B2B Lead Generation When the Buyer Runs a Store

B2B lead generation is the work of identifying organisations that plausibly need what you sell, finding a person there who can decide, and starting a conversation at a moment when it is worth having. That definition holds across every B2B market. What changes from one market to the next is what an organisation actually looks like.

Ours is a specific market, and the disclosure belongs up front. We build Keaz Signals, an ecommerce leads database with live buying signals. We are a vendor in the category this article describes. What follows is a structural argument about a market, not a neutral survey of tools, and you should read it that way.

The readers we have in mind are agencies and software companies selling to ecommerce stores — not store owners, but the people who sell to them. If you run a Shopify store yourself, this is not written for you.

The generic answer to the question assumes a company with a corporate website, a buying committee, job titles that map cleanly onto functions, and a procurement process with stages you can track. For a large share of the businesses agencies actually sell to, none of that is true. The definition survives; the playbook built on top of it does not.

It breaks in three places: the organisation is smaller than the definition expects, the decision has no committee, and the buying stage is invisible from outside. A three-person Shopify brand has no procurement process to enter and no role accounts to sequence against. The enterprise playbook does not degrade gracefully here — it stops matching.

Take the standard motion as it is usually taught. You build an account list from firmographics: industry code, headcount, revenue band, technology installed. You map the committee. You sequence by role. You wait for a trigger like a funding round, a leadership change or a new office.

Now apply it to a store turning over real money with four employees. Headcount tells you nothing useful about spend, because a small team can run a large ad budget and often does. Industry codes flatten every direct-to-consumer brand into one undifferentiated bucket. There is no committee — there is a founder and perhaps one marketer. Funding rounds are rare, and leadership changes rarer still.

What you are left with is a list of storefronts that look identical on paper. Everyone buying from the same category of source receives the same rows, which is why the same founders get the same message in the same week from three of your competitors. That is a targeting failure, not a copywriting failure, and no amount of rewriting the first line fixes it.

The unit is a storefront, not a company record. A store is a live thing you can observe from outside: it is running ads or it is not, it launched a product this week or it did not, it installed a retention tool or it has not. Those observations are the qualification, and they need nobody to fill in a form.

This changes what a row means. A company record tells you what a business is. A storefront observation tells you what it is doing. The first is stable and mostly useless for timing. The second changes week to week and is almost entirely about timing.

Our own pool gives a sense of the proportions. As of August 2026 the database holds 4.1M Shopify and WooCommerce stores across Europe and North America, and 41,200 carried a fresh buying signal in the preceding seven days. Both numbers move, which is why they are dated here.

The gap between them is the argument. On those figures about one store in a hundred is doing something this week that gives you a reason to write. The 4.1M is a directory. The 41,200 is a work queue.

A buying signal here is an observable change on the storefront itself, not an inferred intent score. New Meta ads going live. Active ad count rising. A product launch. An email marketing tool appearing. Newsletter activity starting or stopping. Social cadence jumping. A storefront rebuild.

What makes those usable is that each carries its own reason to make contact. New ads running means budget is moving and creative becomes the bottleneck before spend does. An email tool going live means retention is being built right now, setup unfinished. A product launch is when marketing support is needed most. On how this differs from scored intent, see what each intent data category can see.

Stores are re-scraped weekly and every signal carries a last-seen date, so what you act on is days old rather than weeks old. An ad campaign in its first week and in its sixth are not the same message, and only one reads as though you were paying attention.

Two signal types people ask for are not there. Registry and firmographic triggers, and hiring signals, are both marked Planned on our own catalogue and are not live. If your motion depends on either, it does not work yet.

Contacts arrive with the store rather than as a separate enrichment step, and they include founder addresses that are not published anywhere on the site. For a four-person brand that is usually the only address worth having, because the founder is the buying committee.

We build these ourselves — from store pages including legal notices, from the Meta ad library, from Instagram and TikTok, and through our own contact discovery and verification. Bought lists are a starting point, never the answer. On why the usual enrichment chain runs out, see where the enrichment waterfall runs dry.

We are not going to publish a hit rate. We have not measured, in a way we would defend in public, what share of stores in the pool carry a reachable founder address — and a number invented for a blog post is worth less than the admission that we do not have one. If that figure decides your purchase, ask for it on a call where we can be specific about your market instead of quoting an average.

Three things this article will not tell you: that a signal-led list lifts your reply rate by some figure, that this approach is GDPR-compliant for your particular use, or how we compare per unit of data against anyone else. Each is either unmeasured, not ours to assert, or dishonest as a comparison.

On reply rates, we offer no guarantee. Any number quoted without your offer, your market and your sending history in it is marketing rather than evidence. Timing helps because the reason to write is real; how much it helps depends on things we do not control.

On data protection, our position is deliberately narrow: we are GDPR-conscious by design and we will not claim more than that. Responsibility for what you send stays with you. A blog post is not the place that gets widened, and anyone telling you their database makes your outreach lawful is selling you something other than software.

On price, what products in this category meter is not the same thing — stores, contacts, credits, exports, sends — and a table lining them up as though it were would mislead you. Our pricing is public. The comparisons worth reading are the ones that name the rows where we lose.

If your buyers are not ecommerce stores, this is the wrong tool and the wrong article. Everything above depends on the buyer having a public storefront that changes in observable ways. A company selling to hospitals, manufacturers or law firms has no equivalent surface, and generic B2B data is the right choice there.

Four more limits, stated plainly because they are real and worth planning around:

• We cover Shopify and WooCommerce and nothing else. Other platforms are not in the pool.
• We cover Europe and North America. Other markets are not in the pool.
• We do not export. Leads move into campaigns and stay there. If your workflow needs a CSV, that is a dealbreaker — and it should be.
• We have no LinkedIn signals and no LinkedIn sending, and no plans to add them. If your channel is LinkedIn, buy something built for it.

Sending runs through Instantly, in your own workspace, with replies syncing back — a strength if you already work there, a constraint if not.

In order: pick the stores you can serve, let a signal decide when to write, write from that signal, and send. Four steps, and the one most teams get wrong is the second — they build the list carefully and then send on a schedule rather than on an event.

Segmenting comes first: country, platform, follower count and other conditions narrow the pool to stores you can actually serve, with a live audience estimate as you build. That is what the ecommerce leads database is for. Timing then comes from the event, and the buying signal catalogue describes each type and why it is worth acting on. The writing is where the AI sales agent composes per store from the signal plus your own services, references and tone — not a template with a first name slotted in.

Signing up gives you 1,000 leads free inside segments you build yourself, which is a cheaper way to test this premise than reading more about it.

• Keaz Signals home page — 4.1M stores, 41,200 with a signal in the last seven days; dated "Updated August 22, 2026". Retrieved 8 September 2026.
• The buying signal catalogue — registry, firmographic and hiring triggers marked Planned. Retrieved 8 September 2026.
• What a B2B lead generation agency actually buys — the agency case, narrowed from this pillar.
• Agency lead generation when your buyers run stores