The costs of the MVP Summit are HUGE. I once calculated hotel rooms alone at close to $2 million. Add to that venue and bus rental, parties, food, banners, etc and it becomes a very large cost.

There are other things like company store purchases, shirts and other goodies, air fare for staff to attend events, etc.


>
>It's interesting to speculate on the mentality behind the MS MVP program.
>
>Superficially, the program is an absolute *screaming* deal for MS. The MVPs, some, if not many of whom are literally the best in the world in their fields, give away significant amounts of their highly-valuable time. Their efforts reduce tech support load on MS and, perhaps more importantly, evangelize MS products world-wide.
>
>The "cost" to MS to run this program is the cost of the software bundle MVPs receive (discussed below), plus the staff required to run the program (also discussed below).
>
>Staffing - it's arguable that whether MS runs an MVP program or not, they should be monitoring cyberspace and the real-world usage of their products. The marginal increase to run an MVP program may be small to none.
>
>Bundle - the marginal cost to create a bundle for an MVP is essentially zero. Not even any distribution media these days - there's just download links and product keys at MSDN. Bean counters would argue that MVPs would have bought some MS software in the normal course, so that represents a loss of revenue. That's probably true, but it's also true that very few, if any MVPs install and/or use the full set of products included in their bundle, and probably would have bought even less.
>
>So, from a bean counter perspective, MS should hand out MVPs to anyone whose contribution to tech support, evangelism etc. exceeds the value of the portion of their reward bundle they're likely to have bought regardless. It's hard to put concrete numbers or values to these nebulous/squishy concepts, so it's not surprising MS doesn't publish its MVP requirements - too much temptation for clever people to game the system.
>
>I agree that "the best volunteers do it whether they get a reward or not". Maybe the abortive closure of the MVP program was a bean-counter thinking the unrewarded volunteers were Pareto-optimal: MS would get 80% of the benefits with 0% of the cost (whatever it may be) of the MVP program, if they relied just on these unrewarded volunteers.