By the time a term sheet lands, the investor and their lawyers have done this dozens of times. You probably haven't. That asymmetry shows up in predictable places, and it costs founders more than it should.

This isn't a guide to fighting your investors. Most institutional investors are reasonable, and a functioning relationship with your cap table matters long after the documents are signed. But reasonable doesn't mean the opening draft reflects your interests. It reflects theirs, built on templates developed by industry bodies that represent VCs, refined over years of transactions, and increasingly generated at speed by AI trained on those same standards.

Nobody is doing that work for founders. So here is where to focus.

Before the Term Sheet: The Problems That Surface at the Worst Moment

Two issues consistently derail or delay funding rounds, and both are fixable long before a term sheet exists.

The first is share promises that were never properly documented. A conversation at the kitchen table, an email thread, a handshake with an early advisor or contractor. If shares or options were promised and never formally issued or documented, they will surface during investor diligence. VCs do not like surprises on the cap table. The momentum built over weeks of negotiation can stall over something that would have taken an afternoon to resolve six months earlier.

The second is intellectual property that the company does not actually own. English law gives the first ownership of any IP to the person who created it, not to the company that paid for it, unless that person was an employee creating it in the normal course of their work.

Contractors are not employees. If your product was built in whole or in part by contractors working under agreements that do not assign IP to the company, the company may not own its core asset. Investors will require this to be resolved before closing. Resolving it after the fact is considerably harder if the contractor is no longer reachable.

Neither issue is complicated to address in advance. Both are serious if left until a deal is in progress.

Your Shares: Understanding What Vesting Actually Means for You

Vesting provisions are standard on institutional rounds, and for good reason: they protect the company if a founder leaves early. What is less often explained to founders is where the standard terms leave room to negotiate, and where they carry hidden risk.

The typical structure runs over 36 or 48 months on a linear basis, with a 12-month cliff during which no vesting applies. On a bad leaver event, which includes fraud, gross misconduct, and breach of any non-compete or employment agreement, all shares are forfeited regardless of the vesting schedule. On a good leaver event, only unvested shares are forfeited.

Three things are worth understanding clearly before you sign.

Resignation used to be treated as a bad leaver event as standard. UK Private Capital changed this position in 2023, recognising that a founder choosing to leave should not automatically forfeit everything. Not all term sheets reflect this change. Some revert to the old position. If your term sheet treats resignation as a bad leaver event, that is a negotiating point, not a fixed term.

If your company is not brand new and you have already been building for some time, potentially having navigated a seed round, there is a reasonable argument that a portion of your shares are already earned and should not be subject to vesting. This is not always conceded, but it is a legitimate position to put forward.

One technical point that matters more than it appears: the standard bad leaver forfeiture applies to all shares, regardless of the vesting schedule. If the intention is that a bad leaver should retain some shares in recognition of time served, that requires a specific drafting change. The default vesting structure does not deliver that outcome automatically.

The Register of Members Is Not Your Cap Table

This distinction matters and is frequently misunderstood.

Your company is legally required to maintain a Register of Members. This is the primary legal record of who owns shares in the company. It is not the same as a cap table and it is not the same as your filings at Companies House.

The cap table should reflect the Register of Members, not the other way around. In practice, many early-stage companies maintain a cap table as their working document and let the Register fall behind. This creates discrepancies that are straightforward to fix early and genuinely painful to untangle later, particularly if you are trying to reconstruct historical share issuances with incomplete records ahead of a funding round.

The cap table itself is also worth structuring carefully. Shares and options are often presented together as a single fully-diluted figure, which is useful for understanding total potential dilution. But consent rights, board votes, and formal resolutions all operate on issued shares only. A cap table that does not clearly separate shares from options, and allocated options from unallocated ones, can obscure your actual voting position until a moment when it matters.

Investor Consents: What You Are Actually Agreeing To

Investor consent rights are now largely standardised, which makes them easy to accept without reading carefully. That is a mistake.

Consent rights give investors a veto over certain company decisions. The three areas that deserve specific attention are capital-related consents, business-related consents, and the threshold that determines which investors constitute the required majority.

On capital consents: if you are planning to issue further shares or implement an option scheme shortly after closing, those issuances will require investor consent unless specifically carved out. The carve-out needs to be negotiated at term sheet stage. After signing, the conversation is considerably harder.

On business consents: these often arrive as a template. The template was not written with your specific business in mind. Review each consent right against your actual operations. Any consent that covers something routine for your business, something you would need to do in the ordinary course, should be amended or removed.

