Why SEO Can’t Be Judged on ROAS: Payback Curves and Verdict Dates

What you'll take away

• ROAS works where the click and the purchase happen in the same session at a fixed price. It misleads everywhere else.
• Revenue isn't return. A hundred and twenty thousand at twenty per cent gross margin is twenty-four thousand of actual money.
• Measure cost per qualified lead, cost per booked job, and payback period — how long a channel takes to return more than you put in.
• Lifetime value changes the answer. If a client brings repeat work and referrals, you can afford to pay far more to acquire one.
• Over twelve months, paid and SEO can produce the same number of jobs — but SEO's arrive almost entirely in the back half.
• Don't cut paid to fund SEO. The cheap cost-per-job only exists because the earlier months happened.
• Verdict dates: three months for paid, six to nine months for SEO. Compare year over year in a seasonal business, never month to month.
• Stopping SEO looks like nothing happens for three months, then costs you six months later with no event to point at.

---

Transcript

Ray: Welcome to The Visibility Room. I'm Ray — I lead paid media and advertising at Mediaforce, and this is episode six. Today: why SEO can't be judged on ROAS. Fast channels, slow channels, and how long marketing actually takes to pay you back. My guest is Tessa, who co-owns a design-build home renovation company in Calgary and has been in the business sixteen years. Tessa, good to have you.

Tessa: Thanks, Ray.

Ray: So you came in with a question that I think a lot of people are quietly sitting on and never ask out loud.

Tessa: Probably. Okay, so — we run Google Ads and we pay for SEO. Every month I get an ads report with a return number on it, and it's a good number. And I get an SEO invoice with a list of things that were done, and no number at all. And six months in, I keep asking myself: when am I actually allowed to judge this?

Ray: That's the right question, and the reason nobody gives you a straight answer is that the honest answer has two different dates in it, not one.

Tessa: Two dates.

Ray: Two dates, two metrics, two scorecards. And the mistake almost everybody makes — including some agencies, to be fair — is trying to force one number across both. Usually that number is ROAS, and it's the wrong tool for half of what you're measuring.

What is ROAS and when is it the right metric?

Tessa: Alright, define it for me properly, because I use it and I'm not sure I could defend it.

Ray: ROAS is return on ad spend. It's revenue divided by what you spent to get it. So if you spend two thousand dollars and it produces ten thousand in revenue, that's a five to one ROAS. It's a fast, clean number, and it works beautifully in one specific situation.

Tessa: Which is?

Ray: Where the click and the purchase happen in the same session, the price is fixed, and the margin is known. Online retail. Somebody sees the ad, buys the forty-dollar thing, done. You know the revenue immediately, you know what it cost, you can act on it tomorrow.

Tessa: None of which describes a kitchen renovation.

Ray: None of it. Your version is: someone clicks in February, books a consult in March, gets a quote in April, argues with their spouse about it until June, and signs for a hundred and twenty thousand dollars. And the value of that job depends entirely on scope, so it isn't even a fixed number.

Why doesn't ROAS work for lead generation businesses?

Tessa: So what breaks, exactly? Because my ads guy still reports a return number and it looks great.

Ray: Three things break. First, timing — the revenue lands in a different month from the spend, so any monthly ROAS figure is comparing this month's cost to some earlier month's outcome. Second, scale — one job can be worth eighty thousand or two hundred thousand, so one signature swings your ratio wildly and it tells you nothing about the campaign. Third, and this is the big one, revenue isn't return. A hundred and twenty thousand in revenue at twenty per cent gross margin is twenty-four thousand of actual money. Those are very different businesses.

Tessa: Right, and my ads report shows the big number.

Ray: Every ads report shows the big number. It's not dishonest, it's just that the platform only knows what you told it a conversion is worth.

What should you measure instead of ROAS?

Tessa: So what should I be looking at instead?

Ray: For a business like yours, three things, in this order. Cost per qualified lead — not per lead, per lead your team agrees was real. Cost per booked job, which is the one that actually matters. And payback period.

What is payback period in marketing?

Tessa: That last one I don't know.

Ray: Payback period is how long it takes for a channel to return more money than you put into it. Not how much — how long. And it's the single most useful frame for comparing two channels that behave completely differently, because it stops you comparing a sprinter to a distance runner on the same stopwatch.

What is customer lifetime value and why does it change the answer?

Tessa: Where does lifetime value fit? People throw that phrase around and I've never applied it to us.

