Google’s prediction: What will be the “sexy” job in the next ten years?
Here’s a strange prediction from Google’s Chief Economist: “I keep saying that the sexy job in the next 10 years will be statisticians. And I’m not kidding.”
That quote came from a recent New York Times article, which is about the rapidly increasing demand for statisticians.
He’s talking about you, conversion fans. You may not think of yourself as being a statistician—you may not have it written on your business card—but if you’re basing your marketing on data, not on whims, you are already one of the new wave.
We may be in the minority now, but someday all marketing will be carried out this way.
Another article, published by Wired magazine earlier this year, describes how Google’s business models have evolved. Here’s a great quote from the Wired article:
“Hal Varian [Google's Chief Economist] believes that a new era is dawning for what you might call the datarati—and it’s all about harnessing supply and demand. “What’s ubiquitous and cheap?” Varian asks. “Data.” And what is scarce? The analytic ability to utilize that data. As a result, he believes that the kind of technical person who once would have wound up working for a hedge fund on Wall Street will now work at a firm whose business hinges on making smart, daring choices—decisions based on surprising results gleaned from algorithmic spelunking and executed with the confidence that comes from really doing the math.”
Hal is a fantastically insightful guy—and, of course, he has the benefit of sitting at the cutting edge of innovation. So his advice is worth taking.
Here are a few things that any web business can do now to benefit from this approach:
- Start making decisions based on data, not on opinion.
- Get analytics set up well, so you really know which pages, products, acquisition sources, etc, your money is coming from. Knowledge really is power (and profit).
- Start carrying out split tests (A/B tests or multivariate tests) on different aspects of your business.
- Create a culture in which people are rewarded for carrying out tests, and not punished for making mistakes.
This might sound like a lot of work. In fact, it’s really liberating. Imagine being able to end long debates with “let’s test it and see who’s right”. Imagine being able to make business decisions based on insight that your competitors don’t have. And imagine never taking a step wrong, because every decision you make is tested, so you almost-immediately know whether it was the right thing to do or not.
So, that’s today’s word of the day: datarati.
If you missed it when it came out, the Wired magazine article is particularly worth reading: it’s here.
Rate
(rated 4.80 by 25 people) 
Loading ...
Great post, I can’t wait to see the forthcoming breakthroughs regarding online customer intelligence.
Hal Varian said “The analytic ability to utilize that data”, in my opinion if we could also add a little touch of creativity we shouldn’t be far from conversion rate heaven :)
You know what, I had certainly never though of myself as a ’statistician’ but you are so right. We look at data, we analyse trends and we work out, through probabilities, what an outcome will be…
Excellent article (as always).
Ok Makes sense, but where is Google in all of this? Did it really take 18 months of data to figure out how to make YouTube profitable? Why are even my most tech friendly friends unaware of Google’s voice triggered search app on the iPhone? Why does Google still hide some “other” search query results in the AdWords platform and leaves them out completely in Google Analytics?
The pied piper of the “datarati” needs to get its own house in order or they will find themselves becoming increasingly irrelevant as the datarati find other platforms to work with.
Great article. We’ve been following you since your first article on Multivariate / AB testing. I think you’ve hit the nail on the head with the scarcity in analytic ability to utilize data. Thanks for sharing!
I have learned quite a lot on my own about analytics, conversions, seo and all that we have to learn and do when we have an e-commerce site. I am fascinated with all this stuff. I love doing it.
Here is the question: Where could I get some sort of certification so that perhaps I can start a career as a analytics/statisticians consultant? I really appreciate your advise.
I truly believe these guys are at the forefront of the next generation of SEO. So much so I applied for a job to work for them. (Hint Hint) But seriously, no matter how many hits you get on your site if your not converting them into paying clients what does it matter?
Very true. The idea of accountable marketing is only going to become more mainstream in the next few years.
It’s incredible how few companies are really harnessing the power of data. Even when so many are signed up to Google Analytics.
It’s also unbelievable that the testing and tracking techniques that form the basis of Google Website Optimizer have been around for around 100 years!
Just take a look at Claude Hopkin’s book Scientific Advertising. It’s like it was written in the last few years.
Agree wholeheartedly but, dangerously, it doesn’t go far enough.
The real “money” skill is knowing what statistical information really matters in a business. The person who knows this can easily find a statistician to do the testing and provide the relevant data on which the important decisions can be based.
Very clever statisticians, mathematicians and programmers can be hired on elance etc for a few dollars an hour, all of them very good at what they do but none of them aware of the commercial application of what they produce and how to turn it into commercial value.
Encouraging people to become statistians in the sense of being a number cruncher simply destines them to the cattle market of the jobs boards and other labour pool websites.
I installed Google Analytics about 1.5 years ago. I have already run several tests with Google Website Optimizer and made website changes based on those test, so in a couple of years I should be an expert.
