It’s that time of year again—that time when marketers start having sweet dreams about what they could do with a healthy budget in 2018, and a sense of foreboding about what they’ll have to do with the meager pittance they’re granted.

Some of you are thinking that 2018 might be the time to get rid of that old clunker of a website and start afresh. But how do you budget for a website these days in which websites do, and should, range in price from free to hundreds of thousands of dollars?

You probably have questions. So do we. Lots of them. We’ve grouped them into seven categories, but we counted 31 question marks in this post. When you’re working on your website budget, we think you should ask:

1. Do you really need a new website? Why do you think you do? How do you need it to be different? Is it a matter of arcane sitemap? Clumsy or outdated design? Functionality you wish you had? Are visitors not doing what you want them to do?

Although those can all be good reasons to develop a new website, none of them necessarily requires it. In many cases, you can adjust a sitemap, make design changes, build landing pages, fix metadata, and add or change functionality without scrapping the whole thing. The biggest factor here may be your development platform. If you’re still operating on a custom content management system somebody built you a few years ago, it’s probably time to move to WordPress or something similar.

Of course, you also need to beware of slapping a shiny, new landing page with sleek mobile functionality onto an old site with tired internal pages that confuse and disappoint your visitors. Doing so can actually make poor user experience worse—maybe worse than not doing anything at all.

2. Have you explored your current site in sufficient depth? Now’s the time to take a deep inventory of your site—to audit all your content, your URLs, your metadata, your analytics. Have you dissected the traffic, and where it’s coming from, on what devices, through what door, for how long? Do you know what keywords people are using to find you? Do you know how all this has changed over time?

And while you’re at it, have you taken a good, hard look at user experience? Do you know how different kinds of visitors engage with you? Even a little anecdotal user testing can help you see what’s working and what’s not. And don’t forget site administrators—the people who use your site every day. What are their challenges with the existing site? And will you have to plan for different user levels with different levels of access?

3. What else can you improve? For example: Was your website ever written by a copywriter? We mean a bona fide, write-it-on-her-tax-return writer. This can be a particular problem for hospitals and health systems, law firms, financial service companies, utilities, universities, engineering, and other large organizations with deep websites and a lot of diverse topics to address. There’s just so much content, and it requires input and review from so many subject matter experts, that the end result can be wildly uneven.

Too often, new websites are just a rearrangement of the same, old content. A new website is a fresh start—time to radically improve what you’re putting into the world. We’ve seen a lot of great new websites marred by old, amateur, buzzword-filled content. Don’t let it happen to you.

4. What do you want to protect? Even if you want to change everything about your old website, there are probably things about it you don’t want to sacrifice. For example, if being found in a Google search is important to you—and it almost certainly should be—you’ll want to, as best you can, preserve the organic search rankings your old site has earned. If you’re carrying over old content, be sure you pay attention to your metadata. You’ll also want to set up redirects to new content so people who find you in search results won’t run into 404 Not Found errors.

5. What does your website have to connect with? Do you need your site to work with a user portal? An intranet site? Fundraising software? A customer relationship management system? A blog, a microsite, a landing page, an online store? Your social media platforms? Ad campaigns? All of these connections take careful planning and time to implement. If you care about user experience (and you do), you want all those connections to be integrated as seamlessly as possible.

6. Have you budgeted for ongoing content, maintenance, and upgrades? Some marketers can still get away with brochureware websites. But a content strategy that requires frequent website updating can benefit most modern marketers—and even, perhaps, should be the central driving factor of your marketing communications program. When you’re budgeting, you need to remember what needs to happen after you launch: ongoing blog posts, video, landing pages, social content, e-newsletters, and other content updates. Now’s the time to start thinking like a magazine, planning your editorial calendar a year in advance so you can budget for the content you’ll want to create.

Not to mention technical tweaks, plug-in updates, and routine maintenance. Like an automobile, your website can be a huge asset or a huge liability; a pleasure to drive or a nightmare and a mess. Have a plan for keeping it tuned up.

7. What are your goals? This is the big one—the one that actually justifies the cost of the new site. What do you truly want your website to do for you, and how will you know when it’s done it? There’s so much data available that it’s easy to be overwhelmed with numbers. Pick three things you want your new website to do better than your old website, and figure out how you’ll measure them.

How could this affect your budget? You might want to do some benchmark research so you can measure awareness and consideration before and after launch. You might, as mentioned earlier, want to do some user testing with your old site to see where the bugs live, then again with your new site to make sure you’ve fixed them (and not created new ones).

And, frankly, sometimes big goals require big budgets. Not much is going to happen if you build a nice website and let it just sit there. You’re going to have to get people to visit.

In summary: Audit the heck out of your existing site. Know what you need to improve, protect, and connect with. Have a plan that’s both strategic and tactical. Make sure you really need a new site in the first place.

And get to work. The fact that you have a website at all means you need a great one. A new site is a new chance to be great.