Interview everyone first?

November 23rd, 2010

I see this very often when a company is looking to work with a software development team or persons located outside the US;  high on their criteria list will be excellent language skills. This only makes sense. Everyone needs to be able to talk with each other and to make sure there is understanding of what needs to be done and why it needs to be done a certain way, etc.  But what happens many times is that the company will “rank” higher the person/s who speak English the best, and not take a closer look at the other skills needed to do the job; i.e technical skills, professional skills, etc.  Even if technical interviews are conducted.  Everyone likes a good interviewee, one that talks a lot, gives a lot of clear information.  But does that get you the best person for your team and for the work you need to have done.  Many times it does not.  When you are dealing at a distance, is this standard method of choosing who to work with, really the best way? 

Besides often not getting you the best person for the job, with this method, many companies also often have other concerns such as: 

- Someone out of camera range may be helping the person that I am interviewing, telling them the answers, etc. 
- The person I am interviewing may be taking too long to answer; they must not know what they are doing. The company may not be accounting for the fact that the person does have to do some interpreting in their head to make sure they understand what it is you want. 
- Or the person I talked with does not in the end up working on my project, or I never see that person again. Maybe the vendor I am working with put certain people in place for the interview portion, who could interview well, and then I get stuck with someone else.  Of course this situation more often would occur on a very large project, where it is more difficult to keep track of every single person working on your remote team. 

For these reasons, I believe there is a better way, when working at a distance, to choose who is best to work on your team.  How did companies such as MySQL (now Oracle), which had up to 75% of its work force working virtually, never met in person, decide who to work with? They actually saw their work! Most developers who ended up working closely with MySQL were ones whose work MySQL managers and exec saw as the developers were adding code to the MySQL open source code base. A great way to assess the work someone can do, by actually see what they can do!

Strategy for Success:

If you are looking to work with a person/s, or a team in a distant location, try actually having them do some work first.  Look at the code that they deliver to you, for your actual system. A good way to do this is to work with your vendor to come up with a pilot project or a pilot/test period.  Options for a pilot could be the following:

- One person working on several individual tasks, bug fixes or small enhancements for a set time period, one month, two months;
- Several persons working on several individual tasks, bugs fixes or small enhancements, as many as can be completed during a set time period of one/two months;
- Several persons working on a fixed scope of work, with fixed deliverables, with a fixed deadline.

There are many variants which would let you “try out” who to work with, for a short time period and for a reasonable price, that will give you a much clearer idea of who you are dealing with. This can work as well even if your particular product or application has a long ramp up period, i.e. it takes a while for any developer to get up to speed on your code and application.  Feel free to talk with your vendor about different trial periods/pilot projects which will be comfortable for both sides and give you a clear idea of what the developers working on your project can actually do.

Applying Entrepreneurial traits to software development projects!

October 25th, 2010

Not too long ago I was reading a blog article on entrepreneurship, which I think really applies to software development projects as well.  The article is titled, “7 Warning Signs Your “Big Idea” Is Going to Flop*”.  I am not saying the article should be titled, “7 warning signs your “software development project” is going to flop”, but maybe it could be. Right now I just want to concentrate on the first warning sign – “You keep changing your mind”. Does this sound familiar to anyone who has defined a new project, or who has run a software development project?  Probably it does! 

I love this paragraph from Mr. Chartrand ‘s blog – “Business old schoolers call it “scope change,” and it can seriously hamper your progress. The more you push the boundaries and keep adding to your project, the more it becomes a time-consuming, cost-heavy monster that never ends. Risks go up, your schedule gets trashed, deadlines get blown and quality goes down.”

Wow, the same thing that happens during a software development project. You can change all you want, but realize it could affect your timeline and it could drive costs up!; if doing Agile development, i.e. you may need to add more iterations or you may be at risk of blowing an iteration deadline if you are not tightly controlling it.  But adding more work tends to mean the cost goes up and the timeline extends!

I like the suggestions given as well for how to overcome this warning: “Give yourself a set amount of time to do research and plan the scope of your project before you start. Take a few days, weeks, or months to really think things through. It’s okay to waffle then because no one else is watching, and you haven’t done anything yet, so you don’t have to backtrack.  But once your time has expired, stop, make whatever decisions you need to make, and move forward. Look at it like a deadline. You can change your mind up until a certain day on the calendar, and then after that, you stick with the plan until you’re finished.”

