Shambles

Team Name: 
RubbleDusters

Purpose:

To help people determine whether a rental property is safe and affordable so that they can make better rental decisions. Shambles will help fill the gap between renters and the real information about the property they are renting.

People (target market):

We are currently focusing on ‘Students in Wellington' and comparing Wellington data with Christchurch.

The reasons we have selected students as our (first) target market are:

1.      They are frequently moving and therefore have a high demand for this information

2.      They are often moving to a different town which they don’t know well and therefore they need more visibility

Problem:

Some rental information is available but is scattered e.g. earthquake-stickered properties, some information which renters would be looking for is not available until you visit a property e.g. cold or dark, and still more information is not available until you have been in a property for some time e.g. the helpfulness of the landlord, damp conditions or noisy neighbors.

Through a qualitative survey, our own experience and anecdotal stories of target customers, we have heard about the lack of transparency on the safety and affordability of rental properties at a local and national level.

Feedback from customers

From a short survey which we set up to get a feel for the customers’ needs, we identified key items of information which they thought were most important when looking at a new rental.  They were:

  1. Price – covered below
  2. Close to school, university or work – see ‘future scope’
  3. Cold, damp or dark – we have linked to a heat map of sunshine hours in the capital
  4. Safety and crime rates – covered below
  5. Landlords – see future scope
  6. Leasing types – we have linked to useful information for tenants

Solution:

‘Shambles’ is a website application which collates the information renters want to know when they are renting a property.  The website helps by visualizing various types of data such as rental rates/crime rates by suburb as well as specific information for your address.  The data is mainly displayed in the form graphs (Bar charts, pie charts) and a property-search feature.

Within the application there are two main levels of information – geographical comparison within a region / city and information that relates to an individual property. 

The following data is displayed at regional level:

1.      Rental rates – Region-based data for Wellington, Christchurch and Nationwide. This data is further broken down into:

         a. Suburbs

         b. Type of rental e.g. apartment, 2-bed, 4-bed etc.

         c. Time - patterns over the last decade, the current year and the latest rate

2.      Crime rates – Region-based data for crime rate in Wellington, Christchurch and Nationwide. Crime rate data is subdivided into categories such as burglary and assault (violent crime).

This information is displayed at individual property level:

1.     Earthquake Data – what level of ‘sticker’ this property has.

Additional items

Individual customers also asked for:

  • Internet access – what type of internet would be available at the property

Solution Development:

Technical solution

The website is available at http://jenkins.assurity.co.nz:9090/.

The problem statement and solution was in the Rubble Dusters’ mind from the day the Govhack event was announced.  The actual work required at the event was to implement the idea into a real solution. Here are the details of how Rubble Dusters have built the solution:

Technology Used:

1.          Python Programming Language

2.          Pandas: http://pandas.pydata.org/

3.          Here maps – https://developer.here.com/  

Shambles website is hosted on the cloud utilizing some of the technologies available.  Data is pulled from various data sources as mentioned in the hackerspace project definition.  The data is processed, analyzed and coded into a structure and displayed in the form of graphs.  For further information, see the architectural diagram attached to this project definition in the Hackerspace

Sustainability: Currently the data is sourced from different government data sources e.g. Stats, Police and others.  It is not known how often these sources are updated and whether the format will remain the same.  It is expected that in a future solution a batch process will be scheduled to fetch the data from these data sources so that the data is kept up-to-date.

Team of Teams:

To deliver a valuable piece of information at Govhack, 11 core professionals have worked in 4 mini teams to achieve this solution with great quality.  Teams were:

1.      Analytical Amateurs: Team of people dedicated to doing the analysis of the data sources available from the NZ government

2.      Competent Coders: Team of high end developers to produce a solution which will help solve the decision dilemmas of customers

3.      Media Masters: Team of creative experts producing video and constantly socializing on various social medias such as Twitter, Facebook and LinkedIn

4.      Project Leader: Primarily managing and coordinating with rest of the team to maintain the easy flow of information. Also, working on the project ‘story’ and overall pitch. 

