You really need some kind of positive control sample that does amplify to show that the pcr can work. The bright small blob smaller than the ladder smallest band is primer dimer. The primers are annealing together not to the target sequence and PD amplifies very well and removes all the primer from the reaction so no product. You can try using less ( 1/8 th) the amount of primer or just use a hot start enzyme for your pcr. A really good pcr includes a positive control to show that the pcr can work and a no template control to show that there is no contamination amplifying
Colony PCR is also notoriously tricky. I know it sounds counter-intuitive, but using less bacteria will typically give you a higher success rate. Also, as Paul Rutland said, you should include a positive control (and a negative control).
Typically more than a single colony is tested. It looks like you have room for 5 colonies (with ladder, positive control, negative control in your remaining 3 lanes). One method that can help avoid bacterial overload is to pick single colonies into 100 ul of colony broth with selection, vortex or flick to mix, shake for 1 hour at your typical growth temperature, then use 2 ul of your shaken colony for your PCR template in a 20 ul reaction. Be sure to mix the 100 ul prior to pipetting your 2 ul. Save the rest of the 98 ul in the fridge until you run your gel, then use some of the 98 ul for a 1:1000 inoculation of overnight cultures. This is my lab's usual approach for colony PCR of E. coli, Agrobacterium, and S. pombe.