Multi-site, branded churches take advantage of technology to reach congregations in the tens of thousands.
The Washington Post (Jacqueline Salmon & Hamil Harris) profiled the McLean Bible Church (EIN 54-0763526, church, no Form 990), a 13,000 member church in Tysons Corner, Virginia that has just opened up a satellite operation in Rosslyn, across the Potomac from Washington, DC. Senior pastor Lon Solomon has a plan to open nine satellite operations around the DC beltway and eventually to have a congregation of 60,000, which will hear the same service through broadcast links.
Another church, the Soul Factory (EIN 52-2018347) of Forestville, Maryland, already broadcasts from branch operations in Atlanta and plans to open other sites in North Carolina and Alabama. It currently has a relatively modest congregation of 4,000.
A lot of the information about megachurches in the Post article comes from a survey conducted by the Hartford Institute for Religion Research at the Hartford Seminary (EIN 06-0647016 Form 990 handwritten!). The summary report is available ("Megachurches Today 2005") and they also make available a database of over 1,300 megachurches.
What these reports suggest to me is that the current definition of megachurch as a congregation with more than 2,000 members is no longer adequate. There is a new category of even larger church with more than 10,000 members, yet unnamed. A mere 51 of these 10,000 plus congregations already account for 15% of megachurch attendance, almost 700,000 in the pews on any given Sunday (except that many of them have stadium seating rather than pews).
There are eleven of these churches in California, nine in Texas, four in Georgia, and three in Florida. The rest are mostly in the South and Southwest, with a handful of others in Pennsylvania, Maryland, Michigan, Indiana, Illinois, Kentucky, Kansas, and Colorado. Two, Rick Warren's Saddleback Church and Lakewood Church in Houston, already have membership in excess of 20,000, and they will undoubtedly be joined by others in the next decade.
As is typical with so much nonprofit research, the Hartford group underestimates the impact of these very large scale congregations because they ignore the cumulative impact of their membership: they limit their analysis to counting churches. But a single congregation of 10,000 is the equivalent of five with 2,000 members. So although they may be few in number, these churches will continue to capture a significant percentage of the chuchgoers.
Their growth seems assured, because it will be easier to start a branch of a church by attracting congregants with marketing techniques and then broadcasting content to them than to go through the time consuming one-on-one process of traditional church planting. It seems likely to me that there will be some regional and nationwide churches developing in the next ten years, with branding replacing denomination as an identifying factor.