SEO Agency Not Delivering? A Shopify Founder's Diagnostic

Quick answer: To tell whether your SEO agency is too early or doing the wrong work, run four checks in order: do you control the accounts, can the agency show a dated list of what it changed, are those pages indexed and showing for the agreed queries, and does the timing match the scope you signed. Google says some changes take effect in hours and others take several months, so a flat month 3 can be fair. A month 3 with nothing changed, nothing indexed and no target queries is not.

You are a few months into an SEO retainer and the report arrives. There are charts, a list of tasks and a paragraph saying things are “progressing well”. Your orders from Google look much the same as before you started.

That does not mean the agency is failing, and it does not mean it is succeeding. It means you cannot tell, and that is the part worth fixing. This guide is a mid-engagement diagnostic for separating “too early” from “the wrong work” with tools you already have. It is not a case against agencies: many of the checks protect a good agency from an impatient founder as much as they protect you.

Choosing an agency before you sign is a different question, covered in how to choose a Shopify SEO agency in Australia. Here you already have one and need to know whether to wait, renegotiate or leave.

I will not give you a pass mark such as “X% more traffic by month 3”. Google does not give one, and no single number holds across stores. You can check whether the work happened, whether Google has seen it, and whether the early signals that come before revenue are there.

Before you start: decide what “delivering” means

A lot of confusion about SEO results comes down to scope. Find three things before you open any report.

1. The scope you agreed. The proposal or contract that says what would be done and by when. If it says “ongoing SEO optimisation” and nothing else, that is your first finding.
2. A list of target pages and target queries. The searches the work is meant to win and the pages meant to win them. If nobody wrote these down, nobody can say whether they were won.
3. The reporting you were promised. How often, and what it covers. For reference, our own SEO scope includes a progress report every two weeks and a monthly stakeholder report with every KPI, the data and next month’s plan, tied to revenue and conversion outcomes, not impressions. Yours may differ. What matters is that you know which one you bought.

If the target list or timeframe was never given, ask for it now. Nothing below works without it.

Step 1: check you control the accounts

Before judging results, check you can see the same data the agency sees, with your own login.

Google Search Console. Open Settings, then Users and permissions. Google describes an owner as someone who can add and remove users, configure settings and use all tools, and says only owners can open that page. If the only owner is an agency email address, you depend on them to see your own search data. Fix that now, politely.

Google Analytics 4. In Admin, open Account access management and Property access management. Google’s roles run from Viewer up to Administrator, who has full control and can add and delete users. An Editor has full control of property settings but cannot manage users. You want your own login as an Administrator, with the agency as an Editor or Analyst if they need access.

If you cannot get into your own Search Console or GA4 at all, stop here. You cannot run the other checks, and “we manage that for you” is not a reason to hold the keys.

Step 2: ask what was actually changed

The most useful document in a mid-engagement review is a dated change log: what was changed, on which URL, on what date, and why. It can be a spreadsheet, and an agency should be able to produce one quickly.

“Improved on-page SEO” is not an entry. “Rewrote the title and intro of /collections/vitamin-c-serum for ‘vitamin c serum australia’” is.

Then check about ten entries on the live store:

Two Shopify details help here. Shopify’s help centre says a page title can be up to 70 characters but suggests keeping it to 60 so it is not shortened in results, and recommends 160 characters for a description. And Shopify says redirects start working immediately, but only from an address that no longer loads a valid page. If an old address from the log shows “not found” instead of landing on the new page, no redirect is working, and shoppers and Google both hit a dead end. Test one.

Also check the gap between “recommended” and “implemented”. An audit that lists forty fixes is a deliverable, but it is not forty changes. Was the agency to make the changes on your store, or to hand a list to you or your developer? A list waiting for someone to act is a process problem, not an SEO result.

If the log is empty, vague or “coming soon” three months in, you have your answer for this step. If it is detailed and the changes are visible, the question becomes whether it is the right work.

Step 3: check indexing and the target queries

Work that Google has not seen cannot change your search results, so check that it has, and that the pages show for the agreed searches.

Are the changed pages indexed?

In Search Console, paste a changed URL into the search bar at the top. Google says the results come from the most recently indexed version of the page, not the live page. Read two things: whether the page is on Google, and when it was last crawled. If the last crawl is before the change date in the log, Google is still looking at the old version. Google also says a “URL is on Google” result means the page is eligible to appear, not guaranteed to. “Request indexing” has a daily limit and does not guarantee the page will be indexed, so one request after a real change is reasonable and repeated requests are not a strategy.

