A founder’s two-minute recurring search that keeps pipeline full

The most reliable pipeline habit for a founder is one saved search: your codes, your patch, one signal, re-running weekly and delivering only companies you have never seen. Two minutes to set up, never needs doing again.

The most reliable pipeline habit a founder can build is not a tool stack or a daily prospecting hour. It is one saved search: the SIC codes you sell to, the area you cover, one public record signal, set to re-run weekly and deliver only companies you have never seen before. Setting it up takes about two minutes, once, and it feeds you for as long as you leave it running. With Companies House official statistics counting just over 800,000 new incorporations in the year to March 2025, the feed does not run dry.

Why founders stop prospecting (and how the search fixes it)

Every founder knows the loop: sell hard until the pipeline empties, then scramble to refill it, then sell hard again. Prospecting is the first thing dropped in a busy week because it is the only revenue activity with no external deadline. The fix is not discipline, it is removing the recurring effort entirely: make the finding automatic and keep only the judging and the writing, which are the parts worth a founder's time. That is the whole design of the two-minute search, and why it anchors the founders playbook.

The setup: three choices

Choice 1: who you sell to

Pick the SIC codes that describe your buyers. Type the plain-English trade ("plumbers", "restaurants", "recruitment agencies") and let autocomplete resolve the codes, or work from our SIC code guide if your market spans several. Resist the urge to select everything: a founder can work perhaps a few dozen leads a week properly, and narrower codes mean sharper first lines.

Choice 2: where

Your region, city or postcode areas. If you sell nationally, skip it. If any part of your pitch is local ("we're based here too"), the filter earns its place, because locality is still the cheapest personalisation in B2B.

Choice 3: one signal

This is what separates a list from a reason to write. Pick a single public record signal that matches your product:

• Newly incorporated if new companies need you early: accountancy, banking, insurance, software, websites. (Why the early window matters so much is its own post: first 90 days.)
• No website found if you sell anything web-adjacent: the method behind the flag is in how to find companies with no website.
• A filing deadline approaching if you sell compliance, accountancy or admin relief: see the confirmation-statement window.

One signal, not three. Stacking filters shrinks the pool and muddies the message; a signal per search, and a search per pitch, keeps both clean.

Set it to weekly and walk away

Save the search with a weekly cadence and a per-run cap that matches your capacity. Two guarantees make the recurrence trustworthy, both explained fully in recurring searches: every run covers everything since the last one, so nothing falls between runs, and you never receive the same company twice, deduped against everything you have ever been delivered. You pay only for delivered leads, so the cap is also your budget, fixed and predictable.

That is the two minutes. Codes, patch, signal, weekly, cap, save.

The Monday routine

What arrives each week is a short, fresh batch: companies matching your profile, each enriched with a named director and a verified business email. The founder's working rhythm on top of it:

1. Scan the batch over coffee. Archive the poor fits without guilt; the cap means next week brings more. 2. Write to the good ones the same day, while the batch is freshest. Short, specific, one idea per message. 3. Reply fast to anyone who answers. Speed compensates for everything a solo founder lacks in polish.

Fifteen focused minutes on Monday beats the binge-and-drought loop every time, because it happens every week.

Three worked examples

The same three choices, made three different ways:

• A bookkeeping founder in Bristol: accountancy-adjacent SIC codes are the wrong axis here, so instead: all SIC codes, Bristol postcodes, newly incorporated, weekly, cap 30. Every new company in the city needs the books doing; the signal is the pitch.
• A web-design founder in Leeds: creative and trade SIC codes they have built for before, Yorkshire, no-website flag, weekly, cap 25. The observation that opens the email is the filter itself.
• An IT services founder selling nationally: professional-services SIC codes, no region, confirmation statement due within 60 days as a proxy for "admin is on the director's desk this month", weekly, cap 40.

Three founders, one mechanism, three completely different pipelines. None of them will ever build a list by hand again.

When you outgrow the manual Monday

The natural next step is letting the search feed a Flow: each week's new matches are verified, deduped against everyone you have contacted, sent your intro from your own mailbox, and chased once if nobody replies, with replies stopping the sequence instantly and landing back with you. The search does the finding, the flow does the sending, and the founder's job shrinks to the only part that ever needed them: the conversations. Built-in outreach handles warm-up and sending limits underneath so deliverability survives your enthusiasm.

The honest maths

No invented conversion rates here: yours will depend on your product, your patch and your copy. The claim is narrower and more defensible. A pipeline habit that costs two minutes once, plus fifteen minutes a week, gets kept, and with over 800,000 companies incorporating in a year (Companies House official statistics) on top of the five million-plus already on the public record, the supply side is not your constraint. Consistency is, and consistency is exactly what a recurring search automates.

Start free with 10 leads: build the search this afternoon, and let next Monday be the first one that stocks itself.