A version of the same question comes up on our calls most weeks: my new site isn't live yet — does it make sense to wait, or doesn't it matter? It comes from accountants mid-rebuild, from agencies whose own site is the last thing they ever get to, from web designers quietly embarrassed that theirs is the worst one they've made. Underneath it is a worry that outreach without a website is faintly humiliating, or simply won't work.
Search for the answer and you get a hedge: you don't strictly need one, but you really should have one — usually buried as step five of a nine-step guide to starting an agency, and usually published by a company that sells websites, hosting, or agency software. That advice isn't wrong so much as imprecise, and imprecision is expensive here, because it sends people off to spend three weeks building a brochure site to unblock something the brochure site was never blocking.
The question is really two questions with different answers, because two separate parties look at your web presence. One is a set of systems — mailbox providers, the registry behind business texting, eventually a payment processor — checking whether a real business exists and can be identified. The other is a human who just got a message from a stranger and is deciding whether to reply. They inspect different things, at different moments, and they are satisfied by completely different artifacts. Answer them separately and this stops being complicated.
What actually inspects your domain when you send email
Nothing on the machine side reads your homepage.
What email systems check is DNS. Your sending domain needs records proving mail can be routed to and from it, and authentication records — SPF, DKIM and DMARC — proving the mail claiming to come from you actually did. Receiving servers also check that the sending server's own address resolves to a hostname and back again. None of that involves a web server: a domain with clean authentication and nothing at all on the root passes every one of those checks.
There is a real effect people mistake for this. Newly registered and long-dormant domains get treated cautiously for a window after they start sending — which is why the advice is to buy sending domains early and ramp slowly. That's a penalty on newness, not on sitelessness, and it doesn't go away because you put up a page.
You'll also run into the claim that a parked domain sending outbound is itself a filter signal. It circulates without a published rule, a vendor document, or a test behind it, and the pages carrying it generally sell the fix. What's worth separating are the three states it collapses: a root that doesn't resolve at all, a root serving a registrar's ad-stuffed parking page, and a root redirecting to your real site. The middle one is the only one with an obvious downside — and it's what you get by default when you buy a domain and do nothing with it.
Where the answer genuinely changes: text messaging
If your outreach goes out by text, this is the one place a public URL is genuinely load-bearing — and the part the credibility-focused articles never mention.
Texting businesses in the US means registering your messaging campaign before you can send, and what registration reads is business verification and policy documents. Requirements differ by how your business is registered and which platform files on your behalf — sole-proprietor paths ask for less than standard ones, and individual providers apply stricter rules than the registry's minimum. The common rejections are mundane, and worth knowing before you submit rather than after: a URL that doesn't load, a URL that doesn't identify the business named on the registration, and content behind a login. Where the material isn't publicly reachable, the usual remedy is to host it at a public address, not to build a website around it.
So the binding requirement is narrower and far more specific than "have a website": a publicly reachable page carrying your privacy policy and terms, under a business identity matching the one you registered. Much smaller than the site people go off and build — and unlike a homepage, not optional. Confirm your provider's current requirements before you submit, because this is a carrier and platform process and it changes. This article is general information, not legal advice; talk to counsel about your own situation.
Skip the guesswork on your setup
A free 15-minute demo: the actual businesses we'd reach in your niche, the actual messages, and what your outreach needs in place before day one. At least 3 booked appointments in your first 30 days or you don't pay.
Book your demo →The buyer looks you up later than you think
Now the human side — where nearly all the ranking advice lives, and where almost none of it is measured.
The premise everywhere is that prospects Google you before they reply. Stated as fact, that's folk wisdom told from the sender's chair — the sender can't observe it, and the datasets waved at it are sending datasets, which structurally cannot see what a recipient did in another tab. We're not going to invent a number for it either.
What you can reason about is when the lookup happens, and it's a second-stage filter. A cold message has to clear attention first — relevance, timing, whether it reads like it was written to a person — and only then does anyone bother investigating who sent it. Nobody researches the sender of a message they already dismissed, which means everything you build upstream of that reply is doing nothing to earn it.