Team has applied Agile skills led by an Agile coach to produce a solution using multiple short, hour-long iterations.  The team was self-managed, meaning no-one told anyone what to do but teams took responsibility for items on the list of ideas, broke down the work into individual tasks and worked on them.  We had a standup (short coordination meeting) every hour to discuss the progress and help each other with any impediments.

 

Unique Value Proposition:

There is no one page where renters or landlords can find out about the overall safety and affordability of their rental property and the surrounding street or suburb.

Unfair Advantage:

This is the kind of project which is unlikely to be funded by private enterprise.  The target customers are unlikely to pay for this service as they are low on spare cash and landlords are unlikely to pay for listing their property - particularly when their property is not a good one and these are the exact properties the customers want to know about on the site.  Consequently, the development of a solution like this is only going to happen through government-sponsored initiatives like GovHack.

In terms of this particular team, we are a group of people who are passionate about helping students and other renters because several of us have had bad experiences and want to stop others from having the same problems in the future.

 

Impact:

As described above, Shambles will initially benefit people who are looking to rent a property in the Wellington region.  Simply collating and giving access to the information in one place means that this saves the customers significant time & effort trawling through the internet (which was a complaint of some of our sample customers).  Also, many students do not know about some aspects of renting which they need to know when signing a rental agreement.  Having access to these through the site helps them to become aware that they should know about these areas and explore them.

In the future we hope to raise the standard of rental properties in the selected regions by having means to compare and give feedback on them – therefore forcing owners who have poor quality rentals to do something about their properties in order to continue getting tenants.

Future Scope:

As expected in a hacker-weekend exercise, the actual product is a prototype – we can only do what we can do in the time and there are so many more ideas to include!  Shambles offers visualizations based on static input limited to Wellington and Christchurch. We would expect that the target market would be expanded to include:

  • Location – expanded to other cities and potentially other countries.
  • User Types:
    • Working individuals
    • Low income individuals and families
    • Immigrants
    • Individuals and families with special needs
    • Elderly individuals and families

In terms of features, these are a few of the ideas which could be added:

  • Distance from… - University, workplaces, schools and other amenities.  Currently available on Google.
  • Property condition – the state of the walls, roof etc. is recorded by Wellington City Council, https://homes.co.nz/app/#/homes
  • Property WOF (Warrant of Fitness) – this is recorded by the Dunedin City Council.
  • Tenancy tribunal history – if one landlord or one property has been taken to tribunal may times then this will be visible
  • Meth labs - has your rental ever been meth lab contaminated?
  • Neighborhood – how do other people in the area describe/rate living in the area
  • Other rights and grants:
  • Indidivual property data - Pets allowed, Heating such as central heating or heat pumps, floor heating, modern kitchen
  • Garaging and off street parking - e.g. council zones for 'residents parking'
  • Feedback system – ability to leave notes/comments
  • Crowd-sourced ratings – individuals would be able to rate a property based on criteria like:
    • Cold, damp or dark
    • Landlord/rental agency helpfulness
    • Neighborhood, suburb, street

Problems with the site which we didn't have time for:

  • Collating into one tab per location - i.e. Wellington tab would include rental/crime graphs and a search specific to that city.
  • Searching for your rental would bring up the relevant data dynamically - we have included only a fixed subset, indicating what would be there (using the data sets we have used for the graphs), i.e. average rent, crime rating and internet available.

Promotion:

We expect that the site will be promoted in the future through word of mouth (the student world is very good at spreading memes), social media and possibly local government support.

Prizes - why is this solution relenvant to these categories?:

Best innovative hack utilising Statistics NZ data – we have taken data on rental prices in different locations and in different types of accommodation and graphed it so that renters can see how their proposed rental rates against the local average for the city or suburb.  This is presented alongside NZ Police data on crime statistics and local council information on earthquake ratings to give an overall picture of the safety and affordability of their new flat.