Ray: It probably changes your answer more than anything else we've said. Customer lifetime value is the total profit a customer generates over the whole relationship, not just the first transaction. And in a business like yours it's the difference between a rational decision and a wrong one.

Tessa: Because a renovation is a one-off.

Ray: Is it though? Somebody does their kitchen with you. Four years later, the basement. Six years after that, they move and do the new place. And in between, they're the person at the barbecue telling three neighbours who to call.

Tessa: Referrals are honestly most of our good work.

Ray: Right. So if a kitchen job is twenty thousand in gross profit, but on average that client eventually brings you another job and refers one that closes, the real value of winning that first client isn't twenty. It's closer to fifty or sixty.

Tessa: Which means I could pay a lot more to acquire one than I thought.

Ray: Considerably more. And this is the practical consequence: a campaign with a "worse" return on the first job can be your best investment, because you're not buying a job, you're buying a customer. Businesses that only optimize to the first transaction consistently underspend and lose to the ones who did this arithmetic.

SEO vs Google Ads: a twelve-month cost comparison

Tessa: Okay. Do the numbers for me. Because I'm going to have to take this to my partner and he wants to see it on paper.

Ray: Let's do it properly, over twelve months, with round numbers. Say your gross profit on an average job is twenty thousand dollars.

Tessa: That's close enough.

Ray: Paid ads first. You spend five thousand a month, every month. In month one, you get leads — that week. Let's say across the year it produces twelve booked jobs. So sixty thousand dollars of spend, twelve jobs, five thousand dollars per booked job. And it was profitable in month one.

Tessa: Which is why I like it.

Ray: Which is why everyone likes it. Now SEO. Say twenty-five hundred a month. Months one through four: nothing. Zero jobs. You've spent ten thousand dollars and you have nothing to show your partner.

Tessa: That's exactly the conversation we had in April.

Ray: Month five, first job. Months six through nine, maybe one a month. By month eleven and twelve you're getting two or three a month, because the pages that went up in month two are now ranking. Total for the year — also about twelve jobs.

Tessa: For thirty thousand instead of sixty.

Ray: For thirty thousand instead of sixty. Half the cost per booked job. But look at where those jobs sat on the calendar — almost all of them in the back half. If you had judged that channel in April, you would have killed it two months before it started working.

Tessa: And we nearly did.

Ray: Most people do. And here's the part that makes it worse: the twelve jobs SEO produced this year didn't cost thirty thousand. They cost thirty thousand and twelve months. The time is part of the price, and it's the part that never appears on an invoice.

Should you move budget from paid ads to SEO?

Tessa: So then the obvious next thought — and I've genuinely had it — is: SEO is half the cost, so move all the money there.

Ray: Yeah, don't. And I say that as the paid media guy, which should tell you something.

Tessa: Why not?

Ray: Because the cheap number you're looking at is the month-twelve number, and it only exists because months one through eleven happened. If you cut ads today to fund more SEO, you get roughly six months with no lead flow while the slow channel ramps. Most businesses can't absorb that, and the ones that try tend to panic in month three and turn the ads back on — so they've paid for the gap and got none of the benefit.

Tessa: So it's a sequencing thing, not a switch.

Ray: Sequencing, and the honest version is: paid buys you time, SEO buys you the future, and you generally need both running while the slow one matures. As organic starts carrying real volume, you can pull paid back — but you pull it back on evidence, not on hope.

When should you judge a paid campaign vs an SEO campaign?

Tessa: Alright, then give me the dates. When do I judge each one?

Ray: Paid search, with enough volume: four to six weeks for a first read, three months for a real verdict. Anything shorter and you're reacting to noise — a slow fortnight is a slow fortnight, not a broken campaign.

Tessa: And SEO?

Ray: Six to nine months for a verdict, in a market like Calgary renovation where you're up against established firms. Three months is far too early. Twelve months is later than you need to wait.

How do you account for seasonality in marketing reports?

Tessa: How do I separate any of this from the season? Because our year is not flat. January to March everybody's planning, then it's build season, then it goes quiet.

Ray: Then you have to compare against the same month last year, not against last month. That's the single most common misreading I see, and in a seasonal Canadian business it's almost guaranteed to mislead you.

Tessa: Because August is always down.

Ray: August is always down, and if you compare August to June you'll conclude your marketing collapsed. Compare August to last August and you find out whether anything actually changed. Same spend, same month, previous year — that's the only fair comparison.

