The other week I gave a talk on Gen Y, Gen X, Technology and the Future of the Public Service at StatCan’s managers’ meeting. The speaker before me apparently told the gathering that they “should be more like Google” if they want to recruit young talent. During his Q&A one of the managers asked how a government agency could be like Google (a legitimate question, I thought) and the speaker didn’t have much to say. Frustrating, no?
Definitely.
But I think there is a good case. While the idea of StatCan emulating one of the best performing, young, hottest companies in Silicon Valley may sound far-fetched, it needn’t. StatCan can be like Google. In fact, it already is.
Look, for a second, a Google’s strategy. Google’s mission is encapsulated in its SEC filing statement:
“to organize the world’s information …. and make it universally accessible and useful”. Google explains that it believes that the most effective, and ultimately the most profitable, way to accomplish our mission is to put the needs of our users first. Offering a high-quality user experience has led to strong word-of-mouth promotion and strong traffic growth. Putting users first is reflected in three key commitments illustrated in the Google SEC filing: “1. We will do our best to provide the most relevant and useful search results possible, independent of financial incentives. Our search results will be objective and we will not accept payment for inclusion or ranking in them.
- We will do our best to provide the most relevant and useful advertising. Advertisements should not be an annoying interruption. If any element on a search result page is influenced by payment to us, we will make it clear to our users.
- We will never stop working to improve our user experience, our search technology and other important areas of information organization”.
To organize the world’s information… and make it universally accessible. This a huge part of StatCan’s mission. To organize Canada’s information… (now if only we made it universally accessible).
I think Google’s mission is similar to StatCan’s. Indeed the main difference is that StatCan not only organizes Canada’s information; it also creates that data. However, this is a space that Google has moved aggressively into — why do you think it has created platforms like Google Earth? To facilitate the creation of data so that it has more to organize and offer its users. Indeed, what is interesting about Google is that it knows the more information and data that is out there – for free – the more useful and important it becomes. It means more people doing searches, which means more advertising revenue.
So what does this mean for StatCan?:
First, distinguish and separate what you do: “Creating and organizing information about Canada” from what makes you valuable: making this information universally available to citizens.
Second, make yourself the centre of a data gathering, sharing and analyzing eco-system: There are thousands, if not hundreds of thousands of people out there who could do amazing things with StatCan’s data. The problem is, it isn’t easy to find, often you have to pay for it, and it is usually only available in HTML charts that aren’t easily accessed, and certainly not dynamically available. If StatCan data was available as API’s and Excel spreadsheets, then a whole ecosystem of multimillion dollar businesses, bloggers and other pro-ams would emerge around it as supporters, collaborators and complementors.
Finally, hire young people to make it happen because if you are open, they will come: Does StatCan want young people to come work for them? Then stop behaving like a 20th-century consulting firm whose job is to hoard data and conduct analyses for clients. (Don’t worry you can still do this). Instead, act like the 21st-century Google-like platform that you are. Your job should be to make your data as searchable, taggable, as pluggable, in short, as usable, as possible. This, in addition to collection, should be the top priority. If StatCan’s data were easily available (say as an API) people would start using it in all sorts of creative ways – this, and this alone will drive innovation, excitement, energy and buzz about Statcan into the workplace. In short, it will make Statcan relevant. StatCan should be a place where young Canadians want to work so they can learn how to handle and disseminate HUGE quantities of data to everyone from the smallest bloggers to the largest companies. That skill set is going to be critical in the 21st century and so such a mission will attract talent top talent, if StatCan gives them the freedom to play and build it.
StatCan is like Google — if it chooses to be. It can’t offer the stock options, but it can offer a cool opportunity to help build the country’s most critical data ecosystem for a 21st century economy. That’s a job lots of geeks would be interested in.
eaves.ca 
Can you pass along this message to Elections Canada? Alice over at Pundits' Guide is doing good work with the data they release, but it's usually very old. I think she's working with party donations from the 90s right now.http://www.punditsguide.ca/index.php
This is remarkable David. Not only valid for StatsCan, but any major organization that deals with data at such a scale. Thanks for sharing your thoughts.
Totally agree David, I just wrapped up an Urban Studies course and had to use StatsCan data quite a bit. So much of it is tucked away in seemingly arcane ways, it's a wonder we see StatsCan-sourced information at all. I remember hearing though that this isn't a StatsCan issue, but rather a problem common at most national statistics agencies. If so, is it a culture thing, a security thing? Maybe it's just another case of not keeping up with how information has evolved. If I were at StatsCan I would contact google directly much like the UN did to help unlock their hidden treasures of information.
Pingback: Eaves.ca: Why StatCan is (or could be) Google : Remarkk!
Fantastic post and a great opportunity to share your ideas in a place well-suited to them! In my post, I wanted to add a thought about the culture of many publicly funded agencies that create content.Many of these agencies, under pressure to look for alternative revenue streams, learned the wrong lessons from the private sector. They look to monetize content instead of creating and enabling public value from the content they are mandated to create. Meanwhile, there is a massive shift away from direct content monetization to other forms of value creation in content industries. This shift in orientation is a difficult one for private enterprise to make, but should be easier for public agencies like StatCan.Again, great post!
