Several readers, colleagues, and blogs question MetroFi's move to all free metro-scale Wi-Fi: The move seems absolutely clear to me, but others think it's either reminiscent of the dot-com "get big fast" or "give it away" models or simply ill-advised. I don't see it. In my post yesterday, I noted that MetroFi used to charge $20 per month in Santa Clara and Cupertino; now it's free in those two cities and Sunnyvale, its test market for advertising.
At $20 per month, you're already paying a decent percentage just for credit-card processing, some amount for bill presentment, some amount for billing staff, some unpaid bills, and accounting costs. Even with a decent number of customers, at least several dollars a month is eaten up with pure overhead. CEO Chuck Haas said the company had "thousands, not tens of thousands" of customers before the switch to free. Haas also noted that a significant percentage of his customer support costs centered around billing. (Some hotels have found that not charging for Internet access was only slightly more expensive than charging for it, and that they accrued other benefits, like increased room nights to boot.)
Marketing might figure in several dollars per month per customer. The cost to acquire a customer for MetroFi's service is unknown, but one might suspect that it's from $20 to $80 per customer based on costs found in cellular and ISP markets. Over the first year of acquisition, that's a significant percentage of a $20 per month fee.
Now look at the replacement revenue side. MetroFi would rarely get casual users when it was pay-for-use, whether residential or mobile. Now, they will the network of choice across every city they operate in for users that can see the signal as long as they live up to providing a reliable and consistent data rate. This increases the pool of users immediately by thousands of unique individuals a month. Residential users who might not have wanted to pay $20 per month, but are paying that for dial-up now become immediate converts if they try it out and it works.
The three cities MetroFi covers have roughly 300,000 combined residents. This doesn't include the influx of tens of thousands of workers daily. If MetroFi had, say, 3,000 regular users prior to switching to free (1% of users), picks up 5,000 monthly unique users with laptops and handhelds, and perhaps just 3,000 more residential users, they're suddenly at 11,000 unique monthly users.
3,000 users would generate a gross revenue of $60,000 per month before expenses. Assuming the most optimistically cheap scenarios for billing and marketing, let's say that $45,000 per month was left cover CapX, forward investment, and profit. 11,000 users need to generate a little over $4 per month each on average to exceed that--probably more like $6 per month to handle the per user costs that will go up.
If advertising were still sold exclusively on a banner-ad model with a cost-per-thousand (CPM) rate, then MetroFi might be able to sell ads at a $15 to $25 CPM, and feed 250 impressions over a month to each user on average to achieve the $6 figure. However, local, targeted advertising can be sold at a higher rate or on a cost per objective (CPO) model in which leads or sales conversions result in much higher bounties or results. The Google AdSense model shows that certain keywords are worth vastly more than others resulting in extremely high effective CPMs even with low clickthrough.
When critics of MetroFi's actions say it won't work, my conclusion is: if an average user sees about 250 to 300 banner ads per month, MetroFi's math works. If they see 500, MetroFi makes some real cash.