Tessa: What if we don't have last year's numbers?

Ray: Then this year is the year you start building them, and you'll be in a much better position twelve months from now. In the meantime, the workaround is to judge on a rolling basis — look at the last ninety days against the ninety days before it, and accept that the read is rougher.

Tessa: And the planning season thing — does that change when we should be spending?

Ray: It changes when you should have published, which is the more useful way to think about it. If people are researching in January, the pages need to be ranking in January, which means they went up in the summer. Seasonal SEO runs about two seasons ahead. Paid you can switch on the week you need it — that's genuinely what it's for.

Tessa: And between now and month nine I just — trust you.

Ray: No, and that's important. "Just trust us" is what people say when they don't have interim signals. There are things that should be moving well before jobs show up, and if they're flat at month four, that's a real conversation. But those are diagnostic signals — they tell you the machine is running. They're not the financial verdict. Don't confuse the two, and don't let anyone show you a rising graph of something and call it a return.

Tessa: That's a distinction I've never had drawn for me.

Ray: It's the one that matters most in this whole conversation. Leading signals tell you it's working. The financial number tells you what it was worth. They arrive months apart, and you need both.

What happens if you stop paying for SEO?

Tessa: Okay, here's one my partner will absolutely ask. If we stopped paying for SEO tomorrow, what happens?

Ray: It decays. Slowly. That's the other half of why this is confusing.

Tessa: Slowly meaning?

Ray: Meaning you'd probably see very little for two or three months. Rankings hold for a while, the pages are still there, the traffic keeps coming. And then competitors publish, your content ages, the questions people ask change, and six or nine months later you're noticeably worse off with no obvious event to point at.

Tessa: So people cancel, see nothing happen, and conclude it was never doing anything.

Ray: That's exactly the trap, and it's the mirror image of the start-up lag. Slow to arrive, slow to leave. The same property that makes it frustrating in month three is what makes it valuable in year three — you own it. Paid is rented. The day you stop paying for ads, the leads stop that afternoon.

Tessa: Owned versus rented. That I can explain to him.

Ray: That's the whole thing in two words. And there's a compounding piece to it. Every page you publish this year makes next year's pages rank faster, because the site is more established. Paid doesn't compound — month thirteen costs the same as month one, and often more, because auction prices tend to rise.

Do SEO and paid search help each other?

Tessa: Do the two actually help each other, or is that just something agencies say?

Ray: No, it's real, and it goes both directions. Your ad account knows the exact phrases people typed before they converted — that's the best content brief you'll ever get, and it's free. We use it constantly.

Tessa: So the ads tell you what to write.

Ray: The ads tell you what to write, and they tell you in weeks instead of guessing for a year. Going the other way — as your brand and your content get stronger, more people search your company name directly, and branded searches are dramatically cheaper to bid on than generic ones. So a strong organic presence quietly pulls your average cost per click down.

Tessa: Huh.

Ray: And there's a third thing now, which is that both of them feed whether an AI assistant mentions you when someone asks it who does renovations in Calgary. That's a longer conversation, but it runs on the same foundation — content that exists, on a site that's established.

Tessa: So the myth I've been carrying is that there's one scoreboard.

Ray: That's the myth. One scoreboard, one number, one verdict for "my marketing." And the reality is you're running two engines with different fuel and different top speeds. Judging your SEO on ROAS is like judging a mortgage payment on this month's cash flow — technically you can compute it, and the answer will tell you to sell the house.

Tessa: I'm using that.

Ray: It's yours.

How to set a verdict date for each channel

Tessa: Alright. What do I do this week?

Ray: One thing, and it's a piece of paper, not a tool. Write down, for each channel, three things: the metric you'll judge it on, the date you'll judge it, and what a bad result actually looks like. Paid — cost per booked job, verdict at ninety days, bad means above whatever number breaks your margin. SEO — organic booked jobs, verdict at nine months, bad means still zero at month six.

Tessa: And then?

Ray: And then don't move the dates. Because what actually kills marketing budgets in this country isn't bad campaigns — it's good campaigns getting judged three months early by somebody having a nervous week. If the date's written down in advance, that decision gets made on evidence instead of mood.

Tessa: That's a very unromantic piece of advice.

Ray: It's the most valuable one I've got. Tessa, thank you — this was great.

Tessa: Thanks, Ray.

Ray: And that's episode six of The Visibility Room. Thanks for spending some time with us — we'll see you next time.