Use ground power whenever available and turn the APU off. If you need the APU for ground power, then try to turn the bleed off from the APU if you don't need aircon due to ambient temperature. Even just saving on bleed air requirement from the APU will save a significant amount of burn.
Use an efficient flight planning system that gives the pilots confidence in the flight plan fuel figure. This breeds a 'culture' of confidence and will allow pilots to trust the plan figure and not just add on a few hundred kilos here and there due to perceived innaccuracies.
Never question the commanders decision to load what he feels is needed, but look at trends and work with captains to get to their underlying logic. If their logic is sound then great. If training input could be beneficial then give that input during checks. There is no point 'telling' a captain that he should take less fuel, you need to get him to understand and trust the planning system to have faith that it is accurate and sufficient (ammended by him for on the day 'real world' foreseable increases).
On that note, most pilots have a culture of looking for reasons to add fuel to planned figures, to create that buffer. Why not be more pragmatic and look for things which are likely to reduce your burn also? I am talking mainly shorthaul around EU here, as you are never far away from suitable alternates and my company would prefer us to not load unnecessary fuel and if things don't work out as planned and we need to divert for a tech stop, so be it. They would NEVER question your uplift, and they would NEVER criticise you if your planning and excecution were reasonable but things just conspired against you. It leads to a mutual culture of trust and is a benchmark for efficient fuel planning. An example where fuel can be saved with us is going to Malaga. The flight plan system always plans on the longest STAR, which is some 70nm longer than the shortest, which is the one most usually flown.
In good met conditions use a 'fuel' alternate as your first destination alternate, which can be much closer to your destination than perhaps the 'commercial' alternate (where they would ideally like you to divert to if you didn't get into destination). If the diversion risk is minimal this can save significant amounts of burn and the company accept that if you do end up diverting then you won't be going to their 'more commercially desirable' airport and it will cost them a bit more.
Single engine taxy in wherever suitable. Again, it needs a culture change, but if you can get that and make it SOP to do unless there is a reason not to, you will save huge amounts of fuel over a fleet, over a year.
A huge issue at the moment is ATC. In busy airspace like the UK it is often difficult to geta continuous idle power descent from ToD to short final. This is being looked at by NATS and the airlines to see if anything can be done, but CDA approaches are another area that can save significant amopunts of fuel.
To achieve the above pilots need to plan their descents and energy manage to the best of their ability, which comes from having correct information to base that plan on. Ensure that you program the FMC with the most 'likely' routing to happen to you, rather than just what the flight plan has on it, which comes from route knowledge. Early descents due to ATC can be better managed in my experience - I often see pilots descend early to reach a certain point at a certain level (ATC restriction), only to be given further descent before reaching that point and they continue to descend at idle thrust beyond the limitation when they are now low on profile and increasing burn later on at the lower levels.
Oh, and carry a smaller torch in your nav bag.
PP