Short answer. Add a co-founder only when the business has an enduring ownership-level gap and the relationship survives a real working trial. Do not exchange permanent equity merely for temporary labour, loneliness or investor signalling.
Choosing a co-founder is not the same as filling a vacancy. A capable employee can leave; a shareholder can leave the job and still own shares. An adviser can challenge a decision without sharing control. A contractor can deliver a defined result without becoming a permanent partner in every later result.
There is no credible universal rule that two founders outperform one. The useful question is narrower: which failure risk does another owner reduce, and which new risks does that person create?
Start with the gap, not the candidate
A tempting candidate can make an owner invent a role. Reverse the order. Define what the business must do over the next 18 to 24 months, then identify the gaps that could stop it.
Use five risk categories:
| Risk | Evidence of an ownership-level gap | Evidence that another arrangement may be enough |
|---|---|---|
| Critical capability | The capability shapes the product, proposition or operating model every week and cannot be specified cleanly in advance | Work has a clear brief, market rate and measurable output |
| Sustained workload | Two demanding leadership functions must run concurrently for a prolonged period | The pressure is a launch sprint or a backlog that better systems can remove |
| Capital | The person is investing meaningful capital and will share downside as an active owner | Money is the main contribution and a documented loan or investment route is more accurate |
| Decision quality | The business needs a continuing counterweight with authority in a domain the founder cannot judge alone | Periodic challenge from an adviser, customer panel or non-executive role is sufficient |
| Resilience | A single person's absence would stop a critical function and cross-cover must be built from the start | Procedures, delegation and an early employee can reduce key-person dependence |
Loneliness is real, but ownership is a poor cure for it. So is the belief that investors expect a pair. A second name on a pitch deck does not demonstrate complementary skill, trust or delivery.
If the founder is mainly short of hours, first model whether the business can afford its first employee. If the problem is that the founder is still doing every operating task, read the lesson in knowing when a founder should stop being chief executive. Neither diagnosis automatically requires another owner.
Research can inform the test, not make the decision
Research on entrepreneurial teams is easy to turn into a slogan and difficult to apply to one company. A 2017 meta-analysis of 55 empirical samples and 8,892 observations found that team-composition characteristics have distinct relationships with new-venture performance. It did not establish that adding any co-founder improves any venture.
The US study Early Joiners and Startup Performance examines the people who join young employer businesses and their relationship with later performance. Its practical value here is the distinction it exposes: important early capability may come from an employee, not necessarily an original shareholder. The study's population, data and research design do not settle how a UK microbusiness should allocate equity.
Other research cuts against a simple team-size rule. Sole Survivors: Solo Ventures Versus Founding Teams reported results from a particular crowdfunded-venture sample that challenge the assumption that founding pairs always outperform solo founders. That sample is not representative of all start-ups, and association is not a personal forecast.
Treat team research as a warning against two errors: assuming solo is necessarily fragile, and assuming more founders necessarily create more capability. What matters is the relevant capability, how the people coordinate and the structure around their decisions.
Compare four ways to close the gap
Do not compare “co-founder” with “do nothing”. Compare four complete options.
Co-founder
Best suited to an enduring, uncertain leadership domain where the person will shape strategy, accept broad accountability and share long-term downside. The cost is ownership, influence over major decisions and a relationship that must survive disagreement, low cash and changing roles.
Employee
Best when the company can define a role, supervise it and fund salary plus employment costs. Employment can still carry substantial responsibility. Theflght's guide to the first service-business employee helps test affordability rather than assuming equity is “free”.
Contractor or specialist supplier
Best for bounded work such as a prototype, compliance project, campaign or implementation. Specify outputs, access, intellectual-property treatment and handover. A recurring strategic dependency disguised as a project may indicate a deeper gap, but it still does not prove co-ownership is the answer.
Adviser or non-executive support
Best when the founder needs judgement, introductions or periodic scrutiny rather than daily execution. Define cadence, conflicts, confidentiality and the difference between advice and decision authority.
Score each route from 0 to 2 against duration, ambiguity, required authority, cash affordability, downside sharing and replaceability. A co-founder should win because the work demands an owner, not because salary cash is scarce.
Run a 30-day working trial before discussing the final split
A trial should resemble the work that will strain the relationship, not a series of friendly planning calls. Do not issue shares merely to create the trial. Have a UK solicitor advise on the trial terms, confidentiality, intellectual property, data access and employment or contractor status.