On the majority threshold: consent rights operated by reference to a percentage of investors can give smaller investors disproportionate influence if the threshold is set too high. A minority investor should not be able to block a decision that the majority of your capital supports. This is worth modelling before you agree to the percentage.

Anti-Dilution Provisions: What They Are and What They Are Not

Anti-dilution clauses have been a standard feature of institutional funding documents for over twenty years. They exist to protect investors in a down round by adjusting the conversion price of their shares to compensate for the reduction in value.

There are two main types. The US model, used by the NVCA and more widely across international markets including the Middle East, and the European model used in UK Private Capital terms. The US model is simpler to administer. The current UK standard uses the European model.

Two things are worth understanding as a founder.

First, the rebalancing effect of an anti-dilution clause is rarely substantial unless the down round is severe. For most companies, it functions as an anti-embarrassment mechanism rather than a significant economic correction. It is worth understanding the mechanics, but it is not usually the most consequential term in the document.

Second, the formula has a structural quirk: when applied to multiple successive down rounds, the corrective adjustment from an earlier round is not automatically factored into the weighting used in the formula. The mathematics compounds in ways that are not always intuitive, and founders, who are not typically also investors, bear most of the dilution when the formula runs. A sunset provision, one that terminates the anti-dilution right after a defined period or upon a higher-priced round, is a reasonable position to negotiate. Series B investors who were not part of your Series A will almost certainly resist carrying earlier investors' anti-dilution rights forward.

These clauses are increasingly appearing in seed round documentation. They were designed for Series A and beyond, where valuations are more formally established. Accepting them at seed stage, where valuations are less formulaic and the circumstances of a future round are harder to predict, carries more risk than it might appear.

Warranties: What Happens When Founders Are Asked to Give Them Personally

The standard UK Private Capital position is that warranties are given by the company, not by the founders individually. This matters because the liability caps and limitations set out elsewhere in those terms are calibrated on that basis.

A growing number of VC-driven drafts depart from this standard by requiring founders to give warranties personally in addition to the company. When this happens, the liability protections designed for company warranties do not automatically extend to the founders. Without additional drafting, a founder could theoretically face personal liability at a level that bears no reasonable relationship to their individual circumstances.

This is not always an intentional overreach. Drafts sometimes include the requirement without including the corresponding protections, possibly as an oversight. But the consequence is real. If personal warranties are required, the document needs to include explicit caps on founder liability, typically set at a figure considerably lower than the total funds raised, along with the standard suite of limitations that would apply to the company.

If the opening draft asks for founder warranties without those protections, the protections need to be added before signing, not assumed to exist elsewhere in the document.

Fees: Less Standardised Than Everything Else

Deal fees are the least discussed and least standardised element of a funding round. A few benchmarks are useful to have.

Arrangement fees, where charged, typically run at 2 to 3% of funds introduced, though there is genuine disagreement among VCs about whether they should be charged at all. Larger funds with an international footprint tend to be less likely to deduct an arrangement fee from the capital introduced. This is worth understanding before interpreting an arrangement fee as a fixed market norm.

A contribution to VC legal fees is common and is typically capped at around £30,000 of fees incurred. A quarterly monitoring fee, familiar from seed rounds at roughly £20,000 to £30,000 per year, usually falls away at Series A. For investors with a private equity background, this is sometimes replaced with a director's fee contribution, which can run to £50,000 per year. That is on the high end of the market, and it is negotiable.

None of these fees are fixed. All of them should at least be discussed. The final outcome will reflect bargaining position, but you cannot negotiate terms you have not raised.

One Structural Change Worth Making Early

Non-voting shares are underused at early stages. Most stakeholders who are not founders, early employees holding options, advisors taking small equity positions, have a primary interest in future economic return rather than governance. Separating economic rights from voting rights through a distinct share class is a straightforward change to the articles that preserves founder decision-making authority through later stages of the company's life. It is most practically implemented when the cap table is simple. It becomes harder, though not impossible, to introduce once the shareholder base has grown.

The Consistent Pattern

Every point in this article has the same shape: a term that arrives looking standard, a detail that is not in your favour, and a moment, usually at term sheet stage or in the opening draft, where the conversation is still open. After that moment, reopening the point becomes harder.

The investment process is well-documented from the investor's side. The founder's side is less so. That gap is not usually the result of bad intent. It is the result of a process that has been industrialised in one direction and not the other. Knowing where to look is most of the work.