SEO strategy ROI after 90 days: a teardown

A 90-day case study teardown of SEO strategy ROI—define ROI terms, build a defensible baseline and attribution model, break down costs vs. benefits, and isolate what actually moved (content, tech, and authority) so you know when 90 days is “enough.”

Sev LeoLinkedInTwitter

Sev Leo is an SEO expert and IT graduate from Lapland University, specializing in technical SEO, search systems, and performance-driven web architecture.

90-day ROI question

ROI in 90 days is a blunt instrument. It’s still useful, because it forces clear math and faster decisions. Think “Did this earn or de-risk enough?” not “Did we win SEO forever?”

Define ROI terms

ROI is benefits minus costs, divided by costs. In SEO, costs include cash spend, labor hours, tools, and the work you didn’t do instead. Your “benefit” depends on attribution rules, like “last-click revenue” versus “pipeline influenced.”

A clean baseline helps: $3k writer + 40 hours at $75/hour + $200 tools. Then decide if revenue gets full credit, partial credit, or none without closed-won proof. Your ROI number is only as honest as that rule.

What 90 days captures

In 90 days, you usually see movement signals before money signals.

When it’s worth it

“Worth it” means the expected payback fits your business clock. You need a window, a target acquisition cost, and a reason to keep investing. Otherwise, you’ll chase vanity lifts like “+30% impressions” that never pay rent.

Set criteria like: break even in 6–12 months for content-led SaaS, 3–6 months for local services, or immediate for ecommerce promos. Add strategic value when it applies, like reducing paid spend volatility or owning “best X” queries competitors can’t outbid forever. If you can’t write down the break-even point, you’re not measuring ROI, you’re hoping.

Baseline and setup

False wins happen when your baseline is fuzzy and your tracking is “close enough.” We froze a snapshot, defined conversions, then priced the work like a finance team would.

Starting metrics snapshot

We captured the before-state in one place, on one date, so later lifts had a real reference point.

If you can’t write this table, you’re not measuring ROI yet. If you need a checklist to get the fundamentals right, start with this SEO guide for getting started.

Tracking and attribution

We treated tracking like product infrastructure, not a marketing afterthought.

1. Configure GA4 events for forms, calls, and key clicks.
2. Map each event to a CRM stage and a single primary source.
3. Enforce UTM rules for every campaign link, no exceptions.
4. Add call tracking with dynamic numbers and page-level attribution.
5. Lock conversion definitions in writing, then version changes.

Once attribution stops drifting, optimization becomes math instead of debate.

Cost accounting method

SEO ROI gets distorted when you count only tools and forget labor. We priced hours at fully loaded rates, and we counted work that directly changed outcomes.

We included strategy, content production, edits, technical fixes, and reporting tied to decisions. We excluded general meetings, internal training, and “nice-to-have” experiments. Fixed costs were tools and retainers, while variable costs were hours and one-off production.

If you don’t separate fixed from variable, you can’t forecast the next 90 days with confidence.

Strategy chosen

A 90-day window forces you to pick levers that move before Google’s slower compounding effects kick in. We chose an “optimize what already exists” strategy over a brand-new content engine, because existing pages already had crawl history, links, and some rankings to improve.

Opportunity selection

We picked bets that can show movement without waiting for a full authority cycle.

• Update quick-win pages ranking positions 4–15
• Target bottom-funnel queries with clear commercial intent
• Add internal links from high-authority pages
• Fix technical blockers harming crawl and indexing
• Refresh decayed content with new evidence

If you can’t tie a task to an existing page and query, skip it.

Execution timeline

Twelve weeks sounds long until you lose two weeks to approvals and dev queues.

1. Weeks 1–2: Audit rankings, pages, crawl, and conversion paths.
2. Weeks 3–4: Ship technical fixes and internal linking upgrades, then use AI tools to boost organic traffic to speed up content refreshes and prioritization.
3. Weeks 5–6: Refresh priority pages and publish missing bottom-funnel pages.
4. Weeks 7–9: Add supporting links, expand snippets, and test titles.
5. Weeks 10–12: Iterate on winners and report against revenue proxies.

