I know what you guys are saying. There is just a major major difference when I hear my tracks in a club on a real system vs mixing them on my monitors at home. I can have what I *think* is the same sound as some of my favorite bangers, but head to head in a club they aren't even close. The main difference is always the low end . Either missing or non - existent. If I add what by memory I feel needs to be added, then it sounds wrong in the headphones .
Please read rhythmgj's post above, he makes important points.
One thing you need to be aware of with headphones is how non-flat they are both in absolute and in relative terms, of course some headphones are more flat than others, but when you both have an overall curve that is not flat from headphones by large and then on top of that you have various parts of the frequency range where the frequency response differs a lot between various headphones, then in combination with the fact that the low end is the product of the whole frequency range (also the mids and highs), then all in all it becomes challenging, especially if you on top of that is working with software that by nature produces a less true low end. Modern club mixes today rarely are produced like that, there are almost always critical hardware components that produce the backbone of what you hear in the club, further extended by the DJs.
To break free from this, by far the best solution is to create great monitoring. But even then I think that to be competitive today one has to have at least some rough ideas on RMS, integrated LUFS, True Peak levels and average frequency across the frequency spectrum across the stereo image, so that the mixes also can compete in terms of the impact of the raw sound. For instance for impact a mix needs high enough "good" signal in the mid range and that is difficult to achieve using only software, but with hardware, side chain and two stage compression and brickwall peak limiting you achieve totally different results.
Adding to all of this various stereo processing and all of a sudden we are talking a totally different perception of the low end. So yes monitoring is key but keep in mind that the monitors just reveal what you have in the sound, they do not produce the sound itself. So the issue you are facing stems from the combination of monitoring and content.
In my post above I highlighted how you can use both hardware monitoring signal paths and software monitoring signal paths (with calibration), this helps a lot and you should also combine this with proper A/B and metering of the levels and referencing with multiple monitor and headphone sets. Keep in mind also the impact that saturation has on the low end, when you saturate the low end with hardware and then level the result of that with hardware compressors and limiters you end up with a sweet sound. Combined with clever routing and hardware EQ filtering, the sweetness is separated out so that it becomes more central in the overall listening experience without creating negative density. Also be aware that with software it is very difficult to achieve high information density combined with high quality transients, I would say it is impossible unless the initial frequencies are so near the final that you don't need to do anything more.
So it's not so hard, but because you need to mix together a number of qualities and because it takes some time also to learn to properly balance a mix, it can be challenging. Just follow the advice in this thread and you will succeed with this.