Why is Chempedia Lab Failing?

March 09, 2010

A little over three months since the launch of Chempedia Lab, I have some bad news: it's failing.

For the unfamiliar, Chempedia Lab is a question and answer site dedicated to experimental chemistry. The value proposition is simple: ask your toughest question and get a peer-reviewed answer quickly. At least that's the idea. There's a lot of not-so-secret sauce behind the technology platform, but you can read about that elsewhere. The most important points are that Chempedia Lab is an experiment in community building among professional chemists and advanced students, and that its approach is unique.

This article will discuss the original vision for Chempedia Lab, and offer some speculation on the causes of its failure so far.

A Vision for Collaboration

To say something is failing requires a vision of success. When my company launched Chempedia Lab, the vision was for a Web-based service that would become the premier source on the Web for up-to-date, peer-reviewed information on experimental chemistry. Chempedia Lab would do so by attracting a large number of practicing experimental chemists to ask, answer, and review technically-demanding questions.

Some Numbers to Date

After three months of continuous operation, what do the numbers say about our Chempedia Lab experiment?

Registered Users : 32

Users Asking At Least One Questions : 8

Users Answering At Least One Question: 14

Questions Asked: 42

Unanswered Questions: 6

Given the vision for what Chempedia Lab could become, the metric of highest importance is the number of users asking at least one question (eight). Although probably obvious, this metric matters a lot because the entire Chempedia Lab system is driven by new questions. Questions posed from diverse sources leads to a highly engaged community. Fewer questions leads to the opposite outcome.

Eight people asking questions might make for lively dinner conversation, but for a question-driven online community, it's far too few to be self-sustaining.

Of secondary importance is the total number of questions asked. With fewer than 100 of of these, it's clear that lack of users asking questions is leading to... well, an overall lack of questions.

The Good News

Although Chempedia Lab has so far failed to live up to the expectation of becoming a hub for experimental chemistry, there are some reasons to be optimistic. For example, try a Google search for 'solubility alcohols ammonia', 'dry hydrogen chloride gas', or 'cold bath temperatures'. A Chempedia Lab question is returned in the first four results for each search. In other words, new users are constantly being funneled to Chempedia Lab in response to keyword searches.

Another cause for optimism: since its inception, about 80% of Chempedia Lab's visitors have never visited the site before. New users continue to find the site, although relatively few are returning.

Possible Interpretations

Although any number of reasons for Chempedia Lab's failure to date could be considered, most of them fall into one of four categories:

  1. Chempedia Lab is unknown to the vast majority of experimental chemists. Simple problem and an even simpler solution - publicity, patient explanation... and time. Easier said than done, but not impossible.
  2. Chempedia Lab is widely-known, but does not appeal to those who for whom it's designed. Much more difficult problem to address because users who are turned off just leave, never come back, and don't offer feedback. Specific possibilities might include the notion that research chemists are uneasy about looking foolish in front of their peers by asking a 'too easy' question.
  3. Chempedia Lab is useful and reaching its target audience, but intellectual property issues prevent widespread participation in the asking and answering of questions. If this is the cause, there's little that Chempedia Lab itself can do to address the problem.
  4. The underlying system is flawed as a community-building mechanism in chemistry. In other words, one or more peculiarities in the world of experimental chemistry work against adoption. By comparison, a number of technologically similar sites on different topics have done quite well.

Of the four explanations, (1), (2), and (4) can be addressed by Chempedia Lab directly. At this point, I suspect (1) as the primary cause based on the relatively low traffic on Chempedia Lab to date. How low? Consider this graph, created with Google Analytics, and representing non-robotic visitors:

As you can see, on any given day fewer than 100 (non-unique) visits are made to Chempedia Lab. The traffic spike on January 27 was due to a flurry of referrals from StumbleUpon. The spike in late November came at the launch of Chempedia Lab, when a mass e-mailing to over 500 opt-in only addressees was sent. Traffic has been essentially low and flat from the beginning, marked by periodic spikes.

Reasons (2) and (4) are likely secondary causes. Chemists in industry may barred from participating or otherwise discouraged, reducing the pool of potential users. Lack of participation begets lack of participation.

Where To?

Community building is hard. And the job is made particularly difficult within fragmented groups having strong pre-existing traditions (and taboos) around communication - such as chemistry. It's possible that an idea like Chempedia Lab will never work in such a community.

But assuming that this worst case scenario doesn't apply and taking point (1) above as the main issue to be addressed (few experimental chemists have seen the site), there are a few possibilities that might be tried:

  1. Reach Key Influencers Individually. You know the gal who always seems to know where to find things and always has the latest information? That's the one. The idea being that if she knows about Chempedia Lab and uses it, she'll tell everyone in her immediate range of contacts about it. Chemist-bloggers might be one group. Unfortunately, having sent individual messages to a number of them, only one has responded, and to my knowledge only one has written anything about Chempedia Lab. Reason (2) at work?
  2. Reach the Entire Community at Once. What's the single biggest hurdle in reaching experimental chemists? There are few 'places' they hang out in large numbers. The weekly magazine Chemical & Engineering News would be the top candidate for this position. Although the price for paid advertising is high (thousands of dollars), coverage in the form of a news story is relatively inexpensive. To make Chempedia Lab newsworthy in the absence of widespread use would require something that has so far been missing (e.g., a remarkable story), but may nevertheless be an achievable goal. Lesser venues will have a lower bar to granting coverage.
  3. Do Nothing. That's right - just sit back and observe. The more fragmented the community and the more entrenched its practices, the slower it will be to adopt new ideas. It may be that the pain level of connecting with other experimental chemists outside of one's core personal contacts/institution, and in getting tough questions answered quickly is not yet high enough to overcome the barrier to trying a new way of doing things and risking professional embarrassment. Next year, things might be different.

Conclusions

For Chempedia Lab to have a future as a self-sustaining community resource, it's going to need a lot more participation from people willing to ask questions. It will require key influencers to take a risk and try it. It will take some creative uses of new- and traditional-media. And it will in all likelihood take time.

Imagine a resource that every experimental chemist can turn to for fast peer-reviewed answers to tough questions. A resource that updates itself as the facts are updated. A resource that helps you both with the technical side of your work and the social side. A resource that costs nothing to use - ever.

Chempedia Lab has far to go before living up to its potential, and may never make it. The headwinds it faces are not technical, but social. Still, as long as practical options still exist, it makes sense to continue with the experiment.