Every other service contract has an object in it. Working software. A finished building. Goods on a lorry. A design handed across. You can look at the thing, compare it against what was promised, and say whether it arrived. A consultancy agreement is the one contract where the thing bought is a view — and a view cannot be delivered, which means it cannot be accepted either. Almost every agreement in this field pretends otherwise, usually by naming a report and then arguing about whether the report was any good, which is a question no contract has ever settled. So the honest shape is different. Write effort, access and time, because those three are what a consultant actually controls, and an outcome clause promises something that depends on decisions taken afterwards by people the consultant does not manage. It traps the consultant first and the client second. Then the failure that really happens, which is not the one anybody drafts against. The complaint that arrives is never that the advice was bad. It is that nothing happened. The consultant recommended, the client did not act, and six months later both sides are telling the truth about the other. One honest paragraph prevents it — who does what after the recommendation, by when, with what authority — and almost no agreement contains it. Two more things that go wrong quietly. Scope here does not creep; it grows by questions being asked. Nobody sends a change request; the room simply widens, and the duty to notice has to sit with the consultant because the client has no reason to. And you hired a person while the paper named a firm, which matters more here than anywhere else, because judgement is not transferable between staff. Last, the part both sides avoid discussing: a confidentiality clause handles documents and data properly and handles understanding not at all, because nothing could — and a specialist’s next client is often your competitor. Where we stop: we draft, and we put the awkward questions on the table. We promise no outcome, advise on no restriction, and value nobody’s advice.
What this guide covers
Begin with a comparison, because the difference only becomes visible when this contract is set beside the ones it is usually copied from.
Commission software and there is software. Commission a building and there is a building. Order goods and a lorry arrives. In each case something exists at the end that did not exist at the beginning, and anybody can stand in front of it and form a view about whether it matches what was promised.
Engage a consultant and what exists at the end is somebody’s opinion about your situation, which is a different kind of thing entirely and behaves differently in every respect that matters to a contract.
People resist this because there usually is a physical object at the end, and the object confuses everybody.
A report exists. It has pages, it has a date, it can be printed and filed and emailed. So it feels like the deliverable, and the contract names it as one, and everybody relaxes.
But nobody engaged a consultant because they wanted ninety pages. The pages are the container; what was bought is the thinking inside them, and the container tells you almost nothing about whether the thinking is any good.
Which produces the structural fact this entire page is built on.
A view can be expressed. It cannot be handed over, counted, inspected on arrival, or returned as defective.
That is not a complaint about consultants and it is not an argument that the work has no value. It is a description of what kind of thing an opinion is, and contracts that ignore it do not become more rigorous by ignoring it.
Place Order — no advance, pay after work
And acceptance is the mechanism every other service contract turns on, which is why its absence here matters so much.
In an ordinary supply arrangement, acceptance is what moves the risk, releases the money and ends the argument. Somebody looks at the thing, applies a stated test, and the matter is settled. Even where the test is imperfect, there is a test.
There is no equivalent test for advice. Nobody can inspect a recommendation and declare it conforming, because conformity to what would be the question, and the only honest answer is conformity to a view that has not been formed yet.
And yet nearly every consultancy agreement we are shown contains an acceptance clause, because it was adapted from a contract for something else.
Usually the ancestor is a supply agreement or a development contract, where acceptance belongs and works. Somebody changed the nouns, left the machinery, and handed it over. The resulting document promises a delivery, provides for its inspection and sets out what happens if it fails — of a thing that cannot be delivered, inspected or failed in that sense.
It is not usually anybody’s fault. It is what happens when a document is produced by substitution rather than by thinking about what is being bought.
So the machinery attaches itself to the only object available, and the report is pressed into a role it cannot perform.
The agreement says a report shall be delivered by a date, in a stated form, covering stated headings, and shall be accepted or rejected within so many days. Everything in that sentence is checkable except the only thing anybody cares about.
A consultant can satisfy every word of it and still have been no use, and can breach half of it while having been worth three times the fee.
Place Order — no advance, pay after work
Which is where these disputes go when they go, and it is a place from which nothing returns.
The client says the report was superficial, generic, or told them what they already knew. The consultant says it was thorough, that the uncomfortable parts were ignored, and that being told what you already knew is sometimes exactly the finding. Each has a case. Neither has evidence of the kind a contract can weigh.
And because the question is one of professional judgement about professional judgement, the only people who could resolve it are other practitioners in the same field, which is not what either side signed up for.
So the plain statement, which is better heard before signing than discovered afterwards.
No drafting available to anybody makes the quality of advice contractually determinable. A clause can describe what will be done, who will do it, when, and with what access. It cannot establish that a view was a good one. Any agreement that appears to do so has simply hidden the problem inside a word like satisfactory, and that word is where the argument will start.
