I know you say you're not the idea guy, but that's basically how this comes off. Kudos for putting yourself out there to get advice and feedback though. Not knowing much about your background, here's everything I would tell a non-technical person who wants to do a startup (as a non-technical person doing a startup myself):
Ideas are worthless until they're validated. Get out, talk to customers in a market that you think you can make an impact in. If it's something that you care about, are knowledgeable in, or if there's a problem that pisses you off that you think you could solve with your startup, those are good places to start. Still test in the market though, because you're only one data point. In the words of Steve Blank (go read everything by him), get out of the building! ...or at least get on the phone to start.
If you're trying to build something that appeals to the general population, just go stand at a subway or bus stop where people are just waiting and have nowhere to go and ask them questions. People are usually bored there anyway. If you're building a B2B product, call business and set up appointments to meet them. Learn customer development.
Put up an Unbounce "coming soon" page and start engaging with folks that care about the problem you're trying to solve. Start a blog on Tumblr (better sharing features, but any blogging platform would do). Link the blog to your "coming soon" page and vice versa. Get your social media pages in order (Twitter, Facebook, LinkedIn, Pinterest if it makes sense, Quora etc.) ...with a marketing background, this should be your bread and butter. Use all of these to be authentically helpful. This will drive traffic back to your site.
Eventually, you should have generated a bit of a community if you're solving a problem that people actually care enough about. Use this to keep testing ideas and iterating on what you already know. Ask if this is an idea that these prospective customers (whether they know they are or not) would pay to have solved. If so, ask how much they think the average person in their profession spends, in time or money, to deal with the problem with current methods. Ask how much they think the average person in their profession would spend to solve the problem.
Try to learn to code (CodeAcademy & Udacity are good places to start). If you learn enough to build a prototype, fantastic. If not, you will at least have a new respect for the art. Start with HTML and CSS... this isn't the hard stuff, so you should learn it anyway and, if you're handy, you might turn into a serviceable front end dev. Read dev blogs, poke around on GitHub and StackOverflow, and even read documentation. This stuff won't make any sense at first. Take it slow and after awhile you'll develop an understanding of how all the pieces fit together. You don't actually need to be able to build anything with it, but having an understanding of what language or tool is used for what sort of task will help keep you from getting completely lost in technical discussions.
Learn stats cold, and learn tools to do it well (R is all the rage, but even Excel will get you far). As a marketer, knowing how to handle data will give you solid insights into how to tailor the experience to users and will put you on solid footing with your technical cofounder. It's hard to argue with data. You can even go ahead and start calling yourself a growth hacker.
Learn design, or at least try to learn a bit. This will help you better understand what makes one UI better than the other and is an important component to UX.
Go to meetups at first to bounce ideas off people, crystalize everything you're learning, and tell folks about what you're working on. Don't let it get in the way of actually getting things done though. These things can be productive and sometimes lead to chance meetings with people who can really help you or know someone who can, but they can also be distracting.
Wireframing isn't hard. You should at least learn that much. In fact, you can do a lot of work to validate your idea just by talking to prospective customers. There are ways to build wireframes and run usability tests that don't require coding. If you're creative, you can even use things like Mechanical Turk to fake backend processes (ex. Testing an app that aggregates real estate listings. Someone submits a request on a form you put up on Wordpress. You take that and turn it into a Mechanical Turk job... this may be a bad example. I just came up with it now.)
Here are a couple of things that will help you get started:
https://www.udacity.com/course/ep245
http://nathanbarry.com/finding-ideas-project/
http://steveblank.com/2011/09/22/how-to-build-a-web-startup-...
http://businessmodelgeneration.com/
http://theleanstartup.com/
http://assets.unbounce.com/s/images/noob-guide-to-marketing-...
http://blog.kissmetrics.com/infographics/
http://www.slideshare.net/dmc500hats/startup-metrics-for-pir...
If all of this seems overwhelming, don't worry. You don't learn it overnight and it's easier to learn in the context of actually doing a startup. Get around some good people who are working on startups as well, and make sure to spend time with people who aren't to defuse once in awhile.
I think a decent path would be to leverage the strengths that you have now around marketing by getting a firm understanding of statistics and how to measure growth. Start working at a startup that's big enough to be able to take on a junior employee and use that time to learn everything you can, work on side projects, and try to find mentors among the founders. They've already done it, so they'll be able to guide you in your journey.