Applying this to a software development project, means it is ok to do a little planning, think and write out what you want. It doesn’t mean you are not “agile” or not “flexible” just because you think about something or write something down. It just means you have actually thought the idea through. I am not saying write a book, but writing down can help you think about aspects you have never even thought of and it can help orient others to the idea. Also great to have someone read those specs over besides yourself. Can they understand it?  What questions do they have? 

Another great suggestion is to give yourself a deadline. For software development deadlines can be imposed by government regulations, i.e. new banking regulations require you to track and report transactions over $10,000 as of the 1st of the year. Deadlines can also be imposed by your clients, i.e. they promised to their user base that this new feature will be available by this date.  Well it had better be in by then. The harder situations are when you do not have a deadline imposed on you and you need to set one yourself, but set one you must, otherwise you will keep incurring costs and never finish your project.  Even doing agile development, set a rule that the latest that changes will be accepted is 5 business days before the delivery date for each iteration (or whatever is appropriate for your organization). Making changes up until the day before the due date, or even the day of delivery just gets you this, (from Mr. Chartrand ‘s blog – “Risks go up, your schedule gets trashed, deadlines get blown and quality goes down. “ Good advice for software development projects!

* To read the original blog text: http://blog.kissmetrics.com/product-flop/#ixzz13QohVRzb

I want my team to be more proactive!

September 24th, 2010

Have you ever thought this about your remote software development team?  I want my team to be more involved, I want them to be proactive, I want them to come up with ideas, etc., etc.  Well there may be some ways that you are unknowingly setting a different tone for the team, (In reality, these issues can come up for both co-located teams and distributed teams; outsourced or in-house).  

Centralized decision making? Do you have a team that has been working with your application for a long time, they are very knowledgeable and can quickly make changes, fixes, etc., but have you centralized the technical decision making?  For example, are all decisions as to changes in the development environment, frameworks to use, database changes (including such things as adding tables, fields, etc.), porting to a new database, etc., etc.,  made centrally, i.e. at the headquarters without consulting at least some of the developers at the remote location (at least the lead one or two persons)? Certainly you can do it that way that is your right as a company to make those decisions wherever you want.  However if you do that, and continue to do it, you are showing several things to your remote team (or your in-house team if you do not include at least some of them in the decision), that you do not respect the knowledge that they have, that you do not value their opinions, and so on. If you do not include them in these kind of decisions how can you expect them to share with you their ideas or suggestions when you are showing them you do not value their opinion on some of the infrastructure decisions for your application?  They may try for a while to make suggestions, but if they continue to see this kind of centralized decision making, without input from others, their enthusiasm may soon die. 

Why you would want to include more members of your team in your decision making process, at least asking their opinion:  Take advantage of the knowledge they have gained, their day to day work with your system may be able to point out where another technical solution may be better, or where the solution you suggested may cause issues in the long run.  It can be done so that it does not add so much time to the decision making process.

So the next time you are wondering why your software development team is not as proactive as you would like them to be, or not providing you with their suggestions, think about how you work with the team when you have research and decisions to be made for your application.  If you make all decisions yourself, or with a small group of co-located people, next time think about including some additional opinions from other team members.  In the long run it will not only help with that particular decision, but will also help with future ideas for improvements to your system.

Continuing the “I have a feeling….” situations, and how to handle them!

August 24th, 2010

Continuing the “I have a feeling….” situations, there is one more I wanted to talk about which often occurs when working remotely, that is the, I have a feeling something is wrong, but I am not sure what?”  This is a harder to dissect, because it could be anything.  I have a feeling they are not telling me the truth, I have a feeling that they are not telling me something, I have a feeling they are keeping something from me, and the list goes on and on. 

Certainly these feels are exacerbated if you, as the project manager, are located remotely and the rest of the team is located somewhere else.  Or a large part of the team is co-located with smaller groups of people in other locations.