It also means the load-bearing statistic in the "you need a website" argument is doing less work than advertised. The figure usually quoted — that roughly three-quarters of people judge a company's credibility from its website design, attributed to Stanford — isn't what the research found. The Stanford study behind it (Fogg et al., 2003, 2,684 participants) reported that "design look" appeared in 46.1% of participants' comments about credibility. Those participants had been sat down in front of two live websites and asked to evaluate them. It's a finding about how people judge websites they are already looking at, from a year before the smartphone, and it says nothing at all about whether a stranger who just received your text will go looking for one.
When someone does look, they run two checks in order. First the fraud check: is this a real business, is there a human attached, does the name match the message I just got. Then the competence check: do they work with people like me, has anyone else hired them and said so. Different evidence, and only the first can be failed instantly.
Your channel decides what they land on
Here's the part no one covers: the artifact that matters is whichever one your channel points at, and the channels point at different places. If you texted or called, they don't have your domain at all — they have a business name, so they search it. Only email hands them a domain to type, which is where the advice to redirect a sending domain to your main site comes from, and where it quietly becomes circular for someone with nothing to redirect to.
| Channel | What they search with | What they land on first | Minimum that survives the check |
|---|---|---|---|
| Cold SMS or cold call | Business name and a phone number | Search results for the name — business profile, reviews, social | A claimed business profile, working phone, consistent name |
| Cold email | The sending domain and your signature | Whatever the domain serves, or the signature link | A domain that resolves somewhere real and matches your signature |
| LinkedIn or DM | Nothing — your profile is already open | The profile itself | A profile that looks lived-in: history, a face, some activity |
One mismatch in the standard advice is worth naming. Deliverability guidance says send from a secondary lookalike domain; credibility guidance says be findable. A prospect who types your sending domain and lands on a name that doesn't match your signature has failed you on the fraud check, not the competence one — and the fraud check ends threads silently. Which channel to lead with has its own economics, covered in what an appointment really costs on each channel.
Run the thirty-second test. Open a private window on your phone, search your business name the way a prospect would, and start a timer. Can you tell it's a real business, find a human, and get a working phone number inside thirty seconds? That's the whole bar for the fraud check — and most people discover the problem isn't a missing website but an unclaimed business profile, or a number nobody answers.
What "no website" actually signals — and why it reads differently on you
We have an unusually direct view of this from the other side of the table, because a large share of what we run is outreach for web designers — whose entire market is defined by this exact state.
Businesses with no website are a visible, targetable segment, and designers prospect them deliberately. What surprises people is how many experienced designers refuse to. Several have told us flatly they'd rather message a business with an ugly, outdated site than one with no site at all, and the reasoning is consistent: a business that has gone years without a website usually doesn't believe it needs one. The absence isn't reading as unprofessional. It's reading as not in the market — a targeting judgment from the same family as the other filters that keep a list clean, covered in qualifying filters that keep junk off your calendar.
Sit with that, because it cuts the opposite way for you. "No website" is a useful filter precisely because it predicts disinterest, and it works because the designer is choosing who to approach out of thousands who never asked to be contacted. Nobody is running that filter on you. Your prospect isn't screening a list — they're answering a message you sent them. You already skipped the step where sitelessness would have counted against you. (If you're on the other side of this and sell websites for a living, that market is laid out in how to get web design clients and on our web design page.)
The version of the worry that does apply to you is narrower: if a prospect goes looking and finds nothing — no site, no profile, no listing, nothing tying the name to a human — that's a fraud-check failure, and it lands right after they got interested. Worth an afternoon. Not the same project as building a website.
The minimum verification surface, in the order to build it
Replace "it depends" with a sequence, where each item unblocks something specific.
