Why do Shopify SEO projects fail?

In planning and design theory, there is a concept called a "wicked problem".

It describes a problem that is difficult or impossible to solve because of incomplete, contradictory, and changing requirements that are often difficult to recognise.

There is no single "right" answer, only better or worse ones, and the finish line moves every time you think you have crossed it. It's been a subject that's fascinated me for a long time and one I discussed recently on the Retail is Detail Podcast. 

SEO  is a classic wicked problem. The rules and the algorithm change constantly, as does the competition.

eCommerce is also a wicked problem. It involves complex logistics, shifting consumer behaviour, and intricate technical infrastructure.

When you attempt to execute SEO for eCommerce, you are colliding two wicked problems together. When you do this, the outcome is exponentially more complex. 

When we started out as an agency, we didn't look at it this way. We were pretty much like every other agency out there. We did SEO for any kind of business that would work with us. If you had a website and a budget, we would take the project on. Results, as you can probably imagine, were inconsistent.

This was hugely frustrating, both for us and our clients. Why did our tactics work amazing well some of the time, but not others? 

In order to try and solve this, we started to narrow our focus. First, we specialised in eCommerce. Then, we niched down further to Shopify. Finally, we found our true sweet spot: Shopify stores with large catalogues. 

This was a light bulb moment. Finally, if we found something that worked, we could do the same thing across projects and the outcomes would be consistent.

However, this evolution wasn't just about marketing positioning - it was a response to failure.

We have been doing this for a long time and, as I'm sure you can imagine, we have made plenty of mistakes along the way. It has taken us years to get to this point, but through that process - shifting from generalist to hyper specialist - we learned exactly where the cracks appear in a standard SEO campaign.

Here are the eight main reasons why Shopify SEO projects fail, based on many years of learning the hard way.

1. Using the wrong tactics

This might seem obvious, but it is incredibly fundamental. Everyone knows that specialisation is important, but the trap we see so many people fall into is that they don’t actually know what they should be looking for.

When brands look for an SEO partner, the natural instinct is to focus on the market vertical. You might search for a "fashion" agency or a "homeware" agency. On the surface, that makes sense; you want someone who understands your product.

But in reality, "fashion" is way too broad a term to be useful. We have found that the specific products you sell matter far less than how your business is actually built. This is what we mean when we say you need to look at business type, not just vertical.

To see why, let’s look at two businesses. Both are on Shopify, and both turn over £5 million a year:

• The Subscription Brand: They sell deodorant. They have five or six SKUs and a handful of variants.

• The Retailer: They are a fashion boutique with 10,000+ products across hundreds of categories.

On the surface, they are both "eCommerce," but the challenges they face are worlds apart. You couldn't drop a Head of eCommerce from one into the other and expect them to be successful without a total shift in mindset. Their margins, their primary channels, and their tech stacks are completely different.

Feature	Subscription Brand (Low-SKU)	Large Retailer (High-SKU)
Primary Growth Channel	Meta Ads & Social	Google Ads & SEO
Typical COGS	~30%	60–70%+
Tech Stack	Skio, Klaviyo, Rebuy	Algolia, Matrixify, Searchspring
SEO Focus	Brand storytelling & blogs	Technical merchandising & data

The failure happens when you apply the playbook for one business type to the other.

We still see agencies applying the same standard SEO tactics to every client: a few blog posts and some meta title tweaks. 

But if you have 10,000+ SKUs, your customers aren’t reading blog posts. They are filtering, scanning, and comparing attributes. If you hire an agency that specialises in "Fashion" but their experience is limited to low-SKU lifestyle brands, they will try to solve your complex data problem with content marketing.

Getting the right partner - and building the right strategy - comes down to asking questions about platform, catalogue, buying journey, and goals, rather than industry buzzwords. We actually put together a prompt to help you articulate this, which you can copy and adapt here.

2. Unstable merchandising

This is a failure point that is strictly operational, but it often trips up brands in the sub-£5M revenue bracket, and particularly those in the sub-£2M space.

The issue is stability.

