I failed Zden.com in 2002 (launched in 2000). Regrets turned into positive(?) lessons (I hope one day I can say that).
* When you are doomed, you are doomed. Better to start over than try to fix.
If there is good chance that the reason why you do not get traction is mono-dimensional (eg. just the UI or (exclusive) just performance...), it seems ok to go on & try to tweak it, but if you it is multidimensional (several reasons) (Technology, management, wrong market assumptions), then pull the plug ASAP. I did not have the guts to cut 3 fingers, so reality took my arm.
* Even if others "co-founders", I was alone (CEO with masteroftheuniverse equity stake), I had just came back from a 99 boiling BayArea startup, there was a wrong "guru" effect: no real co-founder (especially hacker) to tell me "this is wrong, you are kidding yourself". Selfcomplacent, mono thinking: doomed.
* Focus, focus, focus: you have to make hard choices so that you will NOT do X,Y,Z in order to become better at U. Not only experienced it for Zden, but another company I have worked (could have gone further quicker) for they did not have the balls (at least for 9years) to cut off either distractive product/features, or market sectors that are not going anywhere while another is booming etc.
"Everything for everybody": doomed
* Do not impose new behaviors on users/prospects (or just a super light one maybe) (worst is several at the same time). Users/prospects do not care about you & your service (by definition). We added complexities on top of another by introducing too many new "innovative" things (we thought) at the same time and users "just had" to do x... (1/ a new micro-currency, Zees instead of using plain dollars, 2/ the fact that people could sell or share their files (ecommerce + heavy bandwidth issues), 3/ that they had to classify their files in categories themselves (taxonomies spamming, low value content). Compounding complexities (for users/prospect or in general): doomed
Learnt that simplicity is very hard, it is an humbling ego bruising discipline.
* Do not accept to meet a VC (or key account etc.) overnight. You need to get prepared. Eg. Almost went the next day after to meet for my 3rd meeting Benchmark capital (after the 2nd they let me so much wait that I accepted meeting right away) and... got burnt. I would have studied John Newton bio (EIR at that time which I was meeting), I'd have discovered that a co he co-founded Documentum, he learnt the hard way about the necessity of a "Crossing the Chasm" (Geoffrey Moore) plan. Even after getting extra 5mn of his attention in the corridor, no Beachhead market focus: toast.
Show you are strong, do not say "yes" right away, successful people have busy agendas, any relevant professional will respect that.
* Finally, do not believe that by "believing in it so hard, (like in "against all odds/we believe/success" movies) in the end you will not fail". You can fail. One could say that as entrepreneurs that is our destiny (http://www.youtube.com/watch?v=45mMioJ5szc). That should always be a possibility in your head. "Will I?" creates better outcomes than "I will" study shows: http://www.psych.illinois.edu/~dalbarra/pubs/Wll%20I%20I%20w...
Outcomes of this not expected failure (before I always thought I had a good angel watching over me): I really thought all my juice was gone forever. But then I accepted, I failed, it was nobody else's fault. I just had to shut up. Restarted from scractch (literally zero, zip, nixs cash, I had invested evthg I had, was too ashamed to ask anybody for help) with 1st new non-CEO job which was... putting advertisements in thousands of mail boxes (for real, I worn shoesoles out!), to find realestate properties to sell, and then made my way up from there, one step at a time w/ the help of luck and others (that I thank).
Bottom line, took me 8 years to earn the right to tell the world this:
In the few coming weeks, we will launch another company. We are 2, I am the hustler, Francis he is the Hacker, and together... we are going to kick ass !! ;)