Then open Indexing, then Pages. Google says not to expect every URL on a site to be indexed, and that non-indexed URLs can be fine. So the test is not “is the not-indexed list zero”. It is “are the pages in the change log, the ones that matter for revenue, indexed”. If many sit under “Crawled - currently not indexed” or “Discovered - currently not indexed”, read our guides to Crawled - currently not indexed and Discovered - currently not indexed before concluding anything. Google defines the first as crawled but not indexed, and the second as found but not yet crawled. Neither status on its own says the agency got something wrong.

Also rule out a self-inflicted block. Google’s documentation says a noindex rule drops a page from search results once Googlebot reads it, that Googlebot has to recrawl a page to see a newly added rule, and that the page must not be blocked in robots.txt or Google never sees the rule. A theme change or an app can add one by accident, so ask whoever manages the theme to check. The live test in URL Inspection shows whether Google can access the page for indexing, but Google says it is not a guarantee the page will be included.

Are the target queries showing up?

Open Performance, choose the last three months and use the Compare tab against the previous period, the same approach Google recommends when it explains how to investigate traffic changes. Filter by a target page and look at the queries tab. Two cautions:

• Brand queries flatter the picture. Clicks from people typing your store name are not SEO growth. Search Console has a branded and non-branded queries filter. Google says it is not available for sites with low impressions and that some queries may be misclassified, so treat it as a guide.
• Average position is a secondary measure. Google’s guide to diagnosing traffic changes says impressions and clicks are ultimately the measure of success, and that small fluctuations in position happen at any time.

Google’s traffic-drop guide uses a “Last 16 months” date filter, so export your pre-engagement numbers now: a year-on-year view is the best defence against reading a seasonal dip as the agency’s doing.

Read the combination, not any one number:

Step 4: check the timing against the agreed scope

This is where “too early” and “wrong work” separate. You compare two timelines: what the scope promised by when, and what happened. First, the general picture. Google’s SEO Starter Guide says every change takes some time to be reflected, that some changes might take effect in a few hours and others could take several months, and that you likely want to wait a few weeks to assess whether your work had beneficial effects. Shopify’s own guide to how long SEO takes says results typically take three to six months, and that keywords dominated by established, high-authority sites can take more than 12 months to challenge. Shopify’s month-by-month outline is a general guide from its blog, not Google guidance and not a promise for any one store, but it is a useful yardstick for what to ask at each point. How long Shopify SEO takes goes into it fully.

Shopify also says missing results at month six deserves investigation: technical blockers, tracking, keyword strategy, content quality and search intent. Put that checkpoint in writing, as a review date rather than a verdict.

Two things can muddy the timing:

• Google’s own updates. Google makes significant, broad changes several times a year and recommends waiting at least a full week after a core update completes before analysing your site in Search Console. Check the Search Status Dashboard before reading a dip or a jump as the agency’s doing.
• Reporting glitches. Google’s Data anomalies page lists known problems with Search Console data. As at 30 October 2026 it lists a logging error that stopped impressions being reported accurately from 13 May 2025 to 27 April 2026, marked resolved, with clicks unaffected. The page says only impressions and the related CTR and average position were affected, so if your comparison (a year-on-year view, for example) crosses that window, lean on clicks and check the page for yourself.

Now a worked example. These numbers are illustrative, not from a real store. Say the scope promised an audit by day 30, 20 priority collection and product pages rewritten by day 60, and four articles by day 90. You are at day 90.

Store A. The audit arrived on day 38. Nine of the 20 pages are rewritten (45% of the agreed page work) and one article is live. Delivery is behind, and whether SEO “works” is not yet the question. The conversation is about the delay and a catch-up plan.

Store B. All 20 pages were rewritten by day 60, all are indexed with a crawl after the change, and 14 of the 20 show impressions for their target queries, with clicks small but moving. The work is done, Google has seen it and early signals are present. A flat revenue line at day 90 is not evidence of failure here.

Store A and Store B can show identical flat traffic and need opposite responses.

What you should see in Search Console and GA4 by month 3

By month 3 I would not expect a revenue story. I would expect the evidence that comes before it. This is my rule of thumb, not Google’s, and none of it is a guarantee. These are signals to ask about.

Search Console, signs the work is landing: changed pages indexed with a crawl after the change; impressions on the targeted pages and queries up against the previous period; more distinct queries showing for each worked page, some without your brand name.

Search Console, warning signs: pages in the change log not indexed or not crawled since the change; flat or falling impressions on every worked page, or all growth on brand queries; a report that quotes sitewide totals and average position and never names a page or a query.

GA4, signs the work is landing:

• Organic Search sessions arriving on the pages in the change log. Google defines the Organic Search channel as traffic whose source matches its list of search sites, or whose medium is exactly organic, and says it includes visits from AI Overviews and AI Mode.
• In Reports, Engagement, Landing page, add the secondary dimension Session source / medium (Google’s help page names this example) to see which landing pages each source reaches. The key events and total revenue columns show what happened next, once key events are marked.