Just noticed this post now, thanks to Brenton's link.Thank you for championing open government data. As taxpayers and citizens, we subsidize the costs of its collection, but are often unable to afford the results (try and buy census data on every measure down to the dissemination area, and you'll see the cost puts it outside the realm of any but commercial users).
Ya, I've gotta say that the staff time alone involved in my last purchase of information (that should have been given to me for free), was probably well more than what I paid for it. Would have been a cost savings if they had just published the data to the website and allowed me to download it every few months.
Ya, I've gotta say that the staff time alone involved in my last purchase of information (that should have been given to me for free), was probably well more than what I paid for it. Would have been a cost savings if they had just published the data to the website and allowed me to download it every few months.
Pingback: NRCan is the Google of the Government | blog.GC20.ca
I worked for Statistics Canada for 11 years until 1986 – during that time I continually argued, as did others that Statscan should provide its data free to citizens, since they paid for it in the first place. The amount of money Statscan reaps by selling the data is miniscule compared to the social and economic benefits free access could create. The fight obviously continues. Love to be part of it.
I recently found some data that I wanted for a project (examaning MPs expenses) in PDF format, only to find the pdf to be locked against copying so that one is unable to even cut and paste to a spread sheet. Seems that even when data is published by government they want to make it as difficult as possible to produce reports and comparisons from it!
Yikes! New lows achieved. That really is punitive… Of course maybe this is part of the stimulus package – keyboard ready make work projects in data entry.
Another former StatCan employee here… I like the idea of making StatCan's data free. Were you inspired by this post?: http://shockminuscontrol.blogspot.com/2008/12/m… However: From a client perspective, “Become more like Google” actually means “display targetted advertising next to information”. Government agencies generally can't promote one product over another. On the other hand, from a hiring perspective, “Become more like Google” means offer the fun, perks, coolness, and incredibly hard work expected in a high-tech company. People do join StatCan because of its great reputation, but also for the stability and work-life balance of a civil service job. It's a hierarchical, unionized public sector environment and just isn't going to have the same culture as high tech. But StatCan's data quality does stem from the staff's high regard for tradition.
Hi Alana, thank you for the link – could agree with its sentiment more. This post was actually inspired by a talk I gave a on November 26th to StatsCan at their managers retreat.
Statscan is beyond useless!It's cheaper for me to do my data collection myself…heading downtown to count buildings :S
Fantastic post! HomeZilla has integrated some of the StatsCan data into our site and it easy but not cheap. Like many Canadian companies, they have done some great technical work but don't do a good job marketing or make it available.
Sandy – I've just joined Homezilla and it looks great and love the integration of StatsCan data into the site. Persuading StatsCan to change will be difficult – not because they don't agree (there is a younger cohort that definitely gets it) but because they are trapped by a cost-recovery model. Insiders tell me that the cost recovery doesn't actually generate much revenue (especially once the costs of billing and following up with clients is factored). I'm willing to bet anything that for accounting purposes, the revenue appears in the StatsCan books separately from the costs of tracking and billing clients. This makes data appear like a big cash generator, even though, once recovery costs are factored in, the actual profit is minimal.Sadly, it is the public that loses out – as only large companies can afford to use the data. In the end our tax dollars subsidize the data McDonald's uses to figure out where exactly to place its next franchise, but I'll know next to nothing about the stats of my own neighborhood. Sigh.
They look to monetize content instead of creating and enabling public value from the content they are mandated to create. Meanwhile, there is a massive shift away from direct content monetization to other forms of value creation in content industrieshttp://staffingpower.com/
1. Who are the philistines at StatCan that won't let their data free? I'm talking names here. Let's start a campaign to oust them.2. You can get much deeper into StatCan via university access. Sign up for one course. Get StatCan access. Drop the course (or pay the $300 tuition). 3. One major problem is the Real Estate Board of Canada. They will never let that housing data go (without a lawsuit). 4. StatCan should at least go with an Amazon type recommendation system. Their data is hard to find. Would be better to have stuff like “top data sets”, or “people that downloaded this data set also downloaded these”. They should also recommend fields in a dataset, as, for example, their census has over 2000 fields, many very similar. It'd be nice to download the top 20 most “useful” fields, just to get started with the data.
1. Who are the philistines at StatCan that won't let their data free? I'm talking names here. Let's start a campaign to oust them.2. You can get much deeper into StatCan via university access. Sign up for one course. Get StatCan access. Drop the course (or pay the $300 tuition). 3. One major problem is the Real Estate Board of Canada. They will never let that housing data go (without a lawsuit). 4. StatCan should at least go with an Amazon type recommendation system. Their data is hard to find. Would be better to have stuff like “top data sets”, or “people that downloaded this data set also downloaded these”. They should also recommend fields in a dataset, as, for example, their census has over 2000 fields, many very similar. It'd be nice to download the top 20 most “useful” fields, just to get started with the data.
Pingback: Canada launches data.gc.ca – what works and what is broken | eaves.ca