oConomy
August 5, 2013 by Shoshana Deutschkron

What is a long-tail market?

Long-tail markets fascinate me. The long tail, a term popularized by Chris Anderson in this Wired article, refers to a market in which there’s the opportunity to sell a large number of things with relatively small quantities of each, typically by widening the pool of potential buyers beyond local demand. The Internet is the main driving force of long-tail markets, as suppliers of these more unique items are now able to tap into buyers via online marketplaces (aided by the fact that they no longer have to maintain the costs of physical spaces for their goods or services).

In his article, Anderson used the now classic example of how Amazon opened up the long-tail book market, discussing how the Internet supports a market for previously limited-run titles by suddenly finding them an audience.

Today, we’ve seen the Amazon example played out in thousands of new long-tail markets — the Internet has thrown open the doors to distributed global demand so that suppliers of goods and services can find demand beyond traditional local markets. There’s Netflix for movies, Zappos for shoes, Etsy for crafts (where I bought a custom bandana from Hungary last month), and the list goes on. However, most of the common examples of long-tail markets refer to consumer goods. We need to evolve our thinking as adoption of Internet marketplaces matures.

Is a similar long-tail market emerging in the labor market?

What if the long tail phenomenon spread to a more advanced market, one that would bring a broader economic impact? In other words, what if it spread to the world of work? As oDesk approached $1 billion cumulative spent hiring via our site (a landmark we announced today), I wondered if the long tail was taking hold in our Internet-based marketplace for work (also known as an “online workplace”). With the Internet opening up access to businesses and professionals worldwide, I hypothesized that plotting oDesk skills data would show a long tail.

You can guess what we found. Before going into the methodology, I’ll jump right to the result, which was the following “oDesk Skills Long Tail Chart ”:

LongtailSkillsChart

How did we create this chart and what does it mean?

The oDesk Skills Long Tail Chart is based on data from oDesk’s own database — specifically the skills that were listed in jobs posted on the site during the month of May 2013. This chart is therefore a snapshot of a month on oDesk.

The axes for our long tail chart are:

• X axis = “rank” (in this case, rank based on how many times each particular skill was mentioned in job posts)
• Y axis = “count” (in this case, total number of jobs posted with this skill listed in them)

As an example of rank and count on oDesk, suppose we had just three skills. This is a mock example of how it could work, with count determining rank:

• Writing: Count = 100, Rank #1
• Programming: Count = 50, Rank #2
• Design: Count = 25, Rank #3

The resulting curve shows how concentrated (or not) our market is. The area under the curve approximates the volume of work (in our case, this is approximate because many job posts list multiple skills). For oDesk’s market, the fact that there is substantial area beneath the fairly low-ranked skills is indicative of a long-tail phenomena — our market fulfills substantial demand for these specialized skills.

How do we really know a long-tail work market is emerging on oDesk?

This chart only reflects one month’s data so it’s fair to question, but the pattern held true for every slice and dice of our market data we looked at over the past year. And looking at the chart itself, you can see the classic pattern of more “mass market” offerings to the top left (like website development, SEO, and creative writing). As you continue down the tail, you see increasingly specialized skills (like civil engineering, econometrics, and legal services).

Rounding out our investigation, we looked at two other things:

  1. Skills concentration over time: In 2007, just 4 “mass-market” skills represented 90% of work done on oDesk. Last year, 35 skills accounted for the same percentage of work on oDesk and 41 other specialized skills were emerging rapidly.

  2. The value of specialized skills: When we looked at the list of top-paying skills on oDesk, we saw that demand for specialized skills has an added bonus — they dominated the list of highest-paid professionals. Examples of these top skills include legal services, a myriad of specialized programming languages and approaches (like Cocoa and rspec), cloud server setup skills, and even startup consulting.

What does this mean for the job market?