GA4, warning signs:

• Organic Search sessions rising while key events and purchases on those landing pages stay flat. That points at the store, not the traffic.
• Organic numbers that do not reconcile with Shopify’s orders. Check tracking before judging the channel, the problem covered in GA4 versus Shopify Analytics.

Search Console clicks and GA4 sessions are different measurements, so do not expect them to match exactly. If a report shows numbers you cannot find in either tool, ask where they came from. For metrics worth tracking beyond SEO, see three numbers Shopify founders watch.

Reading the result: four outcomes

An access problem overrides all four: if you cannot see your own Search Console or GA4, fix that first.

Notice what is not on the list: “the agency is bad” or “SEO does not work”. These checks cannot prove either. They tell you which conversation to have.

Not sure which of the four you are in? Tell us about your store.

A fair test to run with your agency

If you are still unsure, run a short written test that either side can pass or fail.

1. Agree ten pages and ten queries. Pages that matter for revenue and the searches they should show for.
2. Ask for the dated change log for those ten pages. Done, in progress, not started.
3. Set a review 30 days out. Check the pages are indexed with a crawl after the change, and compare impressions for the ten queries against the previous period.
4. Agree in advance what each result means. For example: work done and indexed but impressions flat, hold one more cycle; work not done, a dated catch-up plan; done and moving, continue.
5. Put the outcome in an email both sides agree to.

Agreeing the meaning first means nobody can move the goalposts afterwards. Good questions for that conversation come from Google’s own list for choosing an SEO: What kind of results do you expect, and in what timeframe? How do you measure your success? Will you share all the changes you make to my site, and explain your reasoning?

If you decide to leave, leave cleanly

Google says you are responsible for the actions of any company you hire, which is the best reason to hold your own access before a dispute, not after one. Be an owner in Search Console and an Administrator in GA4, and remove any ownership verification tokens you no longer want valid, because Google notes a removed verified owner may be able to regain access. Ask for the handover in writing: the change log, the target page and query list, redirects, content drafts and the tracking set-up.

To compare what you were sold with what a Shopify SEO engagement normally contains, see what Shopify SEO services include in Australia. For the wider map of search and conversion, start with the Shopify SEO and CRO resource hub.

Key takeaways

• A flat month 3 can be fair or unfair. Google says changes take anywhere from hours to several months, so the answer depends on what was done and when.
• You should be an owner in Search Console and an Administrator in GA4. If not, fix that first.
• Ask for a dated change log and check ten entries on the live store. “Recommended” is not “implemented”.
• Judge the pages in the change log, not the whole index. Google says not to expect every URL to be indexed.
• By month 3, look for early signals, not revenue: indexed pages, rising impressions on targeted queries and organic sessions landing on the right pages.
• If organic traffic rises and orders do not, the leak is after the click and more SEO will not fix it.
• Agree in writing what each result will mean before the next review, so the test is fair to both sides.

FAQs

How long should I give an SEO agency before judging the results?

Google says some changes take effect in a few hours and others could take several months. Shopify’s guide says results typically take three to six months. Rather than a fixed number of months, judge against the scope you agreed: was the work done by the agreed dates, and has Google seen it?

How do I tell if it is too early or the wrong work?

Compare the change log with the scope and check the live store. If the agreed work is done, indexed and early signals are appearing, it is likely too early. If work is missing, not indexed, or aimed at pages and queries buyers do not search, it is delivery or targeting, and waiting will not fix it.

What should I see in Search Console by month 3?

Changed pages that are indexed with a crawl after the change, impressions on target pages and queries rising against the previous period, and more distinct queries per page, including some without your brand name. These are signals to ask about, not guarantees.

Should I have my own access to Search Console and GA4?

Yes. In Search Console, an owner can add and remove users, configure settings and use all tools. In GA4, an Administrator has full control and can manage users.

Can an SEO agency guarantee rankings?

No. Google’s guidance on hiring an SEO says no one can guarantee a number one ranking on Google. A fair agency describes the work, how it will be measured and the timeframe it expects.

My organic traffic is up but sales are flat. Is the agency failing?

Not necessarily. Rising organic sessions with flat orders usually points to what happens after the click: the product page, the offer, the checkout, or tracking that is not counting orders correctly. Check GA4 landing pages and key events and reconcile with Shopify. Our guide on which to fix first, SEO or CRO explains how to separate the two.

Want a second opinion before you decide anything?

If you want a second opinion before you decide anything, tell us about your store or book a free 30-minute diagnostic. We look at your store before we recommend anything, and we tell you honestly if SEO isn’t the right move. No pitch. If your organic traffic is arriving but orders are not, the question is what your store does after the click, and our Shopify SEO service and Shopify CRO service cover both sides.