Comment by nroach | Hacker News Reader

Comment by nroach | Hacker News Reader

Parent Full thread

I got a good laugh out of that comic, and it's frighteningly familiar for a lot of enterprise developers.

But taken too literally, it sends entirely the wrong message to a startup. In some cases I guess you really do want to swing for the fences and bypass everything that's been done in your industry. However, I suspect that the failure rate among those that take a completely divergent approach is pretty high.

Maybe you don't have to match the competition feature-for-feature, but you do need to match them need-for-need. If you don't address customer needs, good luck with your conversion rate. And in the end, a feature matrix is just one company's articulation of what it thinks those customer needs are. Maybe you disagree or maybe you have a better way, but if the same 70% of features show up over and over again amongst your competition, you'd better find a way to satisfy those needs.

In the interest of not posting a tome, I've elaborated on my blog (see profile for link).

you do need to match them need-for-need

I disagree. In particular, you may not be addressing the same niche that the "competition" addresses.

To use an example I've been thinking a lot about today: it is stupid and suicidal for any startup to go head to head with Google on a core Google competency, right? Well, Wingify just released an A/B testing product which is about 5% as featureful as the Google Website Optimizer/Google Analytics tag team o' doom. It lacks on enterprise features, it doesn't have the Google ops team backing it, etc etc. It also won't be free due to a cross-subsidy from Google AdWords.

Instead of trying to reach feature or need parity with Google, they said "There is a portion of the market here which is untechnical and knows they want to do A/B testing but whom are undeserved by EVERY vendor. We're going to make A/B test easy enough for them to incorporate it into their businesses in a few minutes." With respect to those users, they're going to eat Google's lunch. Granted that might only be a couple hundred thousand customers but, well, how many customers do you really need, anyhow. ;)

Well, my surreply would be that for those customers, there is only one one need, and Google's feature matrix attempts to address more needs than actually exist in that segment of the marketplace.

But segmentation is a good point that I hadn't really addressed. I'd presumed competitors going after the same market. A more mature competitor may simply be seeking a bigger market (and thus requires more features). If you're satisfied with attacking just one segment, everything else is extraneous.

One of the key points that Clayton Christensen Innovator's Dilemma makes is that many customers can be "over-served" by a mature product: it has more features and capabilities than they need or can use. They do this by listening to their best customers, who tend to be power users.

Microsoft Word and Excel are clearly an examples of this, where simpler web-based solutions are gaining traction. What Wingify has shown is that Google Analytics may have crossed the threshold into "Swiss Army Chainsaw"--every blade you might ever use all in one handy package--when a simple handsaw may be all that many folks need.

Yesterday's Dilbert was about how long approval processes take in corps.

This one is also about corporations as well. The 12 month schedule in Dilbert's company includes a lot of overhead that doesn't exist in startups

Granted, I haven't really followed Dilbert since shortly after it's popularity.

But isn't every Dilbert about corporate nonsense?

(At least, those that are about work at all. I recall some odd Arbuckle-esque comics, but those are neither here nor there.)

> But isn't every Dilbert about corporate nonsense?

Nope. :) http://dilbert.com/fast/2009-11-04/

Reply on news.ycombinator.com