The only way you are going to “resolve” this feeling is by asking questions.  Many managers have issues with asking questions, especially it seems Americans. They often take it as micromanaging. But if you are working at a distance, with people from other cultures, how are you supposed to feel comfortable with what is taking place “over there”, if you do not ask questions.  Beyond asking, “How’s it going?”.  I think we expect to get a lot of information out of such a question, but sometimes you have to ask more pointed questions, like, “Did that design that we talked about work out for the function you are working on?” Or something like, “Did you find a work around for that issue with iTunes?”  These types of questions are not micromanaging, they are questions to eliminate the “feeling that something is wrong”, and make you feel confident that everyone is on the same page.  There is always the good question that I mentioned in one of the previous blogs on how to handle “I have a bad feeling” situation, “Show me where we are, what is done, what remains to be done, roadblocks?”, “Let’s walk through the last/current delivery so I understand where we are.”

In another example, if you have this feeling that, “they are not telling me the truth”.  Without getting too touchy feely, think about when you have this feeling?  Is it when they tell you that the project is going well, and you do not have to worry? I don’t know about you, but someone telling me not to worry, it makes me worry.  I prefer the, “show me so I do not have to worry!”  With most application development projects, developed using an agile development methodology, it should be fairly quick that you can actually “see” something, and know if the team is going in the right direction or not. If not an actual running app, but can you see something from the area in question, a written design, database design, or whatever may be relevant.

In summary for project managers, when you have these “feelings that something is wrong”: Do not be afraid to ask questions, beyond, “how’s it going today?”, until those feelings dissipate. It does not have to take all day, or even take that much time, and it will save a lot of time in the long run worrying about something that may or may not be a problem.

In summary for team members: Realize that managers and other stakeholders, who may be located remotely from you, and even at your location, need to understand exactly where the project is at, and they need you to be proactive and alert them to potential areas for issues.  Asking questions, and offering suggestions about the app being developed also lets the manager know you are thinking about the project, how to create the best app possible, and it confirms your level of understanding of the project and gives your manager more confidence in your skills.

Another “feeling” situation and possible ways to deal with it!

June 5th, 2010

Today I wanted to continue our discussion on the “feeling” type of concerns when working with a remote software development team.

For dealing with the situation, “He/she said yes he will do it, or he said he understands, but I have a feeling he/she doesn’t really understand,” I believe it can be handled in a similar manner attempting to obtain concrete data to support or not support your feelings. This feeling can be common when dealing with remote teams located in different parts of the world. If people are dealing with English as a 2nd or 3rd language they may not ask as many questions as you would like which would show they are thinking about the request, or that they understand. The person could be not as comfortable with their English skills to ask the required questions, or they could just be quiet in general, or it could be that they are afraid to ask too many questions because in some cultures it would not be good to question the client. Therefore it is going to be up to you to dig in further and help dissipate your “feelings” that there may be a misunderstanding. There can be several ways to start with this. Ask your remote team to “write down” was is discussing during status meetings and send a recap to everyone in the team, after the meeting. Review these notes to see what is understood or where there may be issues. There is also nothing wrong with asking a team member to repeat what they heard while you are in the online meeting or on the conference call, etc. This request can be validated by simply saying that you just want to make sure everyone is on the same page. No one should have a problem with this type of request for better communication. If you are not satisfied with the oral delivery of what the developer is to do, ask them to put it in an email or an IM. These types of steps should help the PM determine if additional explanation is needed for a particular task. If you now believe you have done all of the explaining you can do, and the developer has explained what he/she thinks should be done, you should at least have concrete evidence as to the level of understanding; i.e. here is what I told them they needed to do, and here is what they think they have to do and the two do not match. If that is the case, it may be time to talk with your vendor about the situation. You may need to make a team member change or you, and your vendor, may need to dig in more deeply as to why there is a disconnect. The last step should bring even more concrete evidence one way or the other. Once the developer starts to actually work on the feature or request, you will also get concrete evidence as to whether or not the developer understands what they are doing. More relevant questions should come up, plus you will be able to look at actual code and very quickly you may be able to look at actual results. All of these results will either help validate or alleviate your feelings that there is a misunderstanding.

I have one more “feeling” situation I want to talk about next time.

I have a “feeling” something is wrong!

