Cost is often the key driver to offshore resources, but is it really as cheap as it seems on face value? This paper looks at the challenges; benefits and potential issues Businesses face when they choose to offshore testing resources.

Introduction

Many companies are looking to offshore all or at least a significant part of either their development and/or testing or have already done it.

1. What are the reasons behind this?
2. Equally, are other disciplines such as Business Analysis and Project Management likely to follow this model?
3. Does the methodology that you are using make any difference on the decision to offshore?

In answer to the three questions above:

1. Primarily cost.
2. Unlikely that the BA discipline will move offshore as they have to work closely with the Business, but for Projects Managers maybe there is an argument for this.
3. Seems not to make a difference, although with the increase in Agile this presents a greater challenge where collaboration and interaction are key to success and reflective of the agile manifesto

Offshoring is a topical subject and everyone seems to have an opinion. I would like to share some of my experiences and thoughts including those of peers during our many discussions. Firstly, what are the key drivers to want to move to an offshore model? This seems to be overwhelmingly linked to cost, a reduction in cost. This is what I would like to explore in greater detail in this paper. I have gathered some of the costs which I have believe are sometimes overlooked or not even thought about. If you have metrics from your projects you wish to share I am happy to hear from you.

In order to keep this paper to a readable length, I have picked the Test Analyst to complete this study against. Maybe there are other papers worth writing for the other roles of testing? In order to keep the comparison as consistent as possible, here are my assumptions:

• The Test Analyst (TA) has at least 3 years’ experience
• TA is ISTQB Foundation certified as a minimum
• TA will complete analysis, design and execution activities
• TA not required to have automation or non-functional skills (awareness only)
• The team consists of only TA’s and they are either all offshore or all onshore
• Test management is done onshore and not by TA’s

My belief is that from the perspective of a flat comparison of charge out rates the offshore TA will always be lower, although even these rates are moving closer and closer. However, there are also hidden costs associated with having offshore resources on your team. Some of these are quantifiable whilst others are subjective although I have tried to put costs on these.

Charge out rate – Onshore TA = $550 and Offshore TA = $275. These rates are based on time and materials and do not include any discounts.

Costing

Up skilling 

Let us look at some of these hidden costs in more detail. Firstly that of training, whether that be system, tools, methodology, etc. At this stage, I will state that hopefully the offshore resource chosen has at least domain experience. Most offshore models have at least one person or part of the team come onshore at the start of the engagement. This may be for a few weeks, or even as long as months. Often, there is a higher rate charged by the vendor for this or even passing on of flight and accommodation costs. If there is only one person used for this model, in my experience this is not very successful. We are only human and in most cases there is a lot of information to absorb for one person. If there are any misunderstandings by that individual, then this will be passed on to the rest of the offshore team. Who knows what ultimate cost this may have in terms of missed defects, incorrect testing, poor coverage, etc.? It is important that a strategy is developed in order to ensure that the knowledge gap is bridged.

Of course, if more offshore resources come onshore to up-skill, this is more expensive but may mitigate some of the risks stated above. When starting a new role there is always some up-skilling required, however this is normally done by pairing testers with those more experienced. Also, is the fact that they will always be onshore so there is less pressure to learn everything in one go, as it is easier to ask the question later when that person is sat near to you. For this paper, I have considered additional costs, for example training. That may be 2 days training on the tool and another 5 days training on the system, as well as where heavy process is used an additional day on the specific company process, use of templates, etc. This can be considered as needed to be completed for all resources as even if you take the initial model and only one offshore resource coming onshore, they will need to go back and train the others.

Infrastructure

Another potentially significant cost is the additional infrastructure required to support offshore teams. This may take the form of the following, although is not limited to this list:

• Telephone conference facilities
• Changes to firewalls
• Additional security
• Facilities for video conferencing
• Instant messaging
• Wiki (Intranet)
• Projectors

Increased telephone charges due to international calls replacing either internal calls or simply walking over to a person to speak to them.

As well as this, the company may be forced into the purchase of a tool, where walls and sticky notes were being used previously. Depending upon the number of users and the requirements of the tool, this could add significant cost to the project.

The Team

One “cost” that is often not looked at is that of the staff themselves. How do they feel about working with people that do not share the same culture/beliefs/work ethics as themselves? Often there are frustrations that start to build up where there may have been a small misunderstanding initially that was not dealt with and has now escalated. Simply seeing someone’s facial expression when you first did the act and recognising its affect could have meant that you took corrective action right away.

If there are time differences this adds further complications. There could be the cost of delay in delivery of coded functions; longer turnaround times for defects; defect fix counts may increase where it may take multiple attempts to resolve issues or annoyance if staff are asked to come early or stay late for meetings. Face to face communication is by far the easiest and most effective. It could take 5 minutes to explain a defect when co-located but because they now need to document it in detail it may take 15 minutes. There may also be several rounds of the defects as further explanation is needed, that classic being “I cannot recreate it”. Often the defect count goes up as well as further compounding these extended timings.

You can often find that a sub team culture is created, almost a “them and us” scenario. Instead of the team all putting together as one, there is opportunity to play the “blame game”. He said this, she said that or how can I fix this as it does not have enough information etc.

Communication, Trust & Respect

Communication, trust and respect are some of the core fundamentals for a high performing team. If you do not get this right then it is unlikely that the team and therefore the project will be successful. Whilst there are plenty of technological solutions to communication mechanisms trust and respect need to be earned. When team members are off-shored this is often difficult or takes a longer length of time.

