Every agency loses a client somewhere between month three and month five. Not because the work was bad — because nobody explained what month three was supposed to look like. The client budgeted for patience they never agreed to.
Explaining SEO timelines isn't a soft skill you deploy when a client gets grumpy. It's a system you build into the proposal, the kickoff, and every report you send. Here's how to do it so the "is this working?" email never arrives.
First, the honest answer: how long does SEO take?
Google's own guidance says four months to a year before you see measurable impact from changes. That's a useful anchor because it's not coming from you. But it's too vague to plan around, so break it into the components clients actually paid for:
- Technical fixes: 2-6 weeks to see movement, assuming the fix ships. Restoring 400 accidentally noindexed URLs shows up fast. A site speed improvement might barely register.
- On-page optimization of existing ranking pages: 3-8 weeks. Pages already sitting at positions 8-15 with real impressions are the fastest wins in SEO, full stop.
- New content: 3-14 days to index, first impressions within 2-6 weeks, meaningful position typically month 4-8. Competitive commercial terms can take 12+ months.
- Link acquisition: effects compound over 3-9 months and are almost impossible to isolate per-link.
- Domain-level authority gains: 9-18 months. This is the one clients feel but never see in a report.
Two sites with identical budgets will move at wildly different speeds. A 40-page B2B SaaS site with a clean crawl and a 5-year-old domain might see meaningful lift by month three. A 12,000-URL ecommerce site with faceted navigation chaos will spend the first ten weeks just getting Google to crawl the right things. Say this out loud in the pitch, with the client's actual site as the example.
The real reason clients get impatient (it isn't speed)
Run a mental audit of your last three churned accounts. In most cases the client wasn't upset about the pace — they were upset about the silence. Six weeks of invisible work, a PDF with a traffic chart that looks flat, and no sense of whether anything is happening.
Impatience is an information problem. A client who can see 14 tasks completed, 9 pages published, and impressions up 22% will wait. A client who sees a flat sessions line will not, no matter how many times you say "SEO takes time."
This is why real-time reporting beats the monthly PDF for retainer work. The monthly cadence creates 29 days of anxiety followed by one day of defensive explanation. Continuous visibility replaces the anxiety with a sense of momentum.
The three-layer model: teach it once, use it forever
Give clients a mental framework on the kickoff call and refer back to it in every conversation for the next year. Three layers:
Layer 1: Work delivered (moves weekly)
Audit items fixed, briefs written, pages published, links placed, redirects mapped. This is entirely within your control and should never be flat. If layer 1 is flat, the client is right to be angry.
Layer 2: Leading indicators (moves in 3-8 weeks)
Pages indexed, crawl requests, impressions, keywords ranking in the top 50, average position on target terms. These are the early smoke signals that layer 1 is working. They move long before revenue does.
Layer 3: Lagging indicators (moves in 4-10 months)
Organic sessions, conversions, pipeline, revenue. This is what the CFO cares about and what the client will judge you on eventually — but it's the last domino, not the first.
Now the sentence that saves relationships: "Layer 3 is a function of layer 2, and layer 2 is a function of layer 1. When you ask if SEO is working in month two, I'm going to show you layers 1 and 2, because layer 3 doesn't have enough data yet to tell either of us anything true."
Build the timeline into the proposal, not the crisis call
The best time to explain SEO timelines is before the contract is signed, when you have leverage and no one is defensive. Put a month-by-month table in every proposal with three columns: what we deliver, what moves, and — critically — what does not move yet.
Example for a mid-market ecommerce retainer:
- Month 1: Technical audit, GSC/GA4 setup, keyword map, 6 briefs. Nothing moves. Expect flat traffic.
- Month 2: Priority fixes shipped, 6 pages live, internal linking pass. Indexed pages and crawl rate improve. Traffic still flat.
- Month 3: 6 more pages, first on-page optimizations on existing money pages. Impressions up 15-30%. Small ranking movement on long-tail.
- Month 4-5: First conversions from new content. Average position on 20 target terms improves. Traffic up 10-20% YoY.
- Month 6-9: Compounding. Top-10 positions on head terms. This is where the retainer starts paying for itself.
Then ask a closing question on the kickoff call: "So in month two, when traffic looks flat, are we going to be okay?" Make them say yes out loud. That verbal commitment is worth more than any clause in the SOW.
Keep the same milestone plan visible inside your project workspace so the phases in your PM tool mirror the phases in the proposal. When your timeline lives where the work lives — which is the whole point of running client projects in something like PeakKR rather than a generic task list — the client can self-serve the answer to "where are we?" without emailing you.
Analogies that work (and one that doesn't)
Analogies aren't fluff — they give a non-technical client something to repeat to their boss when the CEO asks why organic is flat.
- Compound interest: "The content we publish in month two is still earning in month twenty. Paid stops the day you stop paying." Good for finance-minded stakeholders.
- Physical therapy: "You do the exercises for six weeks before the pain goes. Stopping in week three doesn't get you half the result — it gets you zero." Good for clients who want to pause the retainer.
- Auditioning, not buying: "Google isn't refusing to rank you. It's collecting evidence. Every crawl, every impression, every click is another data point." Good for explaining why patience isn't passivity.
The analogy to avoid: planting a garden. Everyone uses it, it implies you can just wait and nature does the work, and it quietly excuses you from showing progress. Clients hear "we're going to do nothing for four months."
The month-five call: what to actually say
It'll happen anyway on some accounts. Handle it in this order:
- Don't defend. Diagnose. "Let's look at it together" beats "actually, if you look at impressions…" Open the data live.
- Restate the layer model. Show layer 1 and layer 2 first. If both are healthy, you have evidence, not excuses.
- Name what went slower than planned, and why.

Nick Quirk