June 1st, 2010

We have all had these types of feelings when dealing with people who are located a distance from us.  Or even from managing a project where everyone is co-located.  Maybe it is just part of being a project manager; we are looking for any possible risks that can cause our projects to fail.  Managing at a distance can exacerbate these feelings when you can’t see each other all of the time, or at a moment’s notice when the feeling comes up.  

I have heard this phrase stated in several ways:  

• “I have a feeling they won’t do what I need them to do”,  or
• “He/she said yes he will do it, or he said he understands, but I have a feeling he/she doesn’t really understand.”   Or a more general fear,
• “I have a feeling something is wrong, but I am not sure what?”

I think there are ways to mitigate these risks or feelings that can come up. One of those ways can be to ask more questions and another is to use the work completed as a guide.  

First make sure you feel comfortable with what you have asked them to do and what you are concerned that they won’t do.  Are you concerned that they won’t code something in a certain way?  Are you concerned that if you ask them to make a change they won’t do it? Are you concerned that they “can’t” do what you need them to do?  A lot of these issues are very different but most of them can be addressed by history, if there has been any history of working together, or by asking questions.   For example, if you have not been working together long, and if you have a feeling they won’t code something in a certain way, ask them for a “design” of how they plan to code the feature (for those who are afraid of written documentation, this does not mean they have to write a large document, you are not asking for the world).  This can also be done orally, and if not good enough, ask for a written explanation.  If the design is not what you expected, you have other issues, and you will have to determine and make a judgment as to whether or not the issue is; they did a poor design, it would work, but it is not how you would have done it, or it would work but it is not how it needs to be done for this application.  But at least you will have validated the “feeling” that you had and now you can deal with a concrete situation and not something abstract. 

I will review another one of the “feelings” of concern later this week.

Avoid the “Do as I say, not as I do” syndrome!

May 26th, 2010

“How do I get my remote software development team to be more productive?”  It is something we all want to do, be more productive. One observation I have made with distributed co-located software development teams (i.e. more than one team located in more than one geographical location) is there can be an attitude of: “Do as I say, not as I do”. This can come from the team that works at headquarters or any team that is working in a multi-team situation.  This situation can apply to both in-house teams in multiple locations as well as if outsourced teams are involved.

We know that all teams working on the same application/product should have the same goal; to make the application the best that it can be and provide the best service to the company’s customers.  What can affect this goal is perception that one team is not living up to that goal, i.e. not being as productive as possible.  And what can affect productivity is the knowledge that there are different rules for different teams, or more specifically the enforcement or said rules.  The rules I am talking about related to productivity may be rules for design, level of design documents, coding standards, conducting peer reviews, commenting, documentation, etc.  I am making the assumption that there is one set of rules for development for the entire system, which most likely there is, every company strives for this.  But productivity, let alone motivation, can go down real quick if one side sees that they have to follow the rules and another team does not. 

It can be hard enough, but not impossible, to make distributed teams feel like they are one team and working towards the same goals, but when one side sees clearly (and in software development it is not hard to see concretely the differences) that another team is not held to the team coding standards (whether they be team specific standards or industry standards), or testing standards or documentation or commenting standards, etc., motivation goes out the door. If everyone is co-located, we tend to more easily enforce standards across a team, but it is an easy thing to let it lapse when there are several co-located teams in multiple locations.  It is easier to dismiss one team as not being productive, and to not delve in to why that may be occurring.

The fix to this issue is simple. Make sure all team members are living up to the standards the team and the company have set for themselves and for the application and system that is being built and supported. There is never any excuse for one team to get away with not following standards because it is felt they have to push some changes through at the last minute always, for example, and therefore never have time to follow the standards. In software development, what team ever has all of the time that they would like?  No team ever does. If it was truly the case where a last minute fix needed to go in, go back and make the necessary updates to meet coding standards, documentation standards, etc (although most teams would let it slide, for the good of the team, if they saw that this was an emergency situation and rarely occurred). Having one team do their own thing, and tell the other teams, well “Do as I say, not as I do”, doesn’t work for teams the same as it doesn’t work for the leaders and managers of teams. Follow the same rules for all team members, no matter where they are located, and distributed teams will work much smoother!

