The Vote Yes For C and D campaigns, which I spent election day working for, was one of the best ones that I have seen.
A big share of a good campaign is basically non-partisan work that election officials should do, but don't do, or don't do well. For example, when a voter was unsure of where to vote, we could ask their address, get a map on the screen of a computer, and not only tell them which precinct they were in, but how to get there from their home. We spent a lot of time reminding people we had reason to believe would support us (by virtue of location or party affiliation or contacts with us) to turn in the absentee ballots that they had requested. We found people to get voters with no viable transportation to their polling places, or to get ballots to them if that wasn't feasible. We answered questions of people who weren't sure what to do, for example, if they had requested, but not received an absentee ballot.
Like all political campaigns, a big part of the effort at the grass roots level is vapid. The campaign spent large quantities of time, money and effort, simply putting up signs all over the state saying "Vote Yes For C and D", and waving them at street corners. If voters were simply logical analysts of their own interests, signs like this shouldn't matter at all, because they contain no information in and of themselves except for a motto ("Moving Colorado Forward") so vague it could have stood for just about any issue or candidate. But, signs are basically mini-endorsements. If people in your neighborhood put up signs, then maybe you'd agree. If they don't, maybe you shouldn't care either.
This isn't to say that there weren't more substantive messages. TV spots, editorials in the paper and press releases and events designed to lead to free media all, at least, impart some idea of what the issue is about, which you wouldn't even know from looking at the signs. Lining up endorsements and getting the word out about them also mattered. So too did talking with a small proportion of people who actually want to discuss the issues, in debates and door to door campaign conversations. But, a remarkably large share of voters barely read the papers, don't attend political events and have never talked to anyone about the issues, so the basics have to be established before anything else matters.
Ultimately, I am not a person who thinks that what is wrong with our political system is that we spend too much money. Indeed, I think that we spend far too little money on informing people about elections and making them run smoothly, which is what campaign money goes towards doing. (Yes, there is the issue of money going to misinform people too, of course). In economics, spending money on a fixed prize at the end of the rainbow is called rent seeking, and this kind of behavior is a concern because money spent seeking rents is often not productive itself (i.e. campaign money doesn't provide government services). But, this certainly wasn't the case for Referenda C and D. Both sides combined spend about 1% of the amount of money at stake on campaigning. If twice as much spending would have left the public better informed, I'd be all for it.
The problem is not that there is too much money, but that the money is too hard to raise and that it often comes with strings attached.