I think that's reasonable. There are a bunch of different ways to do it.
It's worth remembering that referendums are good legitimacy-building tools, to some extent. They are awful policy-making tools. When you run a referendum, you are not looking to build a policy, you are looking to generate legitimacy for a policy that has already been submitted. That's worth keeping in mind.
I would say this. In British Columbia, the referendum actually worked fairly well. The government set an arbitrary 60% threshold to pass it. I think that was ridiculous and unnecessary, but 57.7% of people voted in favour of electoral reform. We forget how remarkable and how unusual that was. The referendum was actually run fairly well. Part of the reason was that the folks at the B.C. Citizens' Assembly, which was well funded, became educators. They went out into the public and essentially became points of contact for the public and taught folks what the option was and why it was good. They became heuristics. They became shortcuts for people.
There is a way to do that, but it takes a lot of time, a lot of money, and a lot of legwork.
What is one of the best predictors of whether someone is going to vote in an election? It's whether they have been contacted directly by a political party on their doorstep. We need to be doing this for electoral reform too.