Cloud computing – Excellent panel discussion

May 3rd, 2010

This was one of the better panel discussions I have heard on cloud computing. Putting it alot more in to perspective, and showing the realities of it.

Finally the recording is available:  http://tinyurl.com/GABA-Cloud

Book Review – Cloud Security and Privacy – An Enterprise Perspective on Risks and Compliance

April 19th, 2010

The authors of Cloud Security and Privacy recommend this book for technically savvy business persons who are thinking about using cloud computing and are interested in protecting their information and are wondering about any security concerns.  This is probably the perfect audience for this book, as well as it can be used by business persons who not as technically astute but who are interested in how cloud computing could be used by their business and what issues there may be with it. They can get an idea of the questions they should be asking (which of course technical people are going to love…..).  It also is a book that can be used as a reference, even for technical persons, parts of it include best practices on securing virtual servers.  If not familiar with that, this book can be a good reference, won’t give the entire how to’s but can introduce many of the security areas. 

Since the “cloud” is a moving target, probably parts of this book can be considered out of date already since it was published in September of 2009, however, if you want to know what the cloud is, how the “industry” defines the evolution to the cloud and to learn how or if your company could benefit from it in a realistic manner, this is the book for you.  If you want to know what the cloud is just out of curiosity, this book is way too much for you. 

Cloud computing puts more decisions in to the hands of business people, rather than IT, I am sure we have all heard that before (about earlier forms of Cloud computing – ASP’s, etc.), but a good example of where this has been true has been with the use of SaaS (Software as a Service – which is now considered to be a Cloud service). A wide range of companies; large to small, are already using cloud services such as Salesforce.com. As well, a large number of small and medium sized businesses are using Intuit’s online QuickBooks service, so more companies are already “in the cloud” than probably realize it. From this book these same people can learn more about the other types of cloud services which may be applicable to their business as well.  

There are still a lot of definitions floating around about what is “the cloud”, and experts still do not agree so the book lays out what may be one of the commonly accepted definitions, or not, but at least it gives a basis for the rest of the book and the range of what will be discussed.  What can be mostly agreed upon by experts with regard to cloud computing are the accepted attributes of the cloud which must be: 

1. Multi-tenancy enables sharing of resources and costs across a large pool of users thus allowing for:
2. Massive scalability – has to allow for massive scale in both compute power, bandwidth, storage. Meaning the ability to scale to thousands and thousands of machines, the type of size that you need if you are an amazon or google and that you needed to build for yourselves, now making that available to others.
3. Elasticity – Users of the cloud must be able to rapidly increase the amount of resources that they need, and then release those resources for others to use when they no longer need them
4. Pay as you go – Traditionally for getting your app out you paid a set price, and often paid for more than you needed, or usually needed because you were building yourself or buying what you would need for peak times
5. Self-provisioning of resources – users can use what they want to use for storage, cpu power, network resources

Also important to define is the three types of Cloud Service Providers (CSP’s); IAAS (Infrastructure as a Service), SAAS (Software as a Service), and PAAS (Platform as a Service). 

Chapters 3 and 4 discuss specific areas of security; infrastructure and data security and storage.  There is a good breakdown for the different types of CSP delivery methods and the different types of security.  The authors make it clear though that many of the security issues are not specifically caused by the cloud and they may or may not be exacerbated by cloud computing. 

A great point of the book is that it emphasizes what the CSP is responsible for, and what the customer is responsible for and where it is still questionable who is responsible for what. This is emphasized throughout the book.  So depending on the service, for example the SAAS model such as Salesforce.com or Google Apps, it explains what Salesforce.com is responsible for, and then what the customer is responsible for such as operational security (such as user and access management). It also goes in to detail as to what type of security review the customer should do of the vendor such as:  requesting information about the provider’s security practices.  This information should include their application security testing, release management, authentication and access control, etc.  Although to date much has already been written about what type of review an enterprise should do of their SAAS providers practices.  But the sections for the IAAS and PAAS providers will be interesting as well. 
Good points in the Platform as a Service (PaaS) delivery model includes software vendors such as:  bungee, Eucalyptus, CSP’s such asL Google App Engine, Salesforce.com’s Force.com, Microsoft Azure, etc. In the multitenant PAAS service delivery model, the main security issues are containment and isolation of multitenant applications from each other.   Since applications are developed by the customer, the customer is responsible for application security.