How can we better understand how councils are performing – the Shambles project is about collating information on the quality of rental properties in order to help renters to make informed decisions about where to rent.  Collated views of this information, using a rating system which combined a number of the features mentioned in this description (current and future) e.g. crime rate, average prices, property condition, number of tribunal cases, and crowd-sourced feedback & ratings, would allow a council to assess relative effectiveness of their policies and target specific areas of the city to improve.

International prize – community resilience hack – the Shambles project is about collating information on rental properties so that we have a comprehensive view of the state of rentals in a city.  The aim is to help the renter make good decisions about choosing their rental but it could also provide an overall picture of the ‘health’ of a city in terms of the number & quality of rental property available.  This will encourage individual landlords, rental agencies and local councils to improve the quality of their property or the local street/suburb to match those surrounding them.  Improved quality of housing has been proven to have many benefits to the people living there and the community as a whole.

International prize – storytelling hack –  Ever wanted to get the a feel for a new city or suburb you are moving to?  Shambles collates a number of different data sets to give you a picture of the safety and affordability of the area - crime rates, average rentals and earthquake ratings.  This way you get to see a little of the character of the area, using actual data, to help you choose where to rent.

NZ best data journalism - the Shambles project is about collating information on rental properties so that we have a comprehensive view of the state of rentals in a location.  The aim is to help the renter make good decisions about choosing their rental but it could also provide an overall picture of the ‘health’ of a city in terms of the number & quality of rental properties available.  The data is collated into one view of many different sources of information including average rental rates, crime rates and earthquake ratings.  See the ‘future scope’ section for further information which could be collated to give the overall view for the customer.

NZ best open government – Shambles has taken earthquake rating information for buildings in Wellington and Christchurch and made it machine readable.  It is currently held in .pdf format and we have made it into a text file which we then made searchable by address.  On the Shambles website you can enter an address and it will tell you whether that building is in the list of ‘stickered’ buildings or not.  Also, the intention is to create a crowd-sourced feedback mechanism on rental properties so it will be possible to see whether the crowd-sourced information matches to that collected by the government.

Suggestions for Improvaement with Data - While much of the data was already in a solid normalized format we did find a number of holes which made it difficult to plot the trends we were after. for instance the list of at risk buildings in Wellington from Earthquake damage only appeared to be found in a PDF file - which increases the work necessary to work with and gained insights from the data set. Additionally many datasets are only highly visible at a regional level. This made it easy to make comparisons between auckland and wellington and christchurch if we wanted but made comparing information between suburbs with respect to rental and safety information difficult. It took a good deal of searching for instance to find the dataset that corresponded with NZ Burglaries per suburb or per mesh point. Visibility and tracability of student loan data was difficult to find and ultimately we were unable to use this in our end result - moving forward, using students as our target market means that being able to show how much rent is in an area, relative to what they may recieve via student loans and allowances would be greatly helpful.

Video Upload: 
Region: 
Used Datasets: 
Dataset Name: 
Mean rents by TA [CSV 85KB], Mean rents by region [CSV 23KB],Detailed mean rents [CSV 7.1MB]
Publishing Organisation/Agency: 
MBIE
Jurisdiction of Data: 
Local Government
How did you use this data in your entry?: 
To display the rental information about Wellington and Christchurch property
Dataset Name: 
burglaries_by_mb.csv
Publishing Organisation/Agency: 
Statistics New Zealand
Jurisdiction of Data: 
Local Government
How did you use this data in your entry?: 
To display the burglary data information at an area unit level primarily for Wellington and Christchurch suburbs
Dataset Name: 
Area Units 2006 Dataset
Publishing Organisation/Agency: 
Stats.NZ
Jurisdiction of Data: 
Local Government
How did you use this data in your entry?: 
We used this data to help map our rental data which had a detailed view of suburbs and used SAU codes as a reference point. Moving forward these codes could be correlated with the suburbs across NZ to better map existing rental information
Event Location: 
Wellington