Which sounds like a retreat and is actually the only honest position, and it produces a better document than the ambitious alternative.
Stop trying to buy a result. Buy the things that are real, describable and within somebody’s gift, and describe them precisely enough that both sides know what they have agreed.
A tightly written effort clause is worth more than a vaguely written outcome clause, because one of them can be complied with and the other can only be argued about.
Place Order — no advance, pay after work
And those real things are fewer than people expect, which is what makes them worth naming carefully.
Five items, all checkable, none of them pretending to be something else.
Underneath that list sits a simpler division, and it is the one we use when a client pushes back on removing an outcome clause.
| Within the consultant’s control | Outside it |
|---|---|
| How much time and attention he gives | Whether your people agree with him |
| What he looks at and whom he speaks to | Whether anybody acts on what he says |
| What view he forms and how plainly he states it | Whether the business improves afterwards |
The left column is what a contract can govern. The right column is what everybody wants to buy.
Because the right column has one thing in common that is worth saying explicitly.
Every item in it depends on your own organisation. Your managers, your priorities, your willingness to do the uncomfortable thing, your capacity to act while also running the business. The consultant has no authority over any of that and usually no visibility into most of it.
Asking a consultant to guarantee an outcome is asking him to guarantee your organisation’s behaviour.
Place Order — no advance, pay after work
And the reason we argue against it even when a client insists is that we have watched what it does, and it damages the side that asked for it more than the side it was aimed at.
It looks like protection. It reads like the client has transferred risk. What it actually does is change the consultant’s incentives on the first day in a direction nobody wanted.
Take the consultant’s position, because it is the easier one to see.
He is now exposed to a result produced by people he cannot direct. A competent one responds either by refusing the engagement, which removes him from your shortlist, or by pricing the exposure into his fee, which you pay whether or not anything goes wrong.
Both of those outcomes cost the client money before a single day of work has been done, and neither improves the advice by a word.
But the real damage lands on the client, and it is quiet enough that most people never connect it to the clause.
A consultant paid on acceptance of results has an interest in recommending what will be accepted. The courageous recommendation — the one that implicates a senior person, or costs more than anybody wants to spend, or says the problem is not the one you asked about — is now the recommendation that threatens his fee.
The most valuable thing an outsider does is say the thing nobody inside will say. A results clause is the most efficient method available for buying somebody who will not say it.
Place Order — no advance, pay after work
There is a milder version of the same mistake that is worth naming separately, because clients reach for it when the strong version is refused: holding back a final instalment until the client is “satisfied”. It sounds reasonable and it is the same clause wearing a coat. Satisfaction is not a test either, and the money at stake is sufficient to make a consultant cautious in the last fortnight, which is precisely the fortnight in which the difficult sentence has to be said.
So the agreement that actually serves both sides looks different from the one most people set out to write, and it is shorter.
It describes a question, a period, a person and a degree of access. It provides for review while there is still time to change direction. It says what the client will make available and what follows if that does not happen. It ends on a date. And it says nothing whatever about whether the resulting view will be correct, useful or acted upon.
Clients sometimes find that version disappointingly modest. It is modest in the way an accurate description usually is, and every agreement we have seen that promised more has been settling arguments rather than preventing them.
Now the half of this subject that gets no drafting attention at all, despite being where nearly every unhappy engagement actually ends up.
Everything above concerns an argument about quality. Quality arguments are what people brace for, and the clauses they write are aimed squarely at them. In our experience those arguments are rare.
Ask an unhappy client what went wrong and you will seldom be told that the advice was poor. You will be told that the report sits in a drawer, that nothing changed, and that the money went nowhere.
Which is a different failure with a different cause, and no acceptance clause in the world touches it.
The work was done. A view was formed and stated. Somebody read it, several people nodded, and the organisation then carried on as before — not through bad faith, but because acting on it required somebody to do something difficult while also running a business.
The commonest failure in consultancy is not wrong advice. It is unacted-upon advice, and it is an organisational failure rather than a professional one.
Place Order — no advance, pay after work
Look at it from each side in turn, because the symmetry is the whole point.
The consultant did the work he was engaged to do. He examined the question, formed a view, said it plainly, and handed it over. From where he stands the engagement was completed and completed properly, and he is entitled to think so.
He also, usually, has no authority to make anything happen afterwards, no standing in the organisation, and no visibility into why his recommendation stalled. Frequently he does not even learn that it stalled.
And the client, from the other side, has paid a substantial sum and can point to no change in the business.
