SEIS Due Diligence: What to Check Before You Invest

SEIS due diligence is two jobs wearing one name. The first is ordinary angel judgement: is this a business worth backing at this price? The second is peculiar to the scheme: will the tax relief everyone is assuming actually survive contact with HMRC? I have passed on more deals than I have joined, and both jobs have earned their keep. Here is the working checklist.

Rule zero: the company must stand without the relief

Write it somewhere visible: a weak company with 50% relief is a slower way to lose money. The relief improves the mathematics of a good decision; it cannot convert a bad one, as the risks guide sets out bluntly. Do the business diligence first, as if SEIS did not exist, and only then let the scheme sweeten terms you already liked.

Layer one: the business questions

• Team: have the founders done hard things before, and do they listen? At this stage you are backing people, not spreadsheets.
• Problem and market: is anyone demonstrably paying for this, or plausibly about to? Letters of intent beat surveys; revenue beats both.
• Runway arithmetic: what does this raise buy, in months and milestones, and what has to be true for the next round to happen?
• The cap table: sensible founder ownership, no dead equity, no mystery holders.
• Terms: valuation you can defend out loud, ordinary shares, no structures that quietly favour someone else.
• The exit story: not a promise, but a plausible shape: who buys companies like this, and why?

Layer two: the SEIS due diligence questions

This is the diligence most investors skip, and the one that protects the reliefs described in the reliefs guide.

• Advance assurance: ask to see the letter and the application behind it. Does the assured raise match the round in front of you, same class, similar amount, same trade?
• Eligibility headroom: where does the company sit against the rules and limits: trade age, assets before this issue, headcount, and remaining capacity within £250,000 after grants?
• The instrument: new ordinary shares for cash. Convertible loan notes fail; casually drafted advance subscription agreements can too.
• Your own connection: count your family holdings against the 30% line, and keep any involvement non-employment.
• Paperwork competence: who files the SEIS1, when, and how quickly have they issued SEIS3s before? An administrator who shrugs is a relief delay in human form.
• The three-year posture: any plans, exit chatter, restructures, that could break conditions inside the window?

Red flags that end my conversations

• Tax relief leading the pitch, business case trailing it.
• No advance assurance and no good reason.
• Guaranteed returns, guaranteed buybacks, anything with the word guaranteed.
• Pressure to complete before a date that serves the founder, not the paperwork.
• Terms that de-risk you contractually: they usually de-SEIS you too, since the scheme requires genuine risk.

A process that fits real life

You are not running an M&A department. A workable rhythm: one call for the story, one pass through the data room against the two checklists above, one conversation with a customer or reference, and one hour of professional advice where anything smells complicated. Then decide, at a cheque size that respects the failure odds and leaves room for the portfolio you are actually building.

The data room: what to actually request

Ask for, and expect within days: the advance assurance letter with its application; the current cap table with any options or ASAs noted; articles and any shareholders’ agreement; last accounts and the live management numbers; the forecast behind the raise; the proposed share terms; evidence of SEIS capacity, grants included; and the founder’s own eligibility answers in writing. The response tells you as much as the contents: organised founders send a link within a day, and chaos in the data room predicts chaos in the compliance paperwork your relief will depend on.

The one-hour red team

Before signing, spend one deliberate hour trying to kill the deal. Argue the bear case out loud: why this team misses, why the market shrugs, why the round after this one never happens. Reread the terms hunting for anything that protects you contractually, then remember that protection and SEIS do not mix, the scheme requires genuine risk. Check your own position last: family holdings against 30%, any work drifting toward employment, and whether this cheque fits the portfolio you planned rather than the enthusiasm you feel. Deals that survive an honest hour of this are the only ones worth your locked-up years.

Diligence proportional to the cheque

A £2,500 cheque inside a planned portfolio does not justify a week of work; a £50,000 conviction bet does. Scale the process, never the principles: at any size, see the assurance letter, confirm the instrument, and check your own connection status, because those three protect the relief itself. What scales is the business layer, one founder call and a reference at the small end, customer conversations, cohort numbers and a professional once-over at the large end. The unforgivable pattern is the inverse one, big cheque, small homework, and it is exactly the pattern excitement produces.

After you invest: the light-touch watch

Diligence does not end at completion, because the reliefs run on for three years. Read the investor updates, and notice silence, three quiet months from a seed company is information. Watch for the events that threaten status: talk of acquisition, listing, odd related-party payments, drift into excluded activity. And keep your own file current, certificate, subscription papers, update trail, so that whichever ending arrives, exit, failure, or the long middle, the relief consequences are a lookup rather than a scramble.

Sources

gov.uk, Apply to use SEIS · HMRC Venture Capital Schemes Manual