How we do SEO forecasting (and why most models are broken)

Written by

Sam Wright

September 16, 2025

3 min read

Switch to light mode

Switch to dark mode

Business cases for SEO are notoriously difficult. Analytics tools tell you one thing, Shopify tells you another, and attribution systems rarely agree. GA4 in particular has made things worse. Sometimes it shows half the revenue Shopify reports. Sometimes it shows double. Either way, the inconsistency tells you everything you need to know: the numbers can’t be trusted at face value.

That doesn’t mean you can’t forecast SEO - but it does mean you need to stop pretending the data is clean.

The pattern we see in Shopify stores

After years of working with large Shopify catalogues, the pattern has become obvious.

• Stores doing £5m+ annual revenue

• 250+ products (often thousands)

• Google Ads as the main acquisition driver

Almost always, these stores show the same thing: organic revenue looks artificially small because so much traffic is dumped into “direct” or “unassigned.”

The reality? If you fit this profile, you should expect around 20% of revenue from organic search. But when tracking is broken (as it usually is), analytics reports often show it as less than 10%.

Our retrospective model

Instead of trying to fix tracking first, we use a retrospective model to correct it. The method is simple:

• Take the last 12 months of reported organic revenue.

• Add 30% of your direct and unassigned revenue.

• Divide by 12 to get a realistic monthly baseline.

That number is a truer picture of how organic is already contributing — even if GA4 doesn’t show it.

What happens next

From that corrected baseline, our standard forecast is straightforward: double organic revenue within 12 months. Growth is modelled as incremental, with minimal uplift in the first two months while implementation happens, then accelerating as new structure and content beds in.

We’ve seen this calculation prove accurate time and again once tracking gets sorted — whether through better tagging, server-side tracking, or just Shopify reporting catching up.

Why this only works in certain cases

This isn’t a one-size-fits-all formula. It’s built for a very specific kind of store:

• Shopify sites above £5m in revenue

• Substantial product catalogues

• Google Ads as the primary acquisition channel

When those criteria are met, the consistency of the pattern gives us genuine confidence in the forecast. For smaller stores, different platforms, or brands without paid as the main channel, the dynamics look very different.

What the numbers don’t show

The model only covers organic revenue uplift. It doesn’t capture the structural upside of a project:

• Taxonomy and product data changes that improve PPC conversion as much as SEO.

• Better curation that makes the store easier for customers to navigate.

• Positioning the brand for the next discovery channel. Meta was the last big one; we believe AI-powered search is the next.

That’s why we think of our forecasting less as a spreadsheet exercise and more as a confidence-building tool. It gives stakeholders a clear baseline to believe in, but the real upside is almost always bigger.

The takeaway

SEO forecasting will never be perfect. The data is too messy and attribution too flawed. But when you focus on the right type of store, recognise the patterns, and apply a consistent retrospective model, you can make forecasts with confidence - and build a business case that stands up to scrutiny.

Sam Wright - Managing Director

Sam is the founder and MD of Blink. He has been working in search engine optimisation since 2007, and is a regular speaker and writer on the subject of eCommerce digital marketing. He is heavily involved in all client projects.