A grasp of statistics always has been empowering, even if simply as self defence against headlines such as “64% of Americans say they would do X”.
My question would be, given the poor levels of basic numeracy in the Western world, where are these ‘datarati’ to come from?
Perhaps now is the moment for those wonderful people who gave us Zero to step in?
Fascinating post. My guess is China will be where the statisticians come from. The tools are already evolving now more people are needed to properly use them and interpret the data.
If you take a look at Salesforce.com’s AppExchange as one example, there are more and more analytics / number crunching systems being built for businesses that tie into existing systems than there have been in the past.
Take a look here for some examples.
Just a side note: the accompanying image is an example of what we do wrong today in data analysis and visualization. If we want to make sense of our data, removing 3D charts from our toolbox is the first step.
Nice one, keep these articles coming. The future’s bright ;-)
It’s great to see all of you guys interested in statistics. It’s probably the number one skill that should be taught more in school. (Probability of a teenager putting down the Xbox for a Stat book: 5%)
If you guys are really interested in statistics, you might be interested in this open source project.
http://www.r-project.org
Google and a lot of other analytics focused companies use this software internally (but not on production level). It’s an insanely powerful program although it has a steep learning curve. If you want some resources for it, feel free to contact as I have some of GPL ebooks that have been provided by the community.
To build more on what Chris said, the real skill in statistics is knowing how to set up tests, what data is statistically valid and how to leverage the data. If you give a statistician a data set of numbers, they can make it say anything you want. People think it’s a completely objective subject, but it also requires self control to keep yourself honest.
A good example is knowing why Taguchi testing can lead you astray in testing landing pages. Taguchi is effective in cases where there is not synergy between elements. It works great in engineering, but can perform quite horribly in marketing data.
Best of luck to you all. You are all ahead of the game!
Dave
Interesting and insightful article. Wish you guys would send more of this to my email. :)
Great and so true!
It is true that we log more and more data from more and more sources. This quite often leads to false interpretations and statistics that do not match. IMHO there is need for decent understanding of the data and its various sources, but no real statistician is needed for analyzing a website.
I can do without them – I have a horde of Elvs that analyze the millions of letters I receive from kids around the world every year – and the results are always well received :) Ho Ho Ho
>>Interesting and insightful article. Wish you guys would send more of this to my email<<
I copy on that – the articles you post (when you post) are always helpful and add a significant value to understand the inner workings of my business. Please, don’t leave us in the dark – bring more interesting posts like this one to your blog. Thx
Hello there conversion rate squirrels!
What advice do you have for a university graduate looking to get into this field?
I’m a fresh grad that’s just started a career in web analytics. I’m a junior web analyst at one of Australia’s largest online publishers. I have about 1-2 years of SEO/PPC experience under my belt, but a specialisation in website/conversion rate optimisation I’ve always wanted to pursue.
You guys might also want to join over in the conversation over here. Eagerly awaiting your thoughts. Thank you.
I think that’s right. All Harvard Students are also taught to work with studies of data and analytics and they get taught to work with this data. And the most growing business on the stock are companies with Harvard. But don’t tell everybody. ;)
I think analytics and data modeling technology has evolved massively over the last decade. But being able to align all this fantastic technology with business objectives, and being able spot insightful trends under mounds of information – without succumbing to ‘analysis paralysis’ – is a skillset which I agree will become increasingly sought after.
What Dave said earlier about the R project was right on…it’s a great free stats software that can do what you want (unlike SPSS and SAS), given you are willing to take the time to learn it.
On another note, it is very important to realize that giving the data to some cheap stats laborers will get you what you pay for…analysis independent of its vital context. Stats is all about the big picture, and even if a person is great at the details in the number crunching, if he/she doesn’t keep the big picture in mind, you are not going to get what you need from the analysis. That’s why it seems like there is a gap between the stats nerds who do number crunching on oDesk for a living and those who work in research using stats as a leveraging tool…and you will rarely see the two parties meet in the middle.
Stats is powerful, but with great power comes great responsibility! Use it carefully and wisely and it will open doors to the future…use it carelessly, and you will see why stats can seem more like an experimental science than an exact math…
Although I worked as a number cruncher myself and this experience helps in running by small ‘business’ on the web, I wouldn’t overrate the analysis part.
Making decisions purely on based on data can also go into the wrong direction – I query my users quite often and adjust my content to the feedback I receive (if possible).
I found this article very interesting and insightful. My wife recently received her PhD in Social Science and was required to take many graduate-level statistics classes. Often times during our conversations, we find correlations between how she analyzes research data and how I analyzes Web user data.
For those of us that have gone to the trouble of calculating the lifetime value of a customer there is no going back. I have been making my living in the direct marketing discipline for over 15 years, and I have never worried about getting a job – ESPECIALLY in hard economic times. When you think about it, lifetime value is a proxy for stock price. Wall Street folks would love to get the numbers we direct marketers get to play with.