I found myself in this position today.  As black I felt I had the better of it but in trying to learn to critique my own games did not grasp what was most important as a next step or immediate goal.  To the more experianced players out there if you feel like it what would your thoughts be.  not looking for a heady analysis just a shoot from the hip "what you would do"  I did manage to gain a favorable position later in the game but lost due overlooking a secondary and thus what I thought was a forced win, alas was not.

I'm not very experienced so feel free to ignore but I probably would have moved queen to c7, aiming to castle long which would put rook on open d file, then try to put bishop on d4.

1...Qd4 (double threat 2...Qxa1 or 2...Qxe4)

1...Qd4 2. Bb2

No idea if it's best, but on Qd4 there's:

Qd4 Bb7 Qxe4 Qxe4 Nxe4 and if Bxe5 then Nf2+ with Ng5 can pick up the bishop.

If instead after Qd4 Bb7 Qxe4 Re1 Qxe2 Rxe2 0-0-0 with Be5 seems to keep the pawn as well.

To answer the OP's question, as black I'm encouraged by white's advanced pawns which may become targets.  e4 and f5 sure, but possibly also on the queenside after some exchanges.

e5 seems annoying to babysit (and is making his bishop better than yours due to the color), so I'd probably just let it go one way or another and plan to win e4 (or c4) in a move or two afterwards.

But all this is moot if you can safely win the e4 pawn directly with Qd4.

As with any plan it's useful to see it from white's perspective.  I think as white I'd want to attack the kingside as my pawns and cornered king don't make my endgame pleasant.  I'd likely want to get in f6 to break up the king's position if I can.  Otherwise I suspect black is just better.

Intersting and ballsy Qd4.  I actually put my bishop on d4 and here is a the game.

I dont mind losing; in a tournement or money game I'd have played differently.

0-0-0 with a kingside attack is not a bad idea, I didn't think of that.

10.g5 offers to keep more kingside files closed, which is to white's advantage, better was g6 (and bad was white capturing 11.fxg6 opening lines for you).

On 32 there's no sense in saying it wasn't good because you point it out yourself.  What may be useful to you though is to be aware of this type of position where you are very strong and your opponent has no threats... so there's no need to be desperate (or flashy).  You have all sorts of threats you can work on, (h2, g2, pushing your e and f pawn) and white is so dead in the water you can even think about Kd7 and winning queenside pawns if that's what you really wanted (although of course a kingside attack is obviously called for).

Honestly I wasnt trying to be flashy.  as soon as 30 Nxa5 I figured white was expecting my to counter 31 Qxa5 with Nb7 and started running the calculation and with every angle I saw it as forced mate.  funny thing I was outside for minute and it hit me, the f'n pawn.  I came in and was like shiiiit.  Actually in the prior game he played the same f pawn ..manuevers as black and I punished him for it so I was perhaps in a hurry to repeat the process.  If I lose a game going out like that its fine with me.  

Unless there's some tactical reason, Qd4 seems right to me. White's poor queenside development has to be taken advantage of immediately before White has a chance to get those pieces out (when you have lead in development you should attack before that advantage goes away). Compare Black's bishop and knight to White's.

And White's three backward pawns are targets ready to be exploited. White has space but his pieces have no coordination and the front door is wide open for Black to step in a create havoc.

How to think (as in the title of the thread)? Just go through a strategic checklist. The one advocated by Silman in his book How To Reassess Your Chess would be fine, that is check the imbalances Material, Development, Piece Activity, Space, Pawn Structure, Square Control and Initiative.

There is an interesting tactic in the air : 1...Qd4 (fork) 2.Bb2 Qxe4 3.Qxe4 Nxe4 4.Bxe5 Nf2+ and Black wins the exchange as 5.Kg1 Nd3+ (discovered check) picks the e5 bishop

1...Qd4 is the first move I'm looking at because it's a forcing attacking move, and it also improves the position of the queen

Qd4 is your best choice, hitting the rook. And if he plays Bb2 then Qxe4 winning a pawn!

1...Qd4 allows 2.Bb2 Qxe4 3.Qxe4 Nxe4 4.Bxe5

Black can retake the pawn on e5 but Black can play for a draw here by playing 4...Nf2+ with perpetual check or White will exchange his rook for the knight. :D

I like to go on for 1...Qc7 followed by 2...0-0-0 since 1...0-0 2.bxc6 bxc6 3.Rd1 Qc7 is also unclear but Black is more or less gained equality in the game.

You don't mind being like "shiiiiiiit I just dropped a rook in a completely winning position"  ?   As an old chess saying goes, it's better to win a knight in 2 moves than go for a mate in 8 (i.e. go for the sure thing because you may have miscalculated the 8  move mate).

Maybe I'm just being bias from my own experience, I only mentioned it because that was a memorable moment of improvement for me when I started recognizing and labeling positions where my opponent had no threats.  At that point you can be as patient in the position as you like and play with a draw in hand.

But also in my experience these little nuggets of wisdom are more or less useless 99% of the time and only make sense when you discover it for yourself in a game.  Some of my personal (and fairly recent) "discoveries" include realizing the center is important in the opening

What do you mean a perpetual check?  As noted in posts 5 and 12 black is better in that line.

4...Nf2+ 5.Kg1 Nh3+ 6.Kh1 Nf2+ is a perpetual check.

What's the point of calling it perpetual check if it's not forced.  White can play Rxf2.

And even if white didn't, why woudln't black choose to win the piece with Ng4+ instead of taking a draw?