Skip to content

Can Semantic MediaWiki cross the chasm?

I recently read “Crossing the Chasm” by Geoffrey Moore, which is a book I’d heard about a few years ago, described by someone as required reading for tech entrepreneurs.

The basic premise of the book is that the rise in popularity of tech products is very rarely a smooth upward curve: even for products that have a nice initial burst of popularity, it’s very difficult to move past the early user base and start to get acceptance among the general population (or the population who might use one’s product, in any case). That’s because those early adopters are usually the ones who already seek out and embrace new technologies; while the majority of people are either, in the book’s vocabulary, “pragmatists” (who just care about getting things done with a minimum of headache) or “”skeptics” (who actively dislike switching from whatever they’re using now).

When trying to move from the early adopters to everyone else, Moore argues, everything has to change: your marketing pitch, your sales strategy, the array of additional services and peripherals around the product (if any even existed), and even your company culture.

“Crossing the Chasm” then gives a prescription for the marketing and sales strategy to use: find a suitable niche market, and then put all your focus on that, trying to monopolize it to the greatest extent possible. The operative theory is that pragmatists and skeptics tend not to switch over to new technology unless they start to see people around them using it; so you need to focus on a very small group of people and keep marketing to it until many of them have become convinced by each other’s examples to start using your software. Once that happens, you can let your success among that group propagate out to related markets, to other departments within organizations you’ve already sold to, etc.

On top of that, there’s then a specific branding strategy, to use when marketing to that niche group: find the main software you’re trying to replace, then describe yourself as like that software, but with one improvement: the pitch should sound like, “it’s like X, but Y”. That applies even if your software has more than one benefit over your competitor: the key is, at all costs, not to confuse your potential new customers or overload them with facts, which would give them an excuse to tune you out.

There are some case studies cited in the book; the one that struck out at me was Documentum, a document-management software company which did just that – in 1993, after several years of flat revenues, they decided (based on reading “Crossing the Chasm”’s first edition, it turns out) to focus on a tiny niche: the regulatory-affairs departments at large pharmaceutical companies. As the book puts it, the company was paradoxically “reducing its market scope from ‘all personnel who touch complex documents in all large enterprises,’ to maybe one thousand people total on the planet.” It worked for them, though – they started doubling their revenues every year, they went public in 1996, and in 2003 they were bought by EMC.

Crossing the Chasm was first published in 1991, and then re-published in 2000 with extensive rewrites; the 2000 version is the one I read. So – is it still relevant?

There are ways in which “Crossing the Chasm” is an outdated book, especially for the open-source web software business. The overriding feeling of the book is one of urgency – in its parlance, you have to, with great discipline, “secure a beachhead”, before you can be attacked by rival companies or by “vulture capitalists”, VCs looking for struggling companies to buy up and exploit. The source of all that angst is that, for a company developing proprietary software, the “burn rate”, the amount of money it takes to keep the company going, pay developers’ salaries, pay rent on the office, etc., can be large – over $100,000 per year per employee, on average. For a 20-person startup, that’s over $2 million a year, so if you’re not making that much in revenue every year, you’re in constant danger.

There’s just less urgency in open-source. It’s true that some open-source software companies do follow the proprietary software model – the software is developed by a small group of people who work together in an office, and they just happen to release the resulting code for free. But I think that’s the exception: most of the time in open source, the development work is split between paid and volunteer developers, and even the software’s paid developers can be split among different organizations. (Also, developers often work remotely, saving on office expenses – though that’s happening more and more for propriety software as well.) All of that serves to keep development costs much lower for software of what I think is comparable quality.

Beyond that, the nature of marketing software in 2011 is almost another world from marketing software in, say, 1995. Back then, the success of the Netscape browser in getting market share through free internet downloads pointed the way of the future; but for the most part, software was sold in stores or through third-party sellers, and people learned about new software from their acquaintances, or through a network of salespeople, or by reading trade publications. These days all of those still exist, but we also have tech blogs, Twitter feeds, and web searches. And in the case of web software, people can instantly check out examples of the software in action. And in the case of open-source web software, they can also download it themselves and try it out for free. (Though proprietary software can also have free evaluation periods and the like.) Which means that, unlike 15 years ago, lots of people can find out about, and use, your software, without a single sales call having to be made (even if people still called on the phone, which apparently no one does any more).

