Category: ideas

Top 5 new business ideas

The month of lists has begun, so I decided to rank the business ideas from the last year that could or should be a big deal in the next year. Most of these ideas and companies have actually been around longer than 12 months, but they either reached a certain critical mass or captured my imagination in a new way recently.

1) Scrobbling
All my listening behavior are belong to Last.fm. They figured out how to not only capture what I listen to but also to incentivize me to keep my behavior data with them. Since my listening data is open for other services to use, I am willingly giving Last.fm the power to broker that data with other providers on behalf. That’s a very strong position to be in.

2) Meta ad networks
Feedburner and Right Media figured out that ad networks can be networked into meta networks. Right Media went the extra step and opened up their APIs so that someone can build a white label ad exchange of their own using the Right Media tools. All you need are advertisers and publishers, and you’ve suddenly got a media market of your own. I can’t help but wonder if these guys have stepped into the big leagues with the next really important revenue model.

3) Pay-as-you-go storage, computing, whatever
Amazon impresses me on so many levels even if they don’t know exactly what they’re doing. They are making it happen just by doing smart things with the resources already in their arsenal. Similarly, Flickr understands that the APIs you use to build your web site are the same APIs you want to open up as a service, and it’s paying off handsomely for them. The formula here is one part optimizing resources and two parts confidence that your business won’t crash if you share your core assets with other people. Stir constantly.

4) People-powered knowledge
I really like the Yahoo! Answers experience. I also really like the concept behind MechanicalTurk where knowledge can be distributed as a service. Machines are at their best, in my opinion, when they make humans capable of doing things they couldn’t otherwise do, not least of which is making the universe of human knowledge more accessible.

5) Widget-mania
It wasn’t until I heard about the big revenues GlitterMaker was earning that I realized just how powerful this idea has become the last year or so. Beck’s customizable CD cover reinforced the idea that everything is a tattoo or a tattooable thing if you look at it that way…and many people do. If only I could run AdSense on my forehead.

Idealizing media business models

Jon Udell notes that WSJ’s landlocked articles may in fact help them drive revenue but at what expense:

“PaidContent.org reminded me that WSJ.com is considered a major success not only in the realm of paid online circulation, but also in comparison to newspapers…This may be a successful model of publishing, but it seems to me a curious definition of success.”


Photo: niznoz

How do you measure the opportunity cost of gating your content? The most obvious way is to estimate the number of page views an article would get outside a gated wall and then extrapolate revenue off CPMs.

However, this equation has a fundamental flaw that is not so easy to calculate. It’s the core question of every media company. How do you measure success? Do you exist to make money or do you exist to connect people?

It’s easy to say that media companies have to be both. But which powerpoint slide are you going to show your board of directors first? One of those will ultimately drive every decision at the company.

But even those two metrics fail to measure success in media as I’d like to see it. I’d like to see media brands measuring success based on the quality of the relationships they are able to catalyze. And I don’t see why that’s not possible…it might be hard, but it should be doable.

For example, it’s not how many people read an article that matters, necessarily. It’s how deeply did a story help a person. It’s not how many ad impressions were served, nor is it how many clicks or even the number of sales that result from an ad on a media property. It’s about the types of ways the media property improved the relationship between a vendor and that vendor’s customers.

Imagine how much easier strategic decisions would be if you could look at a report each morning that showed how many of your site visitors were getting promoted in their jobs or referring their friends to particular vendors that advertise on your site.

Imagine going into the board room each quarter and showing that as a direct result of the activity at your web site the average visitor income level increased or productivity in the industry improved or something more substantial like the number of homicides in the area decreased or more people voted in the last election.


Photo: infomaniac

Imagine telling an advertiser that working with your media brand meant that their customer retention metrics would improve or that people would be talking about their products more or even that they could drive up shareholder value. Imagine the rates they would pay to work with you for those benefits.

And imagine the types of people that would want to work at this kind of company and the amazing products that would come out of it as a result of these measures of success.

Yeah, pipe dream, I know. At the end of the day, most of us just want to get paid for their work and live a simple life. There’s nothing wrong with that, and I certainly fit in that camp a lot of the time.

The question in my mind is: Are qualitative success metrics like these measurable and attainable? If they’re not, why not? And if they are, why would you pursue anything else?

Why (and how) the online ad model needs to change

Somehow I keep expecting some company to break through and solve the problems with the Google AdSense model. As advertisers, buyers and media vehicles get smarter about efficiency, the holes in the system get bigger and bigger.

AdSense revenues help a lot of mid to large-sized web sites, but really more as incremental revenue. By the time you’re big enough for AdSense to support your business there are several other revenue opportunities with larger payouts avaiable to you.

And there’s no doubt that AdSense (and most Internet advertising) is failing to help people find and buy the things that matter to them. How can it be that we have an ad model that is considered wildly successful when a campaign or ad unit gets a click-through rate of 1%? And the reality is that it’s much much worse than that on average.


Photo: DWS

Why are click through rates so low? Because the ads don’t matter to people. They aren’t relevant. They don’t help people identify products or brands that matter to them. They don’t help people locate the right deal at the right time.

Yes, some people get lucky if they’re paying attention. There wouldn’t have been $5B in search ad revenue in the market in 2005 if nobody was clicking on the ads. But the click performance and subsequent conversion rates suggest this kind of ad network is just a spray hose of wasteful bits showering the Internet with clutter.

