No primary source defines a universal window for contacting a newly incorporated company. Timing is a decision your practice tests and measures — not a fact you can look up.
Incorporation date tells you when a company joined the Companies House register. On its own, it does not tell you whether the company is trading, whether a director wants to hear from an accountant, or when they might. Guidance that claims a fixed window — that most founders decide within a month, or that the decision is "made" by month three — is not supported by Companies House, HMRC or ONS material. It is an assumption presented as a rule.
This guide sets out what the public record can and cannot tell you about timing, then gives you a versioned framework for testing initial-contact timing against your own audience, channel and capacity. For the complete workflow this page sits inside, see how accountants can reach newly incorporated companies.
What the public record can and cannot tell you about timing
Companies House gives you three timing-relevant facts about any company, and each has a specific limit.
- Incorporation date. This is the date the company was entered onto the register, not the date it started trading. HMRC treats a company as active only once it is "carrying on a business activity such as a trade or professional activity" or otherwise generating income — forming a company does not itself make it active (GOV.UK, Corporation Tax: trading and non-trading, updated 17 February 2020). Some companies incorporate months before they trade; some never trade at all.
- Company status. "Active" is a register status, not proof of economic activity. It tells you Companies House hasn't dissolved, struck off or otherwise closed the company — nothing about what the company is doing.
- Filing due dates. The confirmation statement and first accounts have statutory deadlines set from incorporation, and you can read them from the public record. They tell you a compliance event is coming up for the company — a reasonable, disclosable reason to write — but not whether the company already has an adviser, wants one, or has even thought about the deadline yet.
The registered office address is worth naming separately, because it is easy to over-read: it is a statutory contact address, not necessarily where the company trades or where a director lives, and it may belong to an accountant, a formation agent or another service-address provider. See registered office versus trading address for how to check before you assume otherwise.
None of these fields, alone or combined, tells you when a director is ready to think about an accountant. Earlier contact does not establish a better outcome than later contact, and later contact does not establish a better outcome than earlier — the record simply doesn't carry that information. The honest response is to test, not to guess.
Build a timing-test card
Rather than a claimed best window, run a small, versioned test: pick comparable cohorts, assign each to a different initial-contact band, record what happens, and let the result decide what your practice does next. The bands below are examples to compare, not a predicted winner.
| Field | What to define before you start |
|---|---|
| Cohort definition | Companies matching your sector and location filters, incorporated within the same rolling period. Record the filter criteria and the incorporation-date range for each cohort. |
| Channel | One channel per test — for example, post only. Testing more than one channel at the same time means you can't attribute a difference in outcome to timing alone. |
| Weekly capacity | The number of companies your practice can review and contact consistently each week. Keep it constant across every band you compare. |
| Initial-contact bands to compare | Assign comparable cohorts to bands by days since incorporation — for example, week 1–2, week 3–6 and week 7–12. These are bands to test, not a claimed winner. |
| Restrained follow-up rule | One follow-up only, at a fixed interval set in advance (for example, three weeks after the first contact), applied identically across every band. See contacting new companies: the legal rules for the channel and objection rules that govern whether and how you may follow up — this page draws no legal conclusions of its own. |
| Outcome definitions | Decide, before you start, what counts as a response, an engagement and an objection or opt-out, and record every contact against the same definitions. |
| Confounders to record | Sector, location, letter version, sender and time of year — anything besides the contact-timing band that could explain a difference in outcome. |
| Stopping rule | Fix, in advance, how many contacts per band and how many weeks of data you need before comparing results, and stop at that point regardless of the interim numbers. |
| Version / date | Label each run (v1, v2…) with the date it started. When you change a field, start a new version rather than editing the one in progress. |
Read the result only after the test has run for at least eight weeks — long enough for the latest band's follow-up window to close and for genuinely slow responses to arrive. Compare outcomes within the definitions you fixed in advance, note the confounders you recorded, and treat the result as evidence for your practice's next version of the test, not as a rule to publish or reuse elsewhere.
Why a steady weekly routine matters
A steady weekly routine matters, but not because it hits a hidden window — it matters because it keeps a test valid and your capacity used well. Companies incorporate every working day, so a one-off batch of outreach only ever covers one cohort; the following week's incorporations are a separate cohort you'll miss entirely unless you review on a regular schedule.
Reviewing at the same time each week also keeps your bands comparable. If cohorts are assigned to initial-contact bands only when you happen to get around to it, the bands stop being comparable and the test stops telling you anything. Recording the contact date and any follow-up for every company you write to — reply, no reply or objection — is what lets you read a test back later, whether you keep that log in a spreadsheet or a CRM.
Once you've decided which timing bands to test, the next decision is what the letter itself should say. See writing an introduction letter to a new company director for a one-page structure to adapt.
Frequently asked questions
Is there a best day to contact a new company?
No primary source defines one, and this guide doesn't invent one either. Use the timing-test card above to find out what works for your practice's audience, channel and capacity.
Is it too pushy to contact a company in its first week?
There's no evidence either way. Some directors are still handling setup tasks in week one; others are already looking for an accountant. Treat "too soon" as something to test, not a rule to follow.
Does incorporation mean the company needs an accountant?
No. Incorporation is entry onto the register, not proof of trading, need or readiness. Filing deadlines and VAT or payroll thresholds are possible reasons a company might want advice, but the record alone can't tell you whether this particular company does.
How Launch Register can help
Running this test manually means keeping a log of contact dates, bands and outcomes up to date every week. Launch Register records the contact date and follow-up decision against each company you review, so you can read a timing test back later without rebuilding the log by hand — the bands you choose to compare and the decision to contact remain the practice's. See how Launch Register works for the full record-keeping workflow.
Test your own timing, not someone else's guess
Review up to 10 personalised company matches at no cost, record your contact dates, and read back what works for your practice.
Personalised sample of up to 10 companies · no card required · £29/month to continue