Case study - Immigration New Zealand (Department of Labour)
This case study is part of the Government use of Short Message Services (SMS) web guide.
Summary
Immigration New Zealand (INZ) discovered that a large number of calls to their contact centres were enquiring about the status of work permit applications. They decided that running an SMS update service could reduce workloads and costs.
- INZ believe that their text message project offers great value for money:
- The SMS scheme takes very little to run – maintaining it is only 0.25 of a full-time employee’s time.
- The scheme is very cost efficient for Immigration. It costs less than $50,000 a year to run whereas the same functionality delivered through outbound calls and letters would cost somewhere between $600 thousand to a million dollars per year.
- They have had overwhelmingly good feedback from the scheme - 80 to 90% of clients queried said that as a result they had made fewer calls to the call centre, 99% approved of the concept and wanted Immigration to investigate further.
- Immigration New Zealand are leaders in their field - they won a TUANZ award, made headlines when they presented at the Asia-Pacific Economic Cooperation (APEC), and have been contacted by immigration departments from other countries who are interested in starting something similar.
Details
First thinking
In 2005 Immigration New Zealand (INZ) at the Department of Labour realised that they had under-used data sitting on their system of internal processing of applications; logging when they received an application, when it went from one branch to another, how long until people's permits expire etc. This data was sitting in their system, but was only used when they were actively checking on an application. INZ started thinking that there might be a way for this data to be put to better use.
As part of their system of customers’ contact details, Immigration had been collecting cell phone numbers, but were not using the numbers as the usual form of contact.
At the same time, their call centre was very busy. After a review of calls was completed, INZ identified a high percentage of traffic as status-checking calls such as "Have you got my application?" "What's happening with my application?" "I haven't heard anything about my application" and so forth.
Having identified these inquiries as being resource-intensive, the question remained: how could INZ improve the experience for their customers by using their data more efficiently? Putting together the requests for status, cell phone numbers, and under-used data, Immigration designed an SMS solution that targeted status checkers putting into use the extra data they had in their system. They decided that SMS would be the best way to reach seasonal workers or new immigrants who may not have fixed addresses or access to the Internet, although they also offer the option of receiving the updates via email.
Initial aim
INZ decided to carry out a pilot of the scheme in Auckland, sending SMS to clients on work permits who had previously provided their cell phone numbers. The message sent was information about the status of work permit applications. A sample message:
Hi Leo. We got your Work appl, no#6527865 on 03/09/2007. We aim to process this in 30 days. Immigration Manakau Branch.
Set up
INZ talked to a few providers of SMS services, and decided to go with Datasquirt's product "Contact" together with Gen-I, the telecommunications company. As well as performing this role, Gen-I acts as a middleman, managing the relationship between Immigration and Datasquirt.
At that time Datasquirt was the only company offering an integrated solution with SMS and email, which was one of the main draw cards for INZ. INZ was one of Datasquirt's first big government clients, so there was an added incentive for Datasquirt to make the relationship work.
Datasquirt as vendor provides training to INZ’s iBranch staff. Datasquirt can also tailor solutions to the client (for a price).
Datasquirt tailored their system (integrated SMS and email) to the way Immigration already stored their data, so no extra formatting by Immigration was required. Changing data from one format to another is time-consuming and can introduce human error, so avoiding this step carries real benefits.
INZ already had the data, so this was a case of better using some under-used information.
Perceived risks
INZ is very aware that a sending the wrong message to the wrong person could potentially land in the headlines of the media. However the text messages do not contain detailed information - first name, the fact that an application has been lodged and where it was lodged. There is not enough information for someone else to do something malicious with the data even if it was to fall into the wrong hands.
As clients provide their cellphone number, INZ views this as permission to contact them via SMS. This is a single opt-in - if people receive texts inadvertently, or don't wish to receive texts, they have to call Immigration to stop receiving the texts. They can also unsubscribe by replying back and this is generally the channel clients choose to use because of its simplicity and SMS messaging is a common form of communication these days.
The transience of migrants and holiday workers led INZ to think that SMS would be a good way to contact them. It is not unusual for these people to move around frequently, so they don't have a fixed address for postal contact, and their landline numbers (if they have one) keeps changing. However people usually keep their cell phone with them, and keep the number if they change handsets.
INZ does not market the text update service, because their clients have to enter contact details on their application anyway. They simply use this already existing form, and clients choose to give their cell phone number when applying for a permit. These messages are going out to people who want to hear from INZ – it is not an inbound contact channel, but rather a way to give information to an existing client base.
In progress
The texts themselves
There is a big focus on accuracy with the text messages as INZ do not want any bad press. Every dataset is quality assured before being queued for messaging. Messages can be sent to a “holding pen", and a staff member or manager is able to carry out further checks to ensure content and customer detail accuracy - this safeguards against any errors.
To start with, INZ sent text messages detailing receipt of applications, letting people know about their application progress, but have evolved their system so that they send text messages out on the following occasions:
- When they receive applications
- When permits are expiring
- When further information is needed
- Ad hoc reminders like office closures over the holidays, changes in office open hours.
- Text numbers fluctuate depending how many applications are in process, or as business requirements dictate.
SMS messages are also sent out when promoting a particular service like job opportunities, such as informing people on working holiday visas that there are demands for fruit pickers or other skill shortage needs.
INZ has quite detailed data, so can easily mine this to target specific groups of clients.
Cost
INZ sends out thousands of texts, which costs approximately $2,800 per month, plus an annual license fee. This works out to less than $50,000 a year. Providing a similar service with outbound calls and/or post would cost somewhere in between $600 thousand to a million dollars per annum. Using text messages means that only the equivalent of 25% of a full time person works on the scheme.