It doesn’t work for advertisers, either. Advertisers want more control over their ads, where they appear and to whom they are shown. Blanketing text links blindly across the Internet does not necessarily result in paying customers. They know they’re wasting money, but they can’t afford not to be present in the network.

The AdSense model does much more to help Google and the Google shareholders than it does to help any of the customers it is supposed to serve.

I think the Amazon affiliate program is much closer to a more sustainable ad model for the future. When you can track clicks all the way to a sale then everybody wins. The weakest link in the Amazon affiliate chain is the media vehicle which has to work a lot harder to drive clicks that convert to sale. But the buyer and the seller are both happy, and that’s ultimately what matters most.

I’d love to see an ad network that is able to let media vehicles optimize the ad content and display rules for the ads. The look and feel of an ad is not going to crank up the conversion rates. Media vehicles need to help the right ad get to the right person.

For example, when I post on my blog, I should be able to flag a stream of ad content and define the type of algorhythm that makes the most sense for that post and the users who are most likely to read it. This post should probably link to lead generation service providers even though I haven’t explicitly used the term “lead generation” anywhere in the post…uh, well, you get the idea.

Likewise, users should be able to self-identify as buyers. I haven’t yet setup a wifi network in my home, so I’d love for every tech-related web site I visit to show me the latest deals and setup guides and retailers for wifi gear. I’d actually like the content on all those sites to adjust, as well. I want to see what’s new and interesting at these sites, but they should be able to surface content from deep in their archives that is relevant to the things I’m actively pursuing. My intent should edit the home page for me.

I guess I’m saying that somebody needs to build a service that on one side connects directly into an advertiser’s sales conversion or transaction systems and on the other side distributes marketing links and images for media vehicles to take and optimize. The system should track performance across the chain and offer optimization options at all points along that chain.

Pieces of this exist and some of it is very complicated, I know, but I don’t see why efficiencies can’t be improved. And if enough advertisers are able to offer affiliate programs to track impression-to-click-to-sale, then they may even start competing with eachother and offer better incentives to media vehicles that find customers for them.

Users would see ads for things they want to buy. Advertisers would sell more product. And media vehicles would earn more from the revenue share.  Where’s the down-side?

Dotcom business ideas over a pint or two

DotCom Challenger
Photo: SimonD

I think everyone has had the experience of brainstorming a “great” dotcom business idea over drinks with friends. There were more than a few good and bad ideas thrown around the offices of The Industry Standard as the editors learned that the companies we were covering didn’t always deserve to make as much money as they did.

I had another one recently that was kind of interesting, though not really new in any way, so I thought I might just post it here and see if anything comes of it. I haven’t come up with a name for it, yet:

ProductX is a toy exchange network for parents of small children. It consists of a directory of toys available for members; a profile for each toy with ratings, comments, and user submitted videos; a queue for each member to track which toys will be shipped next; and a distribution system for receiving, storing and shipping toys. Anyone can become a member by contributing X toy(s) and paying a monthly fee. The toy contribution will be donated to charity. After joining, a member can select any 3 toys available from the directory. Members can only keep 3 toys at any one time and must return a toy to receive the next toy from this or her queue.

There was a variation on this theme that was also interesting that involved essentially ‘platformizing’ a swap system like this. That way you could focus member networks on specific types of products of similar value.

And just to put some perspective on my thinking here, these are some of the other ideas I’ve either heard, contributed to or invented myself (not telling which is which) and about when they were conceived:

  • ThingThunk.com (2005). The ultimate product review directory. Inspired by Wikipedia, ThingThunk allows anyone to add a product, rate a product, comment, submit video, etc. It’s the ultimate shopping assistant where you can get advice from other people who have actually bought and used a particular product.
  • Bioconomist.com (2005). The Industry Standard of the biotech industry. Insider analysis, heavy-hitting journalism, industry metrics, deal tracking and job listings.
  • Makingamockery.com (2002). Jack Ass meets Gong Show online. People submit videos of themselves doing stupid things. Users vote for favorite stunts, but, more importantly, users can request stunts. Successfully performed requests win cash prizes and appear on the accompanying weekly TV show. (I’m now realizing as I type this that it’s more like Jack Ass meets Y! Answers.)
  • PublishingPilot.com (2001). A UK-based CMS and hosting environment for small businesses to crank out web sites without having to know anything about anything. (I got some momentum behind this one including a somewhat high-powered board of directors before my personal bank account ran dry and MovableType appeared on the scene.)
  • LargeLifeStyles.co.uk (2001). Fashion retailer for larger-sized women. (Actually, I pulled this one off using my father-in-law’s resources in textiles in London. We morphed it into a shoe retailer that performed mildly well for a few months until we stopped updating it with new product.)
  • Sugarpimp.com (2000). Sponsored behavior modification. This idea arose out of my desire to keep a floater friend of mine unemployed and travelling so that I could continue to get his emails from strange places around the world. Then it hit me that you could create a market of sorts where you could get sponsors to fund all sorts of things. The obvious twist was that this would clearly turn into some kind of strange fetish/porn market.

And my personal favorite (courtesy of John Masterson)…

  • MyBeltWithHooks (1999). An offline clothing personalization platform. The belt would be very expensive but any third-party developer could create hooks to attach to it. One example of a third-party hook is the real-world search engine, a small child or dog that would attach to your belt and go find things on command.