Every redesign starts with a trigger. The trigger is almost never the real problem. A funding round closes and the site has to match the deck. A new sales hire says the homepage embarrasses them on calls. A competitor launches a new site. You open your homepage one night and cannot stand it. All of these are real. None of them tell you what the site is failing to do. The questions to ask before a website redesign turn that trigger into a diagnosis. Then the project you buy fixes what is actually broken, and you can tell afterward whether it worked.
We run these projects. We have watched founders pay for a redesign that rebuilt the old problem in a new layout. This is the list we wish they had worked through first. The order below is the order that saves the most money.
First: is it a redesign you need?
The word "redesign" assumes the design is the problem. Sometimes it is: the site was built in a rush, the product has outgrown it, mobile is broken, or the CMS is hard to use or unsafe. (The CMS is the content management system, the tool your team uses to edit pages.) More often, the site looks fine but says nothing. In that case the real problem is positioning, meaning what the site claims you do and who it is for. It only feels like a design problem.
Score the current page on five points before anyone talks about layout. The sentence: can a stranger say what you sell, and for whom, in one line? The buyer: does the page name who it is for? The proof: is there evidence a buyer can check? The path: is there one obvious next step? The measurement: can you see what the site brings in? Our scorecard does this in a minute. If the sentence and the proof score weak, you need positioning and evidence first. The layout may be fine. If the page order, the product presentation and the mobile experience score weak, you need the redesign. In that case the positioning work is the first two weeks of the project.
Either way, do not write the brief until you know which one you need. A brief that says "modernize the look and feel" will get a moodboard back. A moodboard is a page of colors and sample images. It does not answer the question you are paying to answer.
About the problem
What is the site failing to do, and how do I know? "It looks dated" is a feeling. "Inquiries dropped from twelve a month to four after the product changed" is a problem. "Sales says every prospect asks the same three questions on the first call" is also a problem. Write your problem as one sentence with a number in it. If you have no number, write a sentence a sales rep would agree with.
Which page do inquiries actually start on? Open your analytics. Look at which page a visitor was on when they filled in the form. If most inquiries start on a pricing page or a comparison page, the homepage may matter less than you think. Then the budget belongs on those pages. If nothing is measured, that is your first finding. Put it in the brief.
What do the last ten sales calls say the buyer thinks we are? Use recordings if you have them. Use notes if you do not. The words buyers use to describe you are the words the site should use. They are usually not the words the site uses now. The calls also show whether the site is attracting the wrong buyer. No redesign fixes that unless the sentence changes.
What has changed since the site was built? Think about the product, the buyer, the price, the market and the team. A site describes a company at one moment in time. If the company has moved on and the site has not, that gap is the brief.
About the sentence
Can we say what we sell, for whom, in one line? Try it now, out loud, without the deck. A good line has three parts: a category, a buyer and a change. It should also be a line a competitor could not paste onto their own site unchanged. If your line passes, the sentence exists, and the redesign can be built under it. If it does not, the sentence is the first deliverable, before any design. Then you need an agency that writes it. We explain why in positioning before pixels.
Does the site say that line today? Open the homepage and compare. Look at the H1, which is the main headline at the top of the page. The gap between the line you just said and that headline is the size of your copy problem. It is often bigger than the design problem.
Who are we deliberately not for? A redesign that has to work for everyone will end up saying nothing. You have to decide who the site is allowed to turn away. That is a founder decision. Make it before the brief. An agency will not make it for you.
About the proof
Which customers can we name? This means naming them publicly, on the site, with a logo or a sentence. Ask them now, not in week six. A real customer name on the page does more for a buyer than any design change.
Which numbers can we defend? Do not use the numbers from the deck. Use the numbers you would say to a skeptical buyer who might check them. Examples: screens deployed, tickets resolved, hours saved. Each one needs a method behind it. If you have no such numbers yet, decide what you will measure from now on. Then you will have numbers by the next redesign. Use named customers in the meantime.