Storage
The outbound texts are stored, however it is very hard to find a specific text after the fact. While possible, it requires searching through the whole data set. Datasquirt could come up with a solution and make it easy to search, however this would have to be customised, and would not be cheap.
Other channels
The same service (informing people about the status of their applications) is offered through email. Initially all people who provided email addresses received emails with status updates, and all people with cell phone numbers received status updates via SMS. INZ soon realised that doubling up was inefficient. Now they email everyone who provides their email address, and send out text messages just to people who provide a cell phone number and no email address. However if something urgent comes up, they will send out a text message (if a cell phone number is provided) to people who they usually email.
If people reply to these emails with questions, an automated response is sent out directing them to an FAQ site, which usually has the answer to their question, however the number for the call centre is on that page so if their question is not answered, they know how to get it answered. INZ do not have a corresponding automatic response functionality for their SMS service.
Feedback
Inbound texts are discouraged, but INZ generally receives about 10 inbound texts a day. These are put through a keyword filter, which sorts messages into those that require specific assistance, compliments and abuse. The complimentary messages are sent through to the call centre staff, and the abuse messages are deleted. The abuse messages are usually from New Zealanders, whose mobile phone number has been keyed in wrongly as the result of simple human error.
Most feedback received is positive - "God bless" "Thankyou" and so forth. People who apply for work permits are happy to hear back from Immigration about their application status and want to receive this information.
When this scheme started, INZ would get the cell phone numbers of 35% of their clients (providing a cellphone number is optional). Now 70-80% of clients give Immigration their cell phone numbers. As INZ have not advertised the service, they assume it must be word of mouth from other migrants/travellers that is driving this uptake, or the general fact that the mobile phone is now one of the most commonly used modes of communication globally.
After the initial pilot, survey follow up showed that 80 – 90% of clients said that as a result they had made fewer calls to the call centre, 99% approved of the concept and wanted Immigration to investigate spreading the service further.
Problems
What INZ did not foresee was that by sending outbound texts, they were opening up another channel of communication - inbound texts. They assumed that they wouldn't receive text messages back because the sender has to pay for each text message they send. INZ did not think that their clients would want to spend money replying via text message, however their clients had other ideas. Sending a text message might be a lot cheaper than a call for clients, especially if they have overseas cell phones.
However this inbound messaging proved problematic - INZ were receiving several hundred texts, and didn't want to use SMS as a two-way channel, because they didn't have the staff to respond to individual messages. Now they are receiving about 10 inbound messages a day.
Immigration doesn’t reply to inbound texts any more.
The team that is currently looking after the scheme is technically the research and development unit known as iBranch. Their role is to create productivity and efficiency gains, test and prove these initiatives and then transition these back into the business. The problem with the text pilot is there is no natural area of the business team to take this solution over as business-as-usual. The contact centre would seem ideal, however as is usual with call centres, there is a high staff turnover, so there would not be a stable group looking after the project. The pilot is staying with iBranch for now, but will have to eventually be passed to another group. They are scoping for a potential successor.
Communication with other branches is still an area that requires improvement. Currently there is no branch reporting, so branches don't know which of their clients have been contacted through this service. A solution to provide this information specifically as it pertains to each branch is currently under development. At the moment the default assumption is that branches assume that all clients who have their cell phone numbers in the system have been contacted by the text service, which is not ideal. All the information about who has been contacted, what and when is in the Datasquirt database, which isn't accessible to the branches, and this information is not easily searchable within the database to start off.
Evaluation
INZ definitely think their SMS service is meeting their goals - they have seen a significant drop in the number of status enquiries to the Immigration Contact Centre.
They have a customer service standard that existed prior to this scheme - they promise to reply to any inquiry within 10 days, and they perceive this scheme to fit in with that.
There was media interest when they won the TUANZ award, and they made headlines after presenting at APEC. The service has also been a success case internationally – it was the first Immigration Service to do this, so other countries are very interested.
The greatest advantage of SMS
Most of the people that Immigration New Zealand works with have cell phones, even when they might not have landlines, fixed addresses or access to computers. SMS is seen as a good way to communicate with these people.
The immediacy of SMS is seen as one of its best features. For example if the airport crisis in Thailand last year had gone on longer, there would be some people in New Zealand from Thailand on visas/permits may have expired. Immigration has the means to contact via SMS all these people affected and let them know that their situation was being taken into account, and what they should do about it.
What would INZ have done differently?
When the iBranch team starts a project, it needs an exit strategy. Research and Development teams are meant to start projects and parcel them off to another department. There needs to be an idea at the start of the project where it is going to end.
iBranch think they should have thought a bit more about the effects this scheme would have on their branches. This scheme contacts individuals, all who fall under the jurisdiction of one branch or another. There needs to be a way to let the branches know who has been contacted about what. Information sharing - keep everybody in the loop!
Also, they would rethink the relationship between Datasquirt, Gen-I and themselves. At present they are using Gen-I as a middleman, however it is not clear that this is adding any value for INZ. It's possible that dealing with Datasquirt directly may have been advantageous.
Looking forward
Spin offs
Initially this service was only offered to people with cell phones, now it is being offered through email as well.
An issue they came across was that some people use agents to help them through the immigration process. These agents don’t want to receive tens or hundreds of text messages a day/week. As a solution, Immigration developed a way of packaging information about agents’ clients and delivering that all to the agent in one go via email, giving regular reports to agents. Agents who were calling 10 times a day are now calling 10 times a week.
What is Immigration looking at doing in the future?
INZ are definitely going to investigate what other unexploited data they have, and see if they can think of a good use for it to improve customer service and lower costs.
They want to work on a way of better sharing information with their branches, and are thinking about ways they can achieve this.