What happened inside is ordinary. Two managers disagreed and nobody adjudicated. The person who commissioned the work moved on. A quarter got busy. The recommendation required a decision one level above whoever was holding it, and that level never looked at it.
None of that is anybody behaving badly. It is what happens to any piece of work that arrives without a route into the organisation.
Which is why these disagreements are so bitter and so unresolvable: nobody is lying.
The consultant says he delivered what was agreed, and he did. The client says they got nothing out of it, and they did not. Both statements are accurate, both are made sincerely, and they cannot be reconciled because they describe different things.
We have sat on both sides of that conversation often enough to say with some confidence that it is not resolved by argument. It is only prevented, and it is prevented at the drafting stage.
Place Order — no advance, pay after work
Because there is a gap in the middle of almost every consultancy arrangement, and it is sitting in plain view.
The agreement covers the engagement and stops at the report. The organisation’s own processes cover ordinary work and were never told this was coming. Between the moment the recommendation is made and the moment somebody acts on it there is no owner, no mechanism and no document.
Both parties assume that stretch is the other one’s responsibility, which is exactly the condition under which nothing is anybody’s.
So the paragraph we press for in every one of these drafts is the one that describes that stretch, and it is not a legal paragraph at all.
It sets out, in ordinary language, what is supposed to happen between the recommendation and the action. Who receives it. Who reviews it. In what forum, and by when. What happens if that forum does not meet.
Clients sometimes object that this is internal management and has no place in a contract with an outsider. Our answer is that it is precisely the part the outsider is affected by and has no power over, which is what makes it worth agreeing rather than assuming.
Three elements make that paragraph work, and the first is ownership by name rather than by department.
Not operations. Not the management team. A person, identified, who is responsible for each thing that is supposed to follow. A recommendation addressed to a function is addressed to nobody, and the people in that function all reasonably believe somebody else has it.
Naming individuals feels unnecessarily formal when everybody is being agreeable at the start, and it is the single thing most likely to be missing when things go quiet.
Place Order — no advance, pay after work
The second and third elements are a date and a power, and the third is the one that is almost always absent.
A date is obvious enough. The power is subtler: whoever owns an action has to be able to decide it, or to reach somebody who can. Most stalled recommendations are stalled at a level that cannot approve them and has no route upward that anybody is using.
When a recommendation requires a decision above the person holding it, say in the agreement who that decision goes to and within what period. One line of text buys it, and it separates a report that stalls from one that gets an answer — even where the decision is no.
And the reason to put it in the agreement rather than in a well-meant conversation is a matter of timing.
At the point of signing, everybody is optimistic, nobody is defensive, and the question of what will happen afterwards can be discussed as a practical matter. Three months later it is a question about blame, and it will be answered defensively by whoever is asked.
The agreement is the last moment at which this can be settled cheaply. After that it is settled expensively or not at all.
So why is it missing from nearly every consultancy agreement we are asked to look at?
Partly because the templates people start from have no such clause, having been written for supply arrangements where nothing follows delivery. Partly because it reads like management rather than law, and whoever is drafting feels it is not their business.
And partly, we suspect, because raising it requires somebody to say out loud that the client’s organisation might not act. That is an uncomfortable sentence to say to a new client in the first week, which is why the person best placed to say it is usually somebody neither side is paying to be liked.
Place Order — no advance, pay after work
Now the other thing that reliably goes wrong, and it is the one that costs consultants rather than clients.
Every field of professional work has a version of this argument, and most of them have a familiar shape: the brief grows, somebody asks for extra, and there is a conversation about a variation.
The consultancy version does not look like that at all, which is why people who have managed scope successfully in other contexts are repeatedly caught by it here.
In a build or a supply, the scope is a list of things, and growth is visible as items being added to the list.
Somebody asks for an extra module, another floor, a further thousand units. The request has an edge, somebody has to make it, and it is at least possible for the person receiving it to say that it falls outside the agreement.
Consultancy scope is not a list of things. It is a question — and questions do not grow by having items added. They grow by being asked next to other questions.
Which is the actual mechanism, and it is worth describing slowly because it is so undramatic.
The consultant is engaged on one matter and is therefore in the building. Somebody mentions an adjacent difficulty over tea and asks what he makes of it. He makes something of it, because that is what he does and because saying nothing would be rude. The next week somebody forwards him a document from that other area, for his thoughts. A month later he is attending a meeting about it.
Every one of those steps is small, reasonable, and done in good faith by somebody who is pleased with the relationship. Nobody involved thinks of it as scope.
Place Order — no advance, pay after work
And so the normal machinery never engages, because the normal machinery needs an event to engage with.
