile must solve some real problems and achieve some real results. Figure 5 summarizes the reasons why respondents want to go agile. Most respondents beyond 60% want to reduce lead times (time to market), increase product quality, increased customer feedback, and better team collaboration.
Most consider such as better deployment of resources, better product ideas, lightweight governance as a good to have things, not something critical to their road towards more effective development.
Figure 5 – Why Respondents Want to Go Agile
So, did the respondents get what they want to achieve from agile. In general, there is positive impact on lead times (time to market), product quality, customer feedback and team collaboration (see Figure 6). There is also increased stakeholder collaboration as well, which is a good thing.
Figure 6 – What Respondents Get Out of Using Agile
However, not everyone achieve such results. About 25% did not get any effect, neither positive nor negative. Perhaps they have not tried long enough to have any visible effect, so the jury is still out. Interestingly, there are some instances when things have gone for the worst. There are two instances when team collaboration have in fact gone worse, and one instance when customer feedback deteriorated. This is something worth investigating. There are also reported instances of governance, product innovation and deployment of resources deteriorating. This is also something worth investigating. But for now, this is evidence that agile is not a silver bullet. Adoption of agile needs to be done with eyes wide open.
5. Scrum and Technical Practices Are Most Popular, but …
When a development team adopt agile, they in fact adopt some agile practices. So, what are these practices? What practices solve their problems? What specific practices have respondents tried? Lets see what we can gather from Figure 7 below.
Figure 7 – Practices Respondents Tried
Majority of the respondents (more than 60%) have tried and will persist in using scrum, some form of agile requirements (like user stories), automated testing, and continuous integration. I deem these practices as fundamental and crucial, and many respondents agree with me.
However, there are some surprises as there are instances when respondents are not considering these practices. There are 2 instances when respondents do not want to try scrum and agile requirements, and there are 3 instances when respondents want to give up scrum. Of course this needs further investigation as to reasons why. However, it does say that there is no one-size-fits-all to being agile. In fact, a 2010 report by Forrester Research indicated that only 27% stick to a particular agile methodology as close as possible. The rest will either mix agile methods and even with non-agile methods. It really depends on what challenges respondents are facing and the problems they deem are their highest priority. There is no single road towards agile.
6. Large Scale Agile May Have to Wait
When looking at what respondents are not considering (i.e. “not for me now”), any form of large scale scrum (whether LeSS or SAFe) takes the lead with 45%. This is not surprising since most development here are small scale with 50 people or less. If you compare this with other countries like maybe North America and Europe, scaling agile seems to be the in-thing at the moment.
7. Respondents Keen on Innovation and Agile Governance
Respondents here are more keen to try practices such as design thinking (46%) and lean startup (40%). I do see a growing community in these areas in Singapore. Perhaps this might be the next in-thing here. Perhaps this is where most respondents are filling the pain – how to design the right product, and how to validate the design. This may also reflect the fact that the Singapore government is putting a lot of emphasis in these areas.
Next in line that respondents are keen to try include such as Agile Portfolio (39%), Agile Governance (35%), DevOps (32.5%). These practices have lots to do with how an organization operate as a whole than individual teams. This might be the barrier that respondents need to break out from.
8. Internal Waterfall Processes and Functional Silos Are Biggest Challenges
The journey towards agile is not easy. It is a mindset change for many. So, what are the roadblocks that respondents say they face (see Figure 8)?
Figure 8 – Challenges Respondents Faced
From Figure 8, it is clear that there are roadblocks everywhere, business, technical, organizational, competencies, etc., but the two seemingly biggest roadblocks are internal waterfall processes and department silos, both of which have most instances that stopped respondents completely. This is confirmation why respondents are keen to try practices such as Agile Portfolio, and Agile Governance, discussed earlier.
From Figure 8, we see that the least among road blocks are executive support, and team competency. Executive support not being an issue seem counter-intuitive to me. I would guess that executive support are key to overcome internal process and department silo roadblocks. I know that poor team competency can be a major roadblock. But this apparently is not the case with our respondents! So, we do have great teams in Singapore. Few dysfunctional teams here!
9. Most Respondents Will Persevere on Their Own and Learn from the Community While Some Want External Help
Overcoming the above challenges is not very straightforward. Overwhelming majority wants to persist on their own or get advice from the community, perhaps through meetups, and events like Agile Tour and Agile Singapore. This data suggest that agile which had began as a community movement, and probably continue to flourish as a community movement. So, do continue to support the organizers of these community activities. They have volunteered and they deserve all our support.
Figure 9 – The Kind of Help Respondents Want
Many want further education and training (some 70%). There are few who see the need for external coaches. This I think goes to show that most respondents see the need to grow in-house competencies.
8. Yes, We Have a Vibrant Community, but More is Needed
There are several take ways from this survey.
- Yes, there is an active and helpful agile community in Singapore. I did not put in tremendous effort asking for survey participants, but I was able to get quite a number of responses. As mentioned, my initial fear of the lack of participation is unfounded. I like to say a big thank you again.
- There is quite a well spread of respondents across the agile adoption journey. Some are just beginning, some have gone mainstream. This means that there is a lot of opportunities to learn from one another during community activities and events. In fact, respondents do desire to get help and advice from the community.
- Most development here in Singapore are small scaled and the practices needed here are accordingly small scaled. Well, small is beautiful. So, probably practices for large scale development will find few opportunities here. There will be more growth opportunities next for design thinking and lean startup.
- There are instances when adopting agile produces negative impact. Well there is no silver bullet, there is no one size fits all. Be very clear what problem you want to solve using agile, rather than blindly following some prescribed agile practice or method.
- The biggest roadblocks seem to be internal waterfall processes and department silos. Well, I think this is about changing mindset, not so much as agile practices per se. The principles are well known, but decision makers and policy makers may be tough nut to crack. This will take time, so please be patient. Teams can always improve in other areas while patiently waiting for opportunities to break down these walls. Least among problems are team issues. So, we do have great teams here in Singapore, at least among our respondents!
- Certainly some education, training, and, sharing is needed. This probably has to focus in areas like agile portfolio and agile governance. These are complex topics involving many parameters and trade-offs. My guess is not so much the content, but about getting the right people to take such education, participate, bring out the issues and discuss.
Surveys like this are low cost and effective way of collecting feedback from the ground. But as with all polls and survey, we need to take the results and hence the interpretation with a pinch of salt.
In all, I have 65 respondents, and 43 completed the survey with 22 skipping quite a number of questions. I have sent out an earlier survey (52 respondents, 28 skipping quite a number of questions), with the same questions but different way of collecting responses (respondents need to drag and rank some question options). This second time survey had a better hit rate in terms of getting respondents to fill in the questions completely.
Anyway, 65 (or 43 if you consider those who fill in the survey completely) is not big sample size. Is this sample representative of the local community? Are my questions and options easy to understand? Is each response valid? As I have said, it is important to take the results with a pinch of salt. Nevertheless, while there are some unexpected instances, like respondents giving up on scrum, I see now show stoppers. They are generally agreeable with my limited personal observations interacting with the local community.
With the free version of survey monkey, I am limited to only asking 10 questions, but I am advised that it is sufficient, which now that I re-think about it, is sufficient, at least for the first time I am doing such a survey in the local community.
If you like this survey report, please indicate a like on this post. If you think more surveys (perhaps on specific topics like agile portfolio, governance, requirements, etc.), please follow with a comment. Your encouragement is most welcome.
— Pan-Wei Ng, Ph.D.