Here's my CnC, remember that anything i do not write about is considered as fine.
I'll also be assuming you made
your actual signature and i'll compare the two when possible.
Firstly, the biggest issue of the image is the lack of depth: as a viewer i only one background layer, containing a background, a character and some effects, plus a second layer on top of everything composed of a black blur on the left and bottom side, some main text and a hardly visible secondary text.
Ignoring the content of each layer, this should be avoided as it gives a "printed on paper" feel on the image.
Now, the first layer (the one with Sej), ignoring the black blur, has a fine amount of blending, however placement and flow are lacking.
Placement lacks because the left side of the image is mostly empty compared to the right, the render should've been placed to the left more.
As for flow, it's lackluster as the blue lines coming out of her body lead to the left and so do the background lines, however the ones on her arm point to the top; this also causes composition's quality to fall as it seems mostly messy. You also forgot to remove a little white brush next to her arm.
Comparing this layer with your signature's image is quite effective here: you see that there are no overly empty spaces, almost every part blends together (the top right corner not so much) and there's a general flow of going to the top, which is also strengthened by the form of the character.
As for the first layer, the blur is done poorly on the left side as it's not well distributed.
On top of that, the text acts as a divired rather than connecting the two layers: the colors of the main text are way too bright compared to the rest, the shadow you've put underneath it absolutely mutilates depth (as our brain would think that if there's a shadow quite similiar to the object, then beneath it there's a flat surface) and the white outline(which isn't really an outline but whatever) makes the text feel old and childish.
As for the suidical graphics text, i understand it's a watermark you chose to apply, however there's a fundamental flaw.
Question is, why do you place a watermark? Simple, to be recognized as the autor of something. Now, if you put a watermark you usually have two thoices: the first is to make it almost invisible, just to annoy who tries to steal it or who actually uses the image (this option isn't recommended), while the second is to pretend you were requested to implement it inside the image as a whole, so you disregard its size and "weight" but you make so it blends perfectly in, this way it's like you'd say "look, i'm a l33t pro sig maker" (preferred choice).
Now, the issue is that your watermark doesn't fall into either of the two categories!
Comparing the second layer to your signature's image, you can see the absolute difference: the text in the signature has the same colors of the image and it actually connects the message you wanted to deliver (the text's meaning) with the bottom layers in a seamless way.
My suggestions to improve is to try your hardest to imagine how the whole image will look, the develop it from the most important point.
Anything that you place on top of already existing layers should look like an extension of whatever is underneath it.