Your checkpoint isn’t traffic. It’s “did the pages that sell move up?”

Tradeoffs made

We did not do a big rebrand, a full site rewrite, or a 50-post content sprint. Those playbooks can work, but they usually pay out after the 90-day mark and add coordination risk.

Instead, we accepted a narrower upside in exchange for faster proof, using KPIs like keyword lifts on money pages, assisted conversions, and lead quality notes like “came in asking for pricing.” The goal was a clean ROI signal, not a vanity growth chart.

Costs breakdown

Your ROI math breaks when you hide costs in the gaps between teams. Put every hour and subscription on the table, including the “quick dev tweak” that wasn’t quick.

Here’s a realistic 90-day cost sheet for one mid-sized SEO push.

The big shock is rarely tools. It’s coordination time disguised as “just a few minutes.”

Benefits after 90 days

You have enough time at 90 days to see momentum, not dominance. Treat most wins as correlated, then isolate what you likely caused.

If revenue jumps without query and page-level lift, assume seasonality or promos first.

Teardown: what moved

Ninety days is enough time to see signal, but only if you tie each action to a measurable outcome. Track cost in hours, then map it to first impact in Search Console and downstream conversions.

Fastest wins

The quickest lifts came from changes that helped Google understand pages faster and helped users click sooner. Think titles, internal links, indexation cleanup, and shaving obvious speed bottlenecks.

Title rewrites moved fastest because they affect the snippet immediately and change CTR without waiting on new rankings. Internal links followed, because they reshaped crawl priority and passed relevance to pages that were already close. Indexation fixes and speed improvements were mixed, but they worked when they removed a hard blocker like “Discovered, currently not indexed.”

If your first 30 days show nothing, you likely improved quality without improving discovery.

Content performance split

Content gains were uneven because intent-match beats volume every time. Separate winners from noise before you write another word.

• Updated pages: clear lifts on pages ranking positions 6–20
• New pages: slow starts unless internal links landed day one
• Cannibalization: two pages traded ranks until one consolidated
• Topics that didn’t rank: broad terms without “how-to” specificity
• Losers: refreshes that changed text, not usefulness

Your content isn’t “good” or “bad.” It’s matched, or it’s not.

Technical fixes impact

Technical work paid off when it changed crawling or canonical signals, not when it chased green scores. Fixes that removed ambiguity produced movement; hygiene mostly reduced risk.

Crawl path work mattered when it reduced depth and removed dead ends, especially for pages stuck behind faceted navigation. Redirect cleanups mattered when they removed chains and fixed wrong canonicals, because equity started landing on the intended URL. Schema helped when it reinforced clear entities and eligibility, but it rarely moved rankings alone.

Do the hygiene, but price it like insurance, not growth.

Links and authority

Links were the slowest lever, but they compounded when paired with pages already showing demand. You learned more from a few controlled pushes than from “more outreach.”

1. Reclaimed unlinked mentions, low effort, small stability lift on branded queries.
2. Built partner links, medium effort, modest improvements to mid-tail rankings.
3. Earned one strong editorial link, high effort, visible bump on a single cluster.
4. Added internal authority links, low effort, faster gains than most externals.
5. Tried broad outreach blasts, high effort, mostly noise and weak domains.

If links don’t change conversions, you’re renting rankings instead of building a funnel.

Use the 90-day teardown to decide your next move

1. Lock your definition of ROI and keep it consistent (revenue, pipeline value, or cost savings) before you compare periods.
2. Attribute the lift to the real driver: new/updated pages, technical fixes, or authority gains—then double down on the lever that produced measurable movement.
3. If results are mixed, separate “leading indicators” (rankings, impressions, indexing, CTR) from “lagging outcomes” (revenue/pipeline) and extend the window only for the parts that inherently take longer.
4. Turn the findings into a 30–60 day plan: scale the fastest-win pattern, cut low-leverage tasks, and adjust tracking so the next teardown is even cleaner.