Quite simply, numbers do not matter. The greater amount of people that want to listen to things you say, does not directly correlate into a great number of people respecting your thoughts and being influenced by them.

You could have the widest distribution network for your ideas, thoughts, concepts and such. This could target thousands of people. Yet, does it matter? How many people are going to read what you said, or more accurately how many people are going to read AND care about what you said?

So don’t get hung up on the numbers, surely a more tightly knitted group of people who DO care about what you say is more important.

Think of it like a mailshot, you can send your leaflet out to the whole of Birmingham, or you could spend some more time being careful and just send out your leaflet to those people in Birmingham who care about it’s content.

Look at the outcomes instead, what has happened. Have people read your article, commented on something, send you an email. People you care about. People who care about the topic.

In summary, if you’ve got 100,000 facebook fans (and no doubt a ton of those have ‘hidden your updates’) and 20,000 Twitter followers it doesn’t mean your message is more effective than someone with 1,000 facebook fans and 300 Twitter followers.

Instead, look at the metrics around engagement on the topic, and whether your message is targeting the right groups. Numbers ARE important, metrics are extremely important, but the size of them does not necessarily reflect your influence.

These make up a smaller proportion of the data than one first might expect, but are still significant [other incidents may involve arrests, IED detection, engagements that didn't result in a KIA].

This is a map of Afghanistan, showing the regions the incidents occured (by colour), the size of the mark denotes the scale of the KIA count, whilst the shape of the mark shows who the attack was on.

Take a look at it ‘full screen’ for more ease of use, hovering over individual points will give you some details, as well as the fact you can filter by how many people were killed in the incident, and filter down to what type of incident.

Some other more detailed analysis was conducted by the Guardian (amongst others), who had pre-release access to the data. Further information on the data can be found on Wikileaks – Afgham War Diary. This visualisation was created using Tableau Desktop, and published to Tableau Public.

The presentation is available here:

It’s all a bit brief, and it’s a subject I’m passionate about, so if you are interested just email me neil@rasga.co.uk

An audio feed was also available which I’ll try and find where I saved that!

• http://flowingdata.com/2010/04/12/data-underload-17-famous-movie-quotes-p-2/ – Movie Quotes as Infographics
• Social media, BMW and the way forward | litmanlive.co.uk – Social media, BMW and the way forward
• Web design birmingham | Web site design west midlands | Branding | Interior design | Design agency birmingham | 383 Project » Such Tweet Sorrow Infographic – Birmingham agency 383 have produced an infographic of the information from the Twitter version of Romeo & Juliet, #SuchTweetSorrow
• IRE – Social-network analysis – The next step in data analysis<br />
  IRE and NICAR members are beginning to explore the uses of social network analysis for visualizing and diagramming relationships between individuals and businesses and institutions. <br />
  <br />
  This is a new approach for journalists, but other professions have utilized this approach for decades for a wide range of research projects such as analyzing:
• iA » Cosmic 140—Final Beta – This time, it’s the 140 most influential people on twitter, sorted by #name #handle #category #influence #activity. Plus: When they started tweeting and what they first said. It took quite some time until we had it in the shape we envisioned.

It took me a second to realise that I had signed up to the Birmingham Community Alert, I gave them my postcode and mobile number (through a Text), and thought nothing of it.

Thankfully, the incident was a hoax, and all is well. The service then followed up with a Text, as seen to the left, to alert me that all was fine.

Overall I’m really impressed by the service (which is opt-in), and feel comforted by the fact that in the case of disaster/alerts I can be sent information.

The points I’d think might need work thought are:

• Use a recorded ‘real voice’ instead of a ‘text to voice’ system
• Allow the person to replay the message (i.e hit # to listen again)
• What happens when the phone networks jam? (An email would be nice as well)

So overall, job well done to Birmingham Resilience, Birmingham City Council, NHS, Police and Fire services.

This week, it appears that Adrian Short, has taken the data released from Royal Borough of Windsor & Maidenhead, and produced Armchair Auditor.

This is a website that simply takes the CSV information provided by the council, pulls it into a database, and allows you to examine it by Directorate (Services), and Supplier (Company paid for delivering a service). It is quite nice and easy to use, and you can simply see that, for instance, Capita Business Services Ltd were paid 15 times, totalling £52,493 between Sept 2009 and March 2010, the average amount of each payment was £3,499.

It also highlights the fact that the Supplier Master may give you a bit more information than you currently have. As Capita are shown as:

• Capita Business Services Ltd
• Capita Education Resourcing
• Capita Education Services Ltd
• Capita Software Services
• Capita Training Ltd

So it looks like these are related businesses, indeed seperate companies, but the ultimate beneficial owner maybe ‘Capita Group PLC’. It also shows that payments were made to Capita Education Resourcing and Capita Education Services. These may actually be one company, and just different divisions within the company structure. If you had a full supplier master, it’s likely that these different suppliers will have an indicator to show that they are ‘related’.

A quick check at Companies House doesn’t actually show a result for Capita Education, nor some of the others. For instance Capita Education Resourcing is a trading name of Capita Resourcing Ltd, likewise this may be the same for Education Services.  Companies may be trading under many different names, all of which are actually present as different suppliers in the system, therefore the actual company may be receiving more funds from the council than you presume at first thought.

