How Long Before Marketing Works? Set a Threshold, Not a Deadline

The most-quoted timeline in marketing is a 2017 video about hiring

You’ll see this statement everywhere: Google says it takes four months to a year to see results from search.

It’s worth knowing where that came from. It’s a remark by Maile Ohye of Google in a video about how to hire an SEO, reported in February 2017. It’s nine years old. It advised engaging a consultant, not stating how long ranking takes. And it’s no longer on the Google page it’s usually attributed to.

Google’s current published wording is much shorter and vaguer. The SEO Starter Guide says some changes might take effect in a few hours, others could take several months, and advises waiting “a few weeks” before assessing whether your work helped.

A few weeks. Not four months to a year.

What the research actually shows

The other figure you’ll see quoted is that SEO takes three to six months.

That’s real, and it’s a poll Ahrefs ran on LinkedIn and X, not a study. It asked marketers for their impressions. It was published in 2022, and its poll responses predate AI Overviews entirely. The article concludes that there’s no definitive time frame.

The more useful number comes from Ahrefs actually measuring instead of asking. Across a million random URLs, only 1.74% of newly published pages reach the top ten within a year.

Then the finding that cuts against everything else: of the pages that did reach the top ten, 40% got there within a month.

Read those together, and the picture changes. Most new pages never rank, and the ones that do tend to show early. That doesn’t mean you should quit at four weeks. It means prolonged, total silence is usually information rather than a phase you’re waiting out, and a plan that treats every month as proof of patience is a plan with no way of being wrong.

How long each channel actually needs

Email to a list you already have: about two days. This is the one where patience is simply the wrong instinct. Across 4.4 billion messages GetResponse measured in twenty-four countries, including the Netherlands, 21% of opens and 44% of clicks land in the first hour, and roughly 70% of opens and 85% of clicks arrive within twenty-four hours. There’s no learning phase. If a send hasn’t worked in forty-eight hours, it hasn’t worked, and waiting three months to decide that changes nothing.

Paid search: thirty conversions, not thirty days. Google’s own guidance for evaluating an automated bidding strategy is to measure the last thirty days, including at least thirty conversions. That single line is more useful than any calendar. If you get four conversions a month, you’re seven months from a readable result, and no amount of patience shortens that. Only volume does. Separately, the learning period itself can run up to three weeks, or one to two conversion cycles, and it resets when you make significant edits, which is why constant tinkering keeps a campaign permanently unreadable.

Paid social: nobody can show you the source. The universally quoted rule is fifty optimization events in seven days. I looked for it in Meta’s own documentation. I couldn’t verify it because those help pages aren’t machine-readable, and every page quoting the figure was a third-party blog repeating other third-party blogs.

Search and content: months, with early signals. Three to six months is the working assumption, with the caveats above. Watch for movement in impressions before you watch for traffic.

Referral and word of mouth: unknown, and nobody has measured it. It’s the highest-rated channel among small businesses in LocaliQ’s 2026 trends report, with 83% naming customer referrals their best lead source, up from 65% the year before, though that survey is 87% North American. I could find no credible data on how long it takes to build. Referrals survive impatience precisely because they have no dashboard, so nobody sets a 90-day deadline.

Brand: years, with something visible inside a quarter. Mark Ritson puts brand building on a three-year-plus horizon against roughly twelve months for activation. But System1’s own analysis of high-scoring advertising finds it drives sales in the eight to ten weeks after airing, so even the slow work should produce a flicker inside three months.

What to actually check at 30, 60, and 90 days

Straight answer first: there isn’t an independently validated 30/60/90 framework for small business marketing. I looked for one from an academic body, a trade association, or a platform and found agency content. Anyone can present you with a tidy quarterly milestone chart they built themselves.

What’s defensible is the distinction between a signal and noise at each stage.

The pattern is that binary questions are answerable early, and rates aren’t. “Is it indexed?” has an answer in week two. “Is our cost per customer improving?” doesn’t have an answer until you have enough customers for the math to hold up.

Pre-commit to two things: a data threshold, such as thirty conversions or two hundred visits to a page, and a date you’ll look at. Decide which you expect to hit first. If the date arrives and the threshold hasn’t, that answers volume rather than the channel, and the fix is usually more spend, more reach, or a narrower target rather than waiting.

If you’re just starting, set the review dates in your calendar before you spend anything. The single most common failure isn’t impatience; it’s never deciding what would count as working.

If you have some traction, stop resetting the clock. Changing a campaign’s targeting, budget, or creative restarts its learning period, and a channel that has been “given three months” across four different configurations has actually been given three weeks, four times.

Frequently asked questions