I think you're right about elections being too small, both from a business perspective and from an impact perspective. The problem we were trying to solve was that people vote blindly for most lower-ballot positions. We wanted to make it easy to vote intelligently all the way down your ballot, without researching each candidate individually. I guess it's understandable we couldn't find a way to support a business doing that.
I wonder if it might be preferable to make it easier to put things on the ballot instead. You could look into the process for that and see if there's a way to humanize it or make it more transparent.
I can't think of a way to business model that, though. Maybe a Kickstarter deal where people vote for inclusion via pledges and you take a cut? That doesn't sound like a great idea. Dunno.
Elections are a really bad example of governance, mostly because they're a minimalistic form. It's really not worthwhile, IMO, to go after anything to do with elections at all if you want to make money. Instead, try to figure out a way to loop the government into dealing with everyday problems. Tax-help companies, like H&R Block and TaxAct, are actually early instances of this: they're providing navigation of arcane bureaucratic nightmares. There is probably no end to the number of areas you could find to niche up and optimize a citizen's interface with their governments.
I mean... I just find your comment odd. Elections are such a tiny slice... what problem are you even trying to solve, besides "how do I extract money from this process"?