It might seem obvious, but it is incredibly difficult to drive organic performance if you are consistently out of stock of your core products. SEO is a momentum channel; it relies on consistent signals. If your best-selling product page is constantly toggling between "In Stock" and "404 Not Found", it becomes very hard to maintain rankings.

But there is a more subtle version of this problem: the "Reinvention Reflex".

Brands in this growth phase are often still refining their positioning. When sales are flat, it is natural to want to overhaul the catalogue - renaming collections, moving products around, or completely restructuring the navigation to find a better fit.

While this often makes sense for merchandising, it can cause friction for SEO.

A lot of non-brand discovery happens on specific collection and product pages that have accumulated authority over time. If you shift your catalogue too aggressively, or delete "seasonal" collections that were actually driving traffic, you risk erasing that progress.

SEO demands a level of permanence that often conflicts with the agility of a startup. If you don't balance the need for merchandising freshness with the need for URL stability, you can end up resetting the clock over and over again.

3. The strategic void

Once you have the right business strategy and a stable catalogue, you need the right leadership to manage it. The third reason these projects fail is not technical - it is organisational.

Over the past decade, marketing has exploded in complexity. Just take a look at the number of available tools - there's data platforms, product feed management apps, CRO tools, the list goes on...

Understandably fractured their marketing efforts. They hire specialists: one agency for email, one for SEO, and one for PPC. This makes sense for execution, but it creates a strategic void at the top.

It isn't that brands explicitly expect their agencies to act as the CMO. It is that without a strong central leader, agencies end up stepping into that void by default. You end up with a room full of conflicting voices, where every channel partner is advocating for their own budget and their own attribution.

SEO is particularly difficult to manage in this environment because it has such an unusual position.

Unlike other channels, SEO involves getting deep into the website infrastructure. It doesn't just live in the same ecosystem as Paid Search, CRO, and UX - it directly influences them.

SEO requires changing site architecture, which impacts UX. It requires changing content, which impacts Paid Search landing pages. It requires technical changes, which impacts site speed and conversion rates.

Reporting on this interplay is incredibly complicated.

If you don't have someone internal who can orchestrate all of this - someone who understands that an SEO fix might lower Paid CPAs but confuse a CRO test - then the channel fails. Your channel partner is not your CMO. We can advocate for the channel, but we cannot be the conductor. You need someone internal to manage the chaos.

4. The tracking black hole

If the strategy is right, and the leadership is there, the next failure point is valuation.

The problem with tracking isn't just about optimisation - it is about value. So many brands simply do not understand the true value of SEO as a channel because GA4 tells them it isn't working.

Here is the reality: for a large catalogue Shopify store, organic revenue should account for around 20% to 25% of your total revenue. When we fix tracking, that is consistently what we see.

But most stores have broken tracking. Previously, 8 out of 10 stores we worked with had tracking problems; now it is 9 out of 10. This is usually caused by a mixture of poor implementation and bad consent management, often exacerbated by tools like cookie banners incorrectly labelling sessions as direct traffic. In those cases, organic revenue often sits at around 8%.

This creates a massive perception problem. If you look at your dashboard and see a channel driving only 8% of revenue, why would you pay attention to it? You wouldn't. You would invest that budget elsewhere.

This leads to a vicious cycle where the channel is undervalued and underfunded, simply because the data is wrong. Fixing this isn't just a technical exercise - it is the only way to see what is actually happening in your business.

5. The failure to educate 

The tracking issues we discussed above create a dangerous secondary problem. It isn't just that the data is wrong; it is that the data creates a trust deficit.

We have lost projects in the past not because the strategy wasn't working, but because we lost the battle for buy-in. This happens when you fail to educate stakeholders about why the numbers in the "black hole" look the way they do.

A common scenario we see is that Google Search Console results look great - impressions and clicks are flying up - but in GA4, organic revenue is flat or even down.

When you dig into this, usually one of two things is happening:

• "Direct" or "Unassigned" traffic is massively up due to the misconfigured tracking or strict consent management we mentioned in the previous point.

• Paid Search is massively up while overall revenue is flat, suggesting a cannibalisation or incrementality issue.

The problem here though isn't just the data - it is the timing.

If you spot this early and educate the client about these moving parts, it builds trust. It is seen as education.