One: a domain that can send and receive mail, properly authenticated. This unblocks email outreach entirely, and nothing needs to be served on it. Two: a publicly reachable privacy policy and terms — only if you're texting, but then it's the thing registration actually reads, and it has to sit at a public address under your registered business name.
Three: one page that names the business — what you do, who you do it for, a working phone number and email address, a real address. This is what a curious buyer and most reviewers both accept. Four: a claimed business profile, if you're prospecting by phone or text, because that is the surface your channel actually points at. Five: proof — one named result you can show, the only element every practitioner and every vendor guide in this space agrees on.
The traps that cost a resubmission are all boring: a URL that doesn't load, a domain that doesn't match the registered business name, anything behind a login, and public link shorteners, which registration generally won't accept. Fix those before you fix your typography. Our current pricing sits on the pricing section rather than in this article, because it changes and this sequence doesn't.
Four events that mean it's time to build the real thing
Page one gives you a hedge instead of triggers. Here they are. Build the real site when a referral partner has nothing to forward. When a payment processor or ad platform reviews your online presence before approving you — they generally want the business name, what you sell, a support contact, and your refund terms, reachable without a password. When a prospect asks for a link twice. And when you start getting inbound you didn't originate, because that's the first moment a site does work outbound can't.
Until one of those fires, the website is a verification artifact, not a lead source — and the order matters more than it looks. A site built before you've had a hundred conversations with your market is a guess about who you serve. A site built after them is a transcript. Build the smallest thing that survives a thirty-second check, go send, and let the real one wait until the outreach has told you what belongs on it. For the message side, the cold SMS playbook covers what to actually put in the text.
Your site can wait. Your pipeline probably can't.
Fifteen minutes, no charge: we'll show you the real businesses in your category, the messages we'd send them, and the profit math — whether or not your website is finished. 3 booked appointments in your first 30 days or you don't pay.
Book your demo →Frequently asked questions
Can you do cold outreach without a website?
Yes, and the constraint is narrower than the advice suggests. Nothing on the email side inspects your homepage — mailbox providers read DNS and authentication records, and a properly configured sending domain passes every one of those checks with no web server behind it. Business text messaging is the exception: campaign registration verifies your business identity and reads publicly reachable policy documents, so you need a page carrying those, under the same business name you registered. The buyer's side is a separate question, and there the artifact that matters is whichever one your channel points at — usually a business profile if you texted or called, and the sending domain if you emailed.
Does a cold email sending domain need a website on it?
Not for deliverability. The checks that decide whether your mail is accepted look at DNS records, authentication, and the reputation of the sending infrastructure, none of which involve a web server. The claim that a parked domain is itself a spam signal circulates widely without a published rule or a test behind it, and usually appears on pages selling the remedy. There is a real recipient-side reason to point the domain somewhere: a person who types it into a browser and finds nothing, or worse finds a registrar's ad-stuffed parking page, learns nothing good. That is why redirecting a sending domain to your main site is standard practice — a recipient experience decision, not a filtering one.
Do you need a website to register for business text messaging?
You need less than a website, but this is the one place in the whole question where a public URL is genuinely load-bearing. Registration verifies that the business exists and looks for reachable policy documents, and the requirements differ depending on how your business is registered and which platform submits on your behalf. The most common rejections are mundane: a URL that does not load, a URL that does not identify the business named on the registration, and content sitting behind a login. Where the material is not publicly reachable, the usual remedy is to host it at a public address rather than to build a site around it. Check your own provider's current requirements before submitting, since this is a carrier and platform process and it changes.
Is a landing page enough, or do you need a full website?
For outreach, one honest page usually clears both checks a prospect runs. It has to name the business, say what you do and who you do it for, carry a working phone number and email address, and be reachable without a login. Add your privacy policy and terms if you are texting. What a single page cannot do is carry proof, and proof is the one element every operator and every vendor guide agrees on — a named result a prospect can verify does more for a reply than a redesign does. Build the page, send the messages, and let the real site wait until it has a reason to exist.