Asking a start-up about their business model is a classic "wet blanket" move at conferences. It's an absolutely critical question, but there are better questions within it that deserve asking, questions that often aren't asked by established businesses.
For instance, new entrants are typically cast in one of two market approaches -- wide or deep.
In the "wide" case, wide adoption is necessary for the business to succeed. A good example is a general search engine or a browser. Both need lots of customers, and getting those customers tends to tilt the market in the long-term favor of the winning firm. The intrinsic nature of the model when purveyed for browsers means end users don't pay, but people pay to access customers through the reach of the major "wide" players. The contracts to be featured as the main search engine are very lucrative, and other bidders pay to be featured services. But wide adoption is a condition of success.
In the "deep" approach, a few customers deeply engaged can make a business work. Bloomberg Terminals provide a good contrast. Comparatively few people use them, but the value of the product is so high to those few that they are willing to pay a lot. The company has to keep the value high, and lock-in is a powerful element of ongoing success. Those few customers have to be deeply engaged and loyal.
In either case, cost-per-acquisition is a critical metric to consider, as it impacts growth potential, cash flows, and the likelihood of having enough money to enter the market effectively. If you are pursuing a "wide" strategy, cost-per-acquisition usually has to be as close to zero as possible. For a "deep" strategy, cost-per-acquisition can be much higher.
I've heard of some publishers who have launched "wide" play products or services with a marketing budget based solely on direct expenses for promotion and creative. Yet, when the cost-per-acquisition is multiplied with their expected number of customers, the total spend far exceeds what they have set aside. In cases like this, the project is likely doomed, or will at least contain some unpleasant financial surprises. Add to this a rolling expense for retaining current and incoming customers, and the expense level can truly astound the unprepared. With a "deep" strategy, in which customers pay a good amount, can be upgraded at a predictable rate, and stay for a long time, the cost-per-acquisition can be much higher. But it still has to be calculated, and has to fit into the overall financial model.
The trajectory for cost efficiency is also important to take into account. Will new customers become more expensive to acquire as you move from the core market to peripheral markets? Will technology become cheaper to build and maintain as adoption grows? Or will the opposite occur? In many cases, core customers are the cheapest and easiest to acquire, so have the lowest cost-per-acquisition, and because they are the easiest to identify, technology costs are lower (e.g., fewer databases to integrate, less data cleanup, smaller databases overall). Going into secondary and tertiary markets to pursue the "wide" strategy can increase expenses unexpectedly. Some customers aren't worth acquiring, or keeping. Do you know who they are?
In scholarly publishing, too often non-financial measures become the currency driving decisions. Will this increase our impact factor? Will we attract more papers? But focusing on non-financial measures, while useful to a degree, should come after the financial costs and benefits are well-understood. What good is a higher impact factor if it breaks the bank?