These days, consumer products are easy to code, but difficult to market. While getting investment, scaling up your infrastructure, and getting PR is easier than ever, getting the first 1,000,000 users is just as hard.
Andy Rachleff, formerly of Benchmark Capital, describes the trend:
The premier venture capital firms know the best investments have high technical risk and low market risk. Market risk causes companies to fail. In other words, you want companies that are highly likely to succeed if they can really deliver what they say they will. Unfortunately, consumer Internet companies don’t follow that pattern. They usually have low technical risk and high market risk. There is very little chance they can’t deliver their product. The big issue is whether the startup’s product is of value to a large enough audience.
If you believe that, then the question is, how do you ensure that you can get a big audience? Enter the idea of the distribution-first startup. Basically, from Day One, the majority of the work goes into figuring out the viral loop of the product.
You can figure out the right distribution channel by making sure:
- the product has natural virality (communication, file-sharing, publishing)
- the channel in which the product spreads is new (and thus high response rate) or where you have a strong advantage
- the value proposition for new viral users is strong- ideally you could prove this using small tests
- you can get a ready stream of initial traffic so that you can optimize (via ads or a previous product or something)
- you want to make sure your team is solid also
Once you figure this out, and you pick the distribution channel, and the viral loop, then you pick the product. The idea here is that the distribution defines the product, rather than the other way around. This can get really dangerous though, so you want to be careful here.
Ideally, you pick a well-understood product category where you know how to make a differentiated experience, and ideally that differentiated experience hooks into the distribution channel. If you are going to be the first Facebook-friendly photo backup solution, and you think that's both a good product idea and also has good distribution hooks, then you're all set.
What you don't want to do, of course, is to create some kind of viral gimmick. That leads to a shitty product with low retention, and while your viral loop spins away, you won't build something sustainable over the long-term. We've already run the experience of high viral-growth with unsustainable retention, and the end result doesn't look pretty.
Ultimately, the core product experience still matters a lot- so even if you have a good understanding of how it spreads, you need to quickly switch back between distribution and product contexts. Even if you start out distribution first, the work around making the product great will still ultimately dominate the work you do.
In the world where 10 million users is the new 1 million users*, it's not enough to have a great engineering team, or to have a well-designed product. If you need a ton of traction to succeed in consumer internet, then you have to figure out your distribution model. In fact if it's such a huge bottleneck, maybe you need to figure it out first. And yet, still, make sure you remember to build a great product once you figure out how to get it traction :)
*via Chris Dixon http://cdixon.org/2012/08/03/ten-million-is-the-new-one-million/