(Finally, there’s a third major difference, which is that much of the software used these days is hosted directly on web sites – or “in the cloud”, as people like to say. And WikiWorks is in that business too, with the Referata wiki-hosting service. But my main topic here is Semantic MediaWiki, which has to be downloaded.)

With all that said, I think the book is still relevant. Human nature hasn’t changed in the last 20 years; most people still won’t switch to using software until they know that other people in their situation are using it. Which means that, given the choice, it’s probably better for one’s future success to have 100 people from the same industry or niche using one’s software, than to have 100 people randomly distributed in different careers around the world.


So where does that leave Semantic MediaWiki? (You could ask the same thing about MediaWiki as a whole, which is another interesting question, but for the sake of simplicity here I’ll just assume that the ideal installation of MediaWiki always involves SMW, and thus that the two are synonymous.) Our usage is right now essentially random, with some interesting pockets of adoption in different markets but no real pattern to it. There’s definitely an argument that we’d be better off focusing our marketing efforts. By “marketing” in this case, I mean the wording we use to describe the software, the kind of marketing documents we create, the types of conferences we attend, the blogs and publications we pitch to, and even potentially the kinds of peripheral software created.

The obvious next question, then, is which market, or markets, should be targeted. Interestingly, pharmaceutical companies – the niche Documentum started with – might be a good choice for us too. Pfizer already uses SMW, and another one, I’ve heard indirectly, almost started to use it for a major project, backed out a few months ago, but still may end up using it later. Pharmaceuticals have a combination of massive amounts of data; an always-changing data structure (because the concepts being discussed are always in flux), which makes creating industry-specific applications difficult; and massive amounts of money. The first two make SMW a good tool to use, and the third makes it definitely worth our while to get involved.

Another interesting niche is biotech companies, for much the same reason.

Another strong possibility is one or more sectors within the U.S. Government. There, too, there are massive amounts of data, and, for better or worse, massive amounts of money (at least at the federal level). WikiWorks has already done Semantic MediaWiki-related setup work for some U.S. federal agencies, and we hope to do more.

So those are some potential niches; what about a branding strategy? To the extent that there’s a single, identifiable competitor, it seems like it would have to be Microsoft SharePoint. So what’s SMW’s one advantage over SharePoint? Is it like SharePoint, but open-source? That’s certainly catchy, but there are already various other applications that bill themselves that way – which doesn’t mean, of course, that SMW can’t join them. (Very interestingly, the main application listed in that article, Alfresco, is, from what I understand, an open-source update of Documentum, produced by some of the same people. It’s all come full circle! Somewhat less interestingly, I once met the guy from Alfresco quoted in that article.)

SMW could also be “like SharePoint, but Semantic Web-enabled”, which is undeniably true, but maybe not very compelling at this point. And it could be “like SharePoint, but a wiki”, which is, I think, easier to understand, and possibly more compelling.

Any thoughts from the crowd? Comments, as always, are welcome.

Update: I just want to clarify that I’m in no way dismissive of all the many varied and interesting uses that Semantic MediaWiki has gotten so far, including among our clients and including wikis run personally by various WikiWorks members. I hope SMW continues to get lots of usage in all different sort of fields – all I’m trying to discuss here is where to focus future marketing efforts.

Categories: Uncategorized.

Comment Feed

3 Responses

  1. I’ve tried SharePoint at work and it stinks awfully. The only reason to try it is if you are in Microsoft’s pocket. I can tell that SMWHalo is aiming at the crowd, given their emphasis on MS Office document integration. As people rely on Open Office and Google documents more and more in the work place, the tie-in with MS products decreases and decreases. The key is to get people to change how they create content in the first place.

  2. Awesome blog post! :)

    I’d go with “like SharePoint, but a wiki” because “but Semantic Web-enabled” might be a little vague and “but open-source” could suffer from people still thinking open source can’t work or is evil in some other way.

  3. Interesting angle. I’m thinking that with all the social media, its more useful than back then to target a specific group regardless of their geographic distance. The “office” is more spread out distance-wise but the entire world is much closer than at the time of authorship.

    As far as branding, I think it depends who you’re targeting. The non-technical have no idea what Sharepoint is. I usually go with “like Wikipedia but with database capability” (or you can reverse that). I find that it works, though Wikipedia (i.e. non-semantic Mediawiki) isn’t usually a competitor.



Some HTML is OK

or, reply to this post via trackback.