All of the above was found through a quick play with Armchair Auditor, it shows the strength of the tool at being able to get to the data quickly, and concisely as well as being able to direct someone to your findings with relative ease.

Whilst at the same time, we can see the challenge at understanding the data, especially when you start to look across councils and try and answer questions like ‘How much is supplier X getting paid for services bought by councils across the country’. I’ll discuss approaches for how you might deal with that situation another day.

For now, I’m going to look at the work Adrian has done, and see what else might be interesting within it.

These are just my personal thoughts, and may be incorrect, and they do not represent the thoughts of my employer.

Supplier Identification
I disagree with the phrase ‘Ideally publication would include the Companies House number (or equivalent in the case of foreign companies) or Charity registration number‘. This depends on the quality of data entered into the system, in my experience such level of detail (at least for private sector companies) is never recorded in a Supplier master. You will have the Supplier Name, Internal Supplier ID, Address fields, basic contact details (with a CRM system likely to be holding more details).

Such information is not easy to populate into a system if it’s not already present, I remember a case where we had to identify Dun&Bradstreet numbers for every supplier – it wasn’t an easy job.

I would say, this boils down to one thing – why do you require that level of information, what is wrong with the company name after all unless we are talking about data quality issues?

I’d also be wary of needing to publish the actual contact details for each company, depending how a companies supplier master is setup, you may have the contact details for the person in Accounts Payable. What use is that to the end user? It would need to be researched to understand exactly what data is present etc.

Payroll/Benefits
Dependent on how the Supplier master is setup, it is possible for employee expense reimbursements to be present (i.e every employee is setup as a vendor). Most systems allow you to identify these, not all do though. So it may be possible to anonymise these.

Payroll, at the General Ledger level is likely to be a few entries (the transactions hitting the GL Payroll account, and the bank account), the Payroll data of actual payments (BACS likely) may be held separately to that of where normal supplier payments are made (again, dependent on system structure – I’ve seen them a few ways).

Overall
The other thing to remember, is that in some cases, the data output is likely to be:

01/01/2010
CHQ
£2256.89
A1 Company
INVXYZ002 (or some document reference)
GL CODE
GL Description (i.e Transportation)
Cost Centre Code
Cost Centre (Parks Department).

The above shows that the Parks department paid A1 Company, £2256.89 on -01/01/2010, for a service that the invoice was booked to transportation, and the cost absorbed through the parks department budget.

You’d be surprised how many times there are some systems where it’s not totally easily to identify the payment, back to the relevant invoice (apart from a manual reconciliation), you need to know the invoice side of the transactions – as that is where the cost will be booked to (as the payment details will just be crediting cash, debiting Accounts Payable).

It all depends what level of detail you want to show, it’s easy enough to show a payment went out (but associated services, dependent on the internal reporting functions of the system – may be harder).

These are just my thoughts, from my experiences of working with a multitude of different financial systems from your SAP, to your BAAN and SunSystem.

• 5 Interesting Data Articles | blog.infochimps.org – Inspired by Pete Warden’s Five Short Links, we decided we’d put up a post about the most interesting data articles we’ve come across in the recent months.
• 30Days to context connection – PDF – A 30 day calendar of useful visualisation, and other learning aides.
• The Evolution of Privacy on Facebook – A nice segmented chart showing how privacy has chnaged over time on Facebook.
• Social circle and content – Google – Google Social Search / Circle
• Nicholas Felton’s Master Class – Nicholas Felton's Master Class – Visual notes, interesting insights on how the Annual reports evolved and onto how to create your own data visualisations.

• Google – General Elections 2010 – From Google, some good coverage of UK websites (Openly Local etc). Britain is going to the polls, and the internet is a great resource to help you make your decision. Quickly find out your MP and candidates below, with links to their blogs and Twitter pages. Then explore further, with fun and interesting election sites from around the web.
• Magazine Preview – The Data-Driven Life – NYTimes.com – An interesting look at people who track their life in great detail.
• About – trst.me – trst.me measures user reputation in a way far more robust than counting the number of followers. The basic idea is to look at how many people interact with you and give you their attention, weighted by how many people interact and pay attention to them
• Eddy – Stame Design – Eddy is a media aggregation platform built for the public display of up-to-the-minute activity on realtime services like Twitter. It was used by Stamen to curate the NBC Olympic coverage, pretty interesting.
• What is Twitter, a Social Network or a News Media? – WWW’10 – Research into Twitter, based off crawling the user network and looking at relationships. Touching on the thoughts of influence.

• NY Times: Driving Shifts Into Reverse – An interesting visual about driving usuage/mileage, the kicker being the X axis is not displaying time. <br />
  <br />
  Works surprisingly well
• 960 Grid System – The 960 Grid System is an effort to streamline web development workflow by providing commonly used dimensions, based on a width of 960 pixels. There are two variants: 12 and 16 columns, which can be used separately or in tandem.
• TweetCatcha v 1.0 – TweetCatcha seeks to uncover the organic nature of news as it travels through Twitter over time, by examining the movement of NY Times articles through Twitter.
• Create cool applications! | dev.twitter.com – New application/dev area for twitter
• Follow Finder by Google – Follow Finder analyzes public social graph information (following and follower lists) on Twitter to find people you might want to follow.