Who will give a reference call or a case study? One case study you can click on, with a number in its title, does more than a wall of logos. If one customer would sit for an interview, put that case study in the redesign scope. Do the interview in week one.
About the scope
What stays? List the URLs that rank in search. List the pages that convert. List the content that took years to write. List the links to your CRM (the system that stores your leads and customers) and to your analytics. A redesign that breaks the URLs loses the search traffic that was paying for it. Write down what must survive. Put it in the brief as a hard requirement.
What is actually in scope? Is it the homepage only? Is it the homepage plus the pages that inquiries start on? Is it the whole site? Does it include the product screens, the docs and the pricing page? Most B2B redesigns that go wrong were scoped as "the website" and delivered as "the homepage and a template". Write the page list.
Is the platform part of the problem? The platform is the software the site runs on. If nobody on your team can edit the current site, the redesign is also a migration to a new platform. That changes the team, the timeline and the price. If the platform is fine, say so. Then no agency can sell you a rebuild you do not need. Our modernization work is for the cases where the platform really is the problem.
What is the budget, honestly? Know the number yourself before the first call. You can decide later how much to tell the agency. The budget decides which of the three kinds of vendor you are choosing between. We have written how to choose between them.
About the team and the time
Who writes the copy? If the answer is "we will", ask when. A founder-led company between rounds does not have a spare copywriter. The most common reason a redesign stalls is a finished design waiting for words. Hire a team that owns the sentence. Otherwise, accept that whoever writes the copy controls the timeline.
How many hours a week can I actually give this? Two hours is realistic for most founders. Say the number. Then choose a process that fits it. That means written deliverables, a preview link that updates, decisions recorded in a document, and calls only at the three or four moments that need one. A process built on weekly workshops will fail the first week your calendar fills up.
Who decides? Name one person. The hero is the first block at the top of the homepage. When a co-founder, a head of marketing and an investor all have to agree on it, the project takes three times as long. It also ends with the safest sentence, the one nobody objected to.
About after launch
What will we measure, and from when? Set up a conversion event on the inquiry form. A conversion event is a record in your analytics of each time someone submits the form. Add a source field so you know where each buyer came from. Take a baseline from the current site, so the new site has a number to beat. If the old site has no baseline, start measuring now, before the redesign. Even one month of data helps.
What does success look like in ninety days? Write down a number and a quality. For example, eight inquiries a month, with a first reply that is a scoped conversation rather than "what do you do". Write it before launch. Then the launch cannot change the target afterward.
What is the rollback? Ask what happens if the new site does worse. A team that has thought about this will have a plan. The usual answer is "we keep the old site ready to deploy for sixty days". Asking costs nothing.
Who owns the site after the agency leaves? Someone has to change the price, add a case study, or fix a typo. If the answer is "the agency, on a retainer", you are buying a dependency. If the answer is "nobody", the site will be out of date within a quarter.
What to bring to the first call
If you have answered the questions above, the first call with any agency, including ours, will be short and useful. Bring these six things:
- The problem, as one sentence with a number in it.
- The sentence you say out loud, and the H1 the site says now.
- Every customer you can name, every number you can defend, and the one who will sit for a case study.
- The page list, the URLs that must survive, and whether the platform is in scope.
- The hours a week you can give, and the name of the person who decides.
- What you will measure and what success looks like in ninety days.
Watch what the agency does with this list. If it comes back with a moodboard, it will treat this as a design job. If it comes back with a rewritten sentence and a re-ordered page, it will treat this as a positioning job. Pick the one that matches the problem you found in the first section.
Step zero
Before the brief and before the shortlist, have the current page read against a standard. The scorecard is the free version and takes a minute. The teardown is the paid version. You get a recorded walkthrough of your homepage against the five points. You get the sentence rewritten with three candidates. You get the page re-ordered on paper. It takes five working days. The fee is credited if you go on to work with us. Either one turns the trigger into a diagnosis. Write the redesign brief from that diagnosis and nothing else.