One of my favorite chapters is Chapter 6 – Security management in the Cloud.  After taking the reader through network, host, application, database, storage and web services which include identity services, this chapter steps though understanding the scope of IT system management and monitoring responsibilities that fall on the users shoulders including: access, change, configuration, patch and vulnerability management and those that are the responsibility of the CSP.  

The authors have reviewed the disciplines for common security frameworks such as ITIL (Information Technology Infrastructure Library )and ISO frameworks and they have identified the relevant processes and the recommended security management focus areas for securing services in the cloud including availability management (ITIL), access control (ISO/IEC 27002, ITIL), etc.  So those that are familiar with these processes will find that they know most of what is in this chapter, but if your organization does not yet use a security management framework they will understand the pros and cons of using one.  But it is good that they took standard security frameworks and based on that same terminology pointed out which ones a CSP would have to think about, which ones a user of a CSP has to think about, etc.  

The authors also have identified what security management processes which they feel are relevant to the cloud, the full list is available on pg 113.  Table 1 is a good chart of the security management functions for each type of cloud deployment/SPI.  
A good point that the authors make, that they feel is relevant to cloud computing is that organizations (people and processes) and information systems are constantly changing.   Management frameworks such as ITIL will help with the continuous service improvements that are necessary to align and realign IT services to changing business needs.  So for example this could mean that continuous service improvement means identifying and implementing improvements to the IT services that support business processes such as sales force automation using a cloud service provider.  Security management is a constant process and will be very relevant to cloud security management.

Chapter 8 on Audit and Compliance also does a good job defining what the CSP is responsible for; good list for the users of CSP’s to understand.  For example within Asset management, access control – data protection/segregation/encryption.

The author’s make it clear that audit and compliance are big issues when working with outsourcing vendors as it will be with cloud service providers.  I would have like dot have seen a chart or something which would have shown: what a user needs to think about when using a cloud service provider and what you would not need to think about any more.  i.e. is it a new issue that you have to think about because you are working with a CSP, or do you no longer have to think about it, or does the CSP have to think about it now?   What would be avoided security issues, what would be the new ones, which ones are the same?   

Ongoing this book can be a great reference for operations managers or business owners or managers wanting to know what research how the ‘cloud” can impact their company.  Conclusions in a lot of books can be “weak”, this one is definitely not weak. It is an excellent summary of the security concerns that are applicable to cloud computing. One could read chapters 1 & 2, get an overview of cloud computing how it has evolved and then actually read the summary, get an overview of the issues and then read the appropriate chapter for the type of security concerns. 

Cloud computing events are still hot and heavily attended.  I was just at another on the 13th of April in Palo Alto, California, which included panel members from SAP, Citrix, T-Systems, and AT&T there was a lively discussion of what people are looking for with regard to cloud computing: on demand computing, as needed consumption of compute power.  http://gaba-network.blogspot.com/2010/04/cloud-computing-2011.html.   Models that they are seeing, dominate capacity in-house yet, elasticity is rented out (bursting in to the cloud as needed).  If you are trying to use cloud services for disaster recovery, for example, or contingency purposes, there are still some issues such as getting a VERY large database server up immediately, transfer rates not there yet.  Web servers can be up immediately, but a database server can be brought up only a day later when the data arrives by disk.  Cloud Interoperability has claimed to be a major issue of cloud computing, since there is still no reason for the cloud service providers to work together.  However, the guys on the panel claim it is not a problem.  In reality I would have to agree with this, depending on what you are running in the cloud, and how it was architected you can technically move clouds.  More of the issue, as with most business decisions, is how much effort will it take, as any move requires some effort, and how much will it cost. 

Cloud Security and Privacy – An Enterprise Perspective on Risks and Compliance
By Tim Mather, Subra Kumaraswamy, and Shahed Latif
Copyright © 2009

Upcoming Webinar on Ukraine as an Outsourcing Destination

December 29th, 2009

The Central and Eastern European Outsourcing Association will be hosting a webinar on January 13th, on Ukraine as an outsourcing destination. 

The webinar is free, to register:  https://www2.gotomeeting.com/register/765874842