If you catch this late - explaining it only after the client asks why revenue is down - it sounds like you are making excuses.

Because the numbers in GA4 are often stacked against you, you have to prioritise education. All of this stuff is really complicated, and there are loads of moving parts. If you don't do the hard work of educating stakeholders upfront, you will lose their trust before the results even have a chance to show.

6. Lack of trust

Beyond the data, there is a fundamental reason why trust in the SEO industry is at an all-time low. It comes back to the concept of the "wicked problem" we mentioned at the start.

I suspect anyone reading this inherently understands what a wicked problem feels like, even if you have never heard the term before. It is that nagging sense that the goalposts are always moving and that nothing is ever truly "fixed".

This is why so many relationships between brands and agencies fail.

When you are living inside a wicked problem every day - dealing with logistics, algorithms, and technical debt - and an agency pitches you a simple, definite tactic like "we just need to build more links" or "we need to create more category pages", it jars.

It creates an immediate lack of trust.

Deep down, you know that a definite, linear solution cannot possibly solve a problem this complicated. When the proposed solution is simple, but the problem feels chaotic, the disconnect is palpable. Most people's experience of SEO is defined by partners who go broad, but never deep enough to solve the specific friction points of a complex business.

7. Lack of platform-specific knowledge

This brings us back to the metric we introduced at the start: repeatability.

As we moved from "eCommerce generalists" to "Shopify specialists," we realised that while high-level strategy is universal, execution is specific.

In the past, the industry treated SEO as platform agnostic. The logic was that a canonical tag is a canonical tag, regardless of the CMS. And at the strategy layer (the top of the diagram below), this is true. Whether you are on Magento or Shopify, the principles of taxonomy and user intent are identical.

But as you move down the chart, you hit a fork in the road: execution.

Solving a technical issue on Magento is a completely different engineering challenge to solving it on Shopify. This creates the "generalist trap" you can see in the centre. 

Agencies that try to remain platform-agnostic fall into this void. Because they are constantly relearning the quirks of different platforms on your dime, their ability to deliver consistent results plummets. They get stuck in the middle, where repeatability is low (~30%).

To achieve high repeatability (~80%), you have to pick a lane.

Shopify is fantastic, but it has specific rigidities that break down as your catalogue grows. Without custom intervention, you face specific failures like having no native subcategories, filters that are not crawlable, and shallow internal linking structures.

By bypassing the generalist trap and committing to Shopify, we turn these variables into constants. We know exactly which lever to pull to fix a taxonomy issue because we have solved that specific problem a hundred times before.

8. Failure to execute

The final failure point is simply getting the work done.

One of the biggest mistakes we made early on is falling into a trap that we see a lot of agencies fall into. We used to send over long lists of "stuff to do" - usually in the form of a PDF audit - without taking any accountability for actually getting it done. Execution is everything.

SEO is such a unique channel in that you need to touch the website to make it work. With Google Ads, for example, you can do almost everything in-platform and you don't really need to make changes to the site infrastructure. SEO is different - you have to get your hands dirty.

If you are not in control of that process, everything goes too slow. Trust gets lost quickly because you can't move fast enough to show results.

This is why we changed our model to bring development in-house. To start with, we hired developers just to implement basic stuff. But as we have gone deeper, we started building our own Shopify modules to handle all kinds of things, from secondary navigations to filtering logic. Quite often now, we will completely rebuild templates to allow us all the components that we need to execute quickly.

The accumulation of friction

We have covered a lot of ground here, but if we had to summarise why these campaigns fail, it is rarely down to just one thing.

It is a combination of factors. It is applying a generic playbook to a complex business model. It is the friction between rapid merchandising and long-term authority. It is operating in a strategic void without a conductor to manage the silos.

And finally, it is the jarring disconnect of trying to solve a wicked problem with a simple tactic.

These are lessons that have been hard won. We didn't know all of this when we started. We learned it by making mistakes, by seeing projects stall, and by realising that the standard agency model wasn't built for this level of complexity.

Whether you are considering working with us, trying to unpick why your current strategy isn't landing, or just mapping out a new project, hopefully these lessons will be useful.