Most SEO programmes aren't killed by bad execution. They're killed in month two, when someone asks what the revenue looks like and the honest answer is nothing yet.
How long does SEO take to work? Most sites need six months or more before revenue is a fair test of whether it's working. That's not an excuse it's mechanics: Google has to crawl the changes, reassess the pages, and see whether users actually prefer them over time, and none of that happens on a monthly reporting cycle.
The mistake isn't being slow. It's measuring the wrong thing early, then drawing the wrong conclusion from it and losing the budget before the programme had a real chance to work.
This is the foundation phase, and almost none of it is visible in a normal dashboard. Track indexation (are the pages you shipped actually in Google's index), crawl health (is Googlebot reaching the pages you care about without errors), and query breadth in Search Console the raw count of distinct queries a page starts appearing for, even at position 60. Query breadth is the earliest real signal you have: it means Google understands what the page is about before it has decided where to rank it.
Rankings will be volatile and mostly bad in this window. That's expected, not a warning sign.
Position bands start to matter more than exact rank track what share of your target queries sit in positions 1-10, 11-20, and 21-50, and watch the bands shift over time rather than obsessing over single keywords. Non-branded impressions should be climbing steadily as Google trusts the pages with more queries and more visibility. CTR on the queries you're already ranking for tells you whether the title and meta are doing their job independently of position. And engaged sessions from organic search not raw traffic tell you whether the people arriving are the right people.
If these four are moving in the right direction, revenue is coming. If they're flat, that's the moment to change something not month one.
Somewhere around month six, enough pages have matured and enough of the early signals have converted into rankings that revenue becomes a fact you can actually attribute to the work, instead of noise. Before that point, judging SEO on revenue is like judging a garden by digging up the seeds to check on the roots.
Attribution still needs care even after month six — organic-assisted conversions, branded search lift from content, and multi-touch paths all complicate a simple last-click number. But at least by then there's enough data for the conversation to be honest.
Everything above is standard advice, and it holds up but "month three to six" is still a promise until someone shows the receipts. So here they are: monthly organic-search sessions, pulled directly from Google Analytics, across five accounts I manage — my own site plus four active clients. Not case-study screenshots. A live query, re-run any time.
Client names are left out the data isn't mine to publish, only mine to learn from. Three more accounts I have access to aren't in the table below: two have never recorded a single organic session in Google Analytics, which is its own finding (a tracking gap, not zero traffic), and one is too new to show a trend at all. Leaving them out quietly would have made the table look tidier than the reality.
AccountMonths trackedGrowth (first 3 months avg → last 3 months avg)Client A34 (Oct '23–Jul '26)5.35×Client B22 (Oct '24–Jul '26)2.46×Client C18 (Feb '25–Jul '26)1.67×Client D16 (Apr '25–Jul '26)1.49×My own site22 (Oct '24–Jul '26)1.65×
None of that growth was fast, and none of it was a straight line. Client A took close to two years, not six months, to reach its clearest sustained level with a real dip along the way, left in rather than smoothed out. Client C shows a month where organic sessions collapsed to almost nothing before recovering; I checked it against the unfiltered data rather than assume it was a fluke, and it held up. That's what month three to six actually produces: a slow, uneven climb, not a switch that flips.
This isn't a controlled experiment there's no isolation of what changed in each account and when, and it measures visibility, not revenue: conversion-tracking quality varies too much across these accounts to use it honestly here. What it does show is real, and it's the kind of evidence "six months" claims usually skip.
This conversation belongs in week one, not month three when leadership is already asking hard questions. Agree on the leading indicators you'll report on for the first two quarters, and agree separately on when revenue becomes the metric that matters. Write it down. The KPI you agree to in month zero determines whether the programme survives long enough to work.
When I started with GRASP, the first phase wasn't content or links it was wiring up conversion tracking so every channel could be judged on qualified demo requests instead of vanity traffic. In a category that new, getting the measurement right first was the difference between proving progress and guessing at it.
If you're building the roadmap that sets these milestones, my SEO roadmap guide walks through the full plan structure. And if the harder part is getting leadership to agree to these terms in the first place, that's exactly what I cover in securing SEO buy-in.
Email me if you want a second opinion on what you should be measuring right now, and whether it's realistic for where your programme actually is.
Mohamed Sahardid SEO & GEO Specialist
mo@sahardid.com · LinkedIn