Now that online workplaces exist, there’s more opportunity than ever for businesses to find specialists, and vice versa. The long-tail work market on oDesk benefits both sides:

  • Businesses can find the specific skills they need in order to be more productive. This is especially the case for startups. Traditionally, companies have only been able to justify hiring someone if the team continuously needs at least 40 hours a week of work. Among early-stage businesses, these full-time roles can be difficult to justify, yet the need for a skill may be acute. Now, they can find workers with the skills they need, and fill roles flexibly by hiring online (using the cloud for work just as they already do for server space).

  • Meanwhile, specialized workers can pursue their expertise with increasing confidence that they’ll be able to tap into ample global demand, banishing local market limitations. Take oDesk freelancer Stan Smith, for example, who specializes in developing algorithms for clients. Stan has been working on his specialty for more than three decades, but said it was difficult for him to find the sort of unique opportunities that were a fit for his expertise. Now, by working online and tapping into global demand for his skills, he’s turning away projects. In fact, Bloomberg recently discussed the increase in demand for data scientists; writing, “One measure of demand: Hours billed for work in statistical analysis grew by 522 percent in the first quarter [of 2013] compared with the same period in 2011, according to data compiled for Bloomberg by oDesk.”

This all ties back to 18th century economic theory from Adam Smith, who said that the degree of specialization you can have depends on the extent of the market. Long before the time of Internet marketplaces and Amazon, he used the analogy of a pin factory. If you can only sell 100 pins a day, he said, it doesn’t make much sense to hire a lot of narrowly focused people. If, however, you can sell to the whole world, it makes complete sense to hire more specialists who provide a specific skill that will increase the business’ overall productivity.

As the online work market grows, the opportunities for businesses and specialized workers will continue to grow with it. Not since the Industrial Revolution have we seen such rapid disruption in the labour market. We expect the long tail phenomenon on oDesk to only grow as an increasing share of the world’s professionals turn online to find markets for any skill they choose to pursue.

Shoshana Deutschkron

Director of Communications

As Director of Communications, Shoshana brings to oDesk extensive experience in strategic communications for the technology industry. Her unique data-driven background helps uncover insights from oDesk’s online workplace data. Prior to oDesk, Shoshana was a Vice President at market research firm Penn Schoen Berland, where she provided research-based communications counsel for companies including Ernst & Young, HP, EA and eBay. Throughout her career, she has… read more

  • http://www.linkedin.com/in/joeohlerjr Joseph Ohler, Jr.

    "Hours billed for work in statistical analysis grew by
    522 percent in the first quarter [of 2013] compared with the same period in
    2011..."

    Wow -- then I'm surprised BuyMyStats never took off. A
    post-mortem reveals inability to penetrate the statistical processing control
    (SPC) market; an emphasis on marketing to poorly paid graduate students as
    primary consumers; and a marked lack of social capital as the three issues that
    inhibited success. RIP BuyMyStats!

  • Ravi Bajracharya

    This is an interesting observation of the oDesk distribution of demand and its comparison to long tail concept. It would be interesting to further see the variation with more data and with more variables such as payout and timeframe.

  • my 2 cents

    It's good we have a better functioning market, but the inbalance between capital wealth and workforce wealth is still the same, and the design failure of the capital/work ecuation where the capital eats up and/or insignifies the work by efficiency gains among others is still present. A design failure that destroys all, the capital, the work, and the market.

  • Adil

    The graph is alright for those who are already into business and doing good. It is less informative for those who need more data understanding for business entrepreneurs. It would be constructive if more details is given.

    Thank you

  • http://gravatar.com/tuibiz David Barton

    Hi - I am a little confused by this graph - can we connect on skype so you can explain

  • Chris

    I am disappointed in your graph. The circle sizes and colors distract from the data points. In my opinion as a scientist the only reason to use point size variation and/or color variation is to show addition variables. Also, the data points are far to densely packed to be distinguished for one another. Either trimming the data set so that each point could be labeled or making the graph interactive so that the user can investigate individual data points would make it more useful.