Days 1 to 3: write the trial brief
Choose one live outcome, such as securing and delivering a paid pilot, validating a technical constraint or building a costed launch plan. Record:
- the customer or operating result required;
- each person's role and weekly time commitment;
- the cash and information each may commit;
- decisions either person may make alone;
- decisions requiring both people;
- what evidence will count as completion;
- what happens to work, access and materials if the trial stops; and
- a fixed review date.
If one founder is testing around employment, check their contract and conflicts, then use the practical limits in testing an idea while working full-time.
Days 4 to 24: create useful friction
Include at least one customer conversation, one money decision, one missed assumption and one piece of difficult feedback. Observe behaviour rather than chemistry:
- Does the person surface bad news early?
- Can each explain a decision with evidence?
- Who does unglamorous follow-through?
- Does disagreement improve the work or merely delay it?
- Are quality standards compatible?
- Does either person make commitments outside agreed authority?
- Can the pair change a view without keeping score?
Keep a short decision log. Memory becomes generous when people want a partnership to work.
Days 25 to 30: review separately, then together
Each person should answer the same questions before comparing notes:
1 What did the other person own without prompting?
2 Which decision improved because both were involved?
3 Which decision became slower or less clear?
4 What important contribution cannot be bought through a normal commercial arrangement?
5 What behaviour would become intolerable under financial pressure?
6 Would I still choose this person if funding were not involved?
“We completed the task” is insufficient. The decision concerns repeated collaboration under uncertainty.
Practical asset: the founder decision-rights worksheet
Complete this before negotiating percentages. It is a commercial brief for professional advice, not a legal agreement.
| Decision area | Proposed lead | Other person's input or veto | Evidence required | Escalation if unresolved |
|---|---|---|---|---|
| Product or service scope | ||||
| Pricing and material discounts | ||||
| Hiring and dismissal | ||||
| Spending above an agreed limit | ||||
| Borrowing, guarantees or security | ||||
| New shares or investors | ||||
| Intellectual property and licensing | ||||
| Brand and public statements | ||||
| Founder pay and expenses | ||||
| Sale, closure or major change of direction |
The worksheet reveals whether “equal partnership” means equal economics, equal votes, equal workload or mutual veto. Those are different propositions.
The GOV.UK shareholder guidance confirms that a UK company limited by shares can have one shareholder and that share rights can differ by class. The formation-document guidance also requires prescribed particulars covering dividends, redemption and voting rights. Do not improvise a split from a headline percentage.
Tax is another reason to stop before issuing equity. HMRC states that there is no special statutory category of “founders' shares”; a founder who is a director from the start can be within the employment-related securities rules. The detail is fact-specific, so have a tax adviser review the timing, valuation and restrictions, using HMRC's founder-share guidance as context rather than treating it as personal advice.
Do not make the share split the first negotiation
First agree contribution, time, cash, decision authority, expected pay, intellectual-property ownership, departure scenarios and the intended financing route. Only then ask solicitors and tax advisers to design coherent documents.
Also decide what the brand depends on. If customers are buying one founder's reputation, an equal operational partnership may still leave concentrated commercial risk. The guide to personal brand versus company brand helps expose that dependency.
Pause the co-founder route if the candidate will not run a trial, avoids money conversations, contributes only a vague network, expects equity for a short project, refuses clarity about other commitments or treats legal and tax review as mistrust.
At the trial review, choose one of four recorded outcomes: proceed to professional structuring; extend the trial for a specific unresolved risk; use an employment, contractor or adviser route; or stop. A clean “not now” is valuable evidence, not a failed partnership.
Sources
- GOV.UK: Choose your shareholders for companies limited by shares - accessed 9 November 2026.
- GOV.UK: Prepare documents agreeing how to run your company - accessed 9 November 2026.
- HMRC Employment Related Securities Manual: Founders' shares - updated 5 May 2026; accessed 9 November 2026.
- Kerr and others: Early Joiners and Startup Performance, NBER Working Paper 28417 - revised October 2022; published in The Review of Economics and Statistics in 2025; accessed 9 November 2026.
- Jin and others: Entrepreneurial Team Composition Characteristics and New Venture Performance - A Meta-Analysis - 2017; accessed 9 November 2026.
- Greenberg and Mollick: Sole Survivors - Solo Ventures Versus Founding Teams - working-paper record; accessed 9 November 2026.
Cover photograph: PO1_0618 by collision.conf. by. Adapted with an editorial headline and The FLGHT branding.
Related guides
Comments (0)