A Shopify App Detector Shows State, Not Change

A Shopify app detector reads a storefront's public source and reports the apps it can fingerprint: the review widget, the subscription app, the page builder, the email capture form. That is a state, a list of what is installed at the moment you looked. It is accurate and it is useful. It is not, on its own, a reason to email anyone this week.

The disclosure first. We build Keaz Signals, which sells ecommerce buying signals to the people doing the selling. We are not a detector, and we have an interest in the distinction this post draws. Better said at the top than left in a footer.

This is written for the person selling to the store, not for the merchant. Most writing about an ecommerce tech stack is a list of tools a shop ought to install. That answers a different question than yours, which is narrower: of the stores I could contact this week, which one has a reason to reply.

State is what a store has installed. Change is what it installed, removed or replaced, and when. A detector gives you state on demand. You only learn change by looking at the same store more than once and keeping what you saw last time. Almost every practical outbound decision runs on change.

Take one row in one table. A store that has run the same email tool for three years is not a prospect for an email implementation. The setup is finished, the budget is spent, nobody inside the business is currently thinking about it. A store that installed the same tool nine days ago is a different business entirely: the flows are still defaults, someone is accountable for making the spend work, and the question of who helps them is open.

Same app, same detection, opposite value. A lookup cannot separate the two, because it has no memory.

That is why we re-scrape the stores we watch weekly rather than maintaining a directory. The point of the weekly pass is not a fresher list. It is that the difference between this week's pass and last week's is the product.

Three stack events carry enough weight to open with, in our experience of building the signal set: a retention or email tool going in, a storefront rebuild, and ad activity restarting or climbing. What they share is not the technology. Each is a decision the store already made and already paid for, which means someone inside the business now has to make it work.

• An email or retention tool installed: the setup window, published as stores that just installed an email tool
• A storefront rebuild or theme change: budget already in motion.
• Meta ads going live, or the active ad count rising: acquisition spend committed.

Product launches, newsletter rhythm and social growth sit alongside them in the buying signal catalogue. The test is always the same: does this event imply unfinished work inside the business.

Most of them are noise. A payment provider, an analytics tag, a cookie banner or a theme that ships by default tells you nothing about whether the store wants to hear from you. Neither does an app that most stores in the category run: a fact that is true of everyone sorts nobody, and a segment built on it is just the whole market with extra steps.

Three more traps are worth naming. Long-installed tools look identical to new ones in a lookup, and they are the majority of any detection. Fingerprinting is inference, not a manifest, which is why two detectors disagree about the same storefront — script tags get inlined, apps get embedded in a theme, and some of what is reported as an app is leftover code.

And absence is weak evidence. A store with no review app might be a prospect, or it might have decided against one, or run headless, or keep the thing behind a login where nothing public can see it. Building a campaign on what you cannot see is the most expensive way to use this data.

Write from the event and the gap it leaves, not from the tool name. "You use Klaviyo" is a fact the reader already knows about their own business, and it reads like a mail merge because it is one. "You put a retention tool in nine days ago" is an observation about their week.

Operationally that changes three things. You segment on the change plus the fit conditions — country, platform, follower count — rather than on the app. You send while the event is still recent, which is the only reason freshness is worth paying for. And you write one message per event rather than dripping a static list, because the second message has nothing new to say unless something new happened.

That is the shape our ecommerce leads database is built around, and the per-store copy comes from the signal plus your own knowledge base rather than a template library.

We are not going to give you install counts, app market shares, or a percentage of Shopify stores running any given tool. We have not measured those, and the figures that circulate for them are not traceable to a method we can check. A number we cannot stand behind is worth less to you than this sentence is.

We are also not going to tell you that a stack signal converts better than any other kind. We have not run that comparison, and anyone quoting a reply rate for a signal type is quoting their own campaign, not yours.

What we will state, dated: as of September 2026 we watch 4.1 million Shopify and WooCommerce stores across Europe and North America, re-scraped weekly, with roughly 41,200 carrying a fresh signal in any given seven-day window. Those numbers move, which is why they carry a date every time we use them.

If your question is what this one store is running right now, a detector answers it directly and we do not. Lookup is a real job: qualifying an inbound lead before the call, checking a client's stack, auditing a portfolio you already own.

Coverage is the second place we lose. Detectors index whatever they can crawl. We cover Shopify and WooCommerce stores in Europe and North America and nothing else. If you sell to Magento shops in Brazil, we are the wrong tool and no framing changes that.

Two more, plainly. We do not export: leads move into campaigns and stay there, so if you need a file to hand to someone else, that is a dealbreaker and it should be. And we have no LinkedIn signals and no LinkedIn sending, with no plans to add them.

If it is the event rather than the state you are missing, the signal catalogue is the honest list of what we watch, and the ecommerce leads database is where that change becomes a segment you can send to.

• Keaz Signals buying signal catalogue — retrieved 19 September 2026
• Our own coverage: 4.1M Shopify and WooCommerce stores watched, ~41,200 with a fresh signal in seven days — as of September 2026
• Weekly re-scrape cadence — own product behaviour, September 2026
• Adjacent: intent data by category · Meta Ad Library signals