A lot of work needs to be invested initially to get this right. The best forms of communication then need to be used to sustain this. Using an offshore contractor can result in delayed, poor or miss-communication, particularly when you consider cultural and time zone differences.

One of the reasons Agile approaches are so effective is open communication. Yes, the regular daily standup meetings are part of that, but arguably these can be scheduled so this should not be too much of a problem. Everyday interactions are another part of that – over the table, over the wall – you can hear the conversations going on and may be able to shed light on their question, or gain insight on something you are working on as well. When teams are divided across multiple locations, this can become more challenging.

Lost communication cues can play a part of this – the uncomfortable expression can be easily missed, even on a video chat. The subtle shifting of the body where people don’t agree and do not want to cause a problem can simply never be noticed. Therefore, communication is a key factor and should take into account different aspects such as culture, time difference, respecting them as an equal partner, visibility of execution through videos or remote desktop, day to day follow up and finally review & appreciation.

When there is no central location where the team is based, remote meetings, internet chat sessions and the like become the norm. If most of the team is in one physical location and a part is somewhere else, it can become very easy to “forget” about the other team. I don’t think this happens intentionally, but it happens. I have seen it happen where one person or one group, tend to get left out of conversations. Or the conversations may happen, but they get left out of these and are only told the result and the actions they need to carry out without the rationale behind it.

The biggest challenges are to keep clear and detailed communication of the work. It would also be best practice to follow up with a governance call with the business at least once a week to relay the updates from the team and highlight any issues that need discussing and / or resolving.

Some ways in which you can keep team members communicating well are as follows:

• Always appreciate if a good job is done on time it motivates the team.
• Provide templates and other documents to offshore to avoid duplicate efforts.
• Allocate the work and build a plan accordingly to minimise the use of time difference.
• There will need to be increased reporting mechanisms but try to keep those to a minimum, avoiding double entry of information.

Methodology 

The test methods would need to be outlined in greater detail to ensure there are no misunderstandings. It is important that both teams have the same understanding of the process. The delivery needs to be iteratively reviewed with feedback provided. This may mean that additional Team Leads or Managers will be required, adding to hidden costs.

Managers will need to ensure that work types match the resource skill base, although this may not always be possible if they do not know the offshore team well. This may then require an additional level of management offshore. There is also likely to be increased management overhead managing time, language and cultural differences. Ensuring that work is forward planned with appropriate completion targets to ensure resources are always actively engaged.

Who is the escalation point if the TM is not there for example in different time zones? This can cause delays in testing or incorrect decisions may be made.

It is likely that there will need to be increased feedback on test deliverables, particularly in the early stages of the relationship to ensure that they meet the required quality and align to the rest of the project. As well as internal feedback there may be a requirement for the business to review tests. This may need to be conducted onshore through workshops in order to provide timely feedback loops and stop questions/answer going backwards and forwards. There may also need to be additional technology support in order to track these changes and review comments.

All of this may result in an increased need for a governance layer of the process. In some cases the existing process and procedures may need fine tuning the process which could also slow down the projects that are not using offshore resources or result in two differing processes.

Culture

On many projects, when offshore services have been engaged, one of the major obstacles has been the cultural gaps and the fact that there is no “one size fits all” approach to cope with the significant variances in cultural attitudes towards work, management or even communications to name but a few.

The cultural dimension is forever evolving; so one cannot label a nationality as having a uniform way of working. We do not have a single way of working in the west; so it is not surprising that there is no single way of working for offshore suppliers. If one is working with multiple offshore supplying countries simultaneously, it can be challenging to accommodate all the different ranges of management styles required and also to rationalise them internally within the project to achieve consistency of results.

Of all the challenges mentioned, I think communication and cultural differences are two of the most important challenges that usually come up in any offshoring engagement. However, one strategy to overcome these two challenges is by having an onsite coordinator who would take care of both the issues i.e. act as a communication link between client and offshore team, take care of the cultural differences and assist in bridging any knowledge gaps between the teams, of course this will have an associated cost attached.

I feel that there is another major factor seen to be missing from the offshore unit which is ownership or responsibility to adhere to the deadlines. The general feeling is that the responsiveness is found to be very casual and vague at time. It is much easier to walk past your staff and see what they are doing and ask questions on progress right there. Suggested above is having an onsite/onshore coordinator to handle communication, however having a strong and efficient offshore coordinator is also a must as the onsite and offshore model works only when the team combination and coordination is right to achieve the deadlines.

Agile Tester Training

Whilst conducting Certified Agile Tester training one of the activities included is a simulation of one team member being offshore for the fourth iteration of a build. They have already been onshore with the rest of the team for the first three iterations so they have knowledge of the application under build plus have built up relationships with the team. We have a retrospective at the end of this fourth iteration where we discuss how much work they do and how they feel about the experience. Some of the consistent feedback and observations are:

• The work rate decreases
• If they finish before the end of the iteration they often cannot help the rest of the team to complete other tasks
• They feel left out of conversations
• They are unaware what other team members are working on
• They do not feel included

This exercise does not even take into consideration time differences. So it is an interesting reflection on real life and sends candidates away with a much better appreciation of what they need to fact in when considering offshoring.

Conclusion

There are lots of reasons to offshore mainly related to perceived cost reductions or possibly certain skills or knowledge not being available in-house. There is no denying that offshoring can work although there are additional challenges. If the offshoring is simply for cost reasons then companies need to consider the whole costs of working in this way including those potentially hidden ones discussed in this paper.

Download Full Whitepaper