What is the utility of the command "ip classless" in running configuration?

When we use default routes we should use ip classless because routers are by default classful.
  The behavior and the way router in order finding the destination subnet is different when we implement using ip classless vs ip classful   Example routing table :       10.0.0.0/24 is subnetted,3 subnetsC      10.1.1.0 is directly connected, FastEthernet0/0D      10.1.2.0 [90/2172416] via10.1.1.254,00:00:33, Serial1/2C      10.1.3.0 is directly connected, Serial1/1 S* 0.0.0.0/0 [1/0] via10.1.3.254   Case when router received packet destined to10.1.4.4,   ip classful behavior : the router will try to find the answer on the10.0.0.0/8 subnet class and will not try to look even further outside that range, it could not find any line match, packet will be dropped ip classless behavior : the router will try to find the answer on the10.0.0.0/24 subnet class, since it can't find the match destination subnet, it tries outside the subnet class, hit the default-route line, packet forwarded to next-hop10.1.3.254   Hope this helps..

If you don't use "ip classless'' command then by default whatever the IP you configure will be counted as classfull.It means if you want to configre81.62.34.0/24 automatically it will configure81.62.34.0/8.Because class A  starts with0 and ends with127.So, when you configure VLSM IP you need that command.In real world it is always better to use 'ip classless'.Because public ip you get from your provider are not classful.