There is no request to approve, no variation to price, no document to sign. Nothing crossed anybody’s desk marked as a change. The agreement’s careful provisions about scope variations sit unused, because they were drafted to catch a kind of event that is not occurring.
Consultancy scope does not expand by request. It expands by conversation, which no clause written in terms of requests will ever catch.
Six weeks on, the shape of the engagement is materially different from the one on the paper.
The consultant is now spending a third of his time on a matter that was never within the brief, has become the person others go to about it, and in practical terms has acquired a role nobody created. His fee is unchanged. His exposure, his hours and his involvement are not.
And from the client’s side nothing improper has happened at all. They asked an expert who was already present what he thought, which is about as ordinary a thing as happens in an office.
Which is why the clause that works here is not about variations but about duties, and it points in one direction.
The consultant has to be the one obliged to raise it, by name, in the agreement. Something to the effect that where he considers the work has moved beyond the stated question, he will say so in writing before continuing, and the parties will then agree whether it is a new engagement.
It looks like a clause against the consultant and it is the clause that protects him, because it converts an awkward social act into a contractual obligation he can point at.
Place Order — no advance, pay after work
A practical form of the same clause is worth mentioning because consultants find it easier to use. Rather than a bare obligation to raise the matter, some agreements fix a periodic review — a short conversation at a stated interval whose only purpose is to ask whether the work still matches the brief. It removes the social difficulty entirely, because nobody has to decide to start the conversation; the calendar starts it, and the awkwardness disappears into a routine.
And the reason the duty cannot sensibly sit on the client’s side is simple arithmetic rather than any want of integrity.
The widening gives the client more of a valued person’s attention at no additional cost. It feels like a relationship going well. There is no moment at which anybody on that side has a reason to stop and ask whether this should be repriced, and no individual whose job it is to notice.
We say this to client-side readers as plainly as to consultants: you are not going to raise it, you will not notice that you have not raised it, and a good consultant raising it is not being difficult with you.
Now a question that matters more in this contract than in any other we draft, and which most agreements answer by accident.
Think about why the engagement is happening. Somebody met somebody, or was recommended to somebody, or read something, or worked with somebody years ago. The decision to engage was a decision about a particular mind.
Judgement sits in a person. Unlike a capability — a factory, a development team, a service desk — it does not reside in an organisation and cannot be reassigned between staff of equivalent seniority.
But the agreement is almost always with an entity, for perfectly ordinary reasons, and the entity is free to resource the work as it sees fit.
Which means the document records a bargain nobody made. You agreed with a person. You contracted with a company. The company has promised consultancy services and may deliver them through whoever is available.
In practice this surfaces at about month two, when the person you met attends alternate meetings and somebody more junior does the work. Nothing has been breached and you have not received what you decided to buy.
Place Order — no advance, pay after work
And the sharper version arrives when the individual resigns, is moved, falls ill or is pulled onto something more lucrative.
There is no general answer to what should follow, because it is a commercial question with several reasonable answers: a right to approve any replacement, a right to end the engagement, a pause while a substitute is agreed, a reduced fee, or genuinely nothing at all where the firm’s method rather than the individual was the draw.
What matters is that it was decided in week one rather than discovered in month five. After the person has gone, the conversation is conducted by people who are annoyed, and the answer is whatever the stronger party says it is.
Which leads to the slowest-moving risk in this field, and the one nobody notices because it has no event in it.
Consultancy arrangements are frequently extended rather than renewed. The first engagement answered its question, the relationship was good, something else came up, and the arrangement simply continued. Three years later nobody can say when the original brief ended.
Nothing was ever decided and a great deal has changed.
And the markers of that change are physical and administrative rather than contractual, which is why they attract no attention.
A desk. An address on the client’s own domain. A place in the weekly meeting. Access to systems. A line on an organisation chart drawn by somebody in a hurry. Introductions to outsiders that do not mention the word consultant.
Each arrived for a sensible practical reason, usually convenience. Together they describe an arrangement that no longer resembles the one on the paper, and the paper is the only thing that has not moved.
Place Order — no advance, pay after work
At which point a different question arises, and we are careful to say that it is not a documentation question.
How such a relationship is properly characterised — for tax, for employment, for anything else that turns on it — depends on how the arrangement actually works rather than on what the agreement is called, and it is assessed on facts we are not qualified to weigh. The employment side of that distinction is set out properly elsewhere and we will not restate it here.
We do not give a view on whether a long consultancy arrangement should be characterised differently, and we will not draft around the question to make an answer more likely. Where the arrangement has drifted far enough that the question is live, it goes to an advocate and to your accountant, who will look at how it operates rather than at the heading on the document.
Now the dimension of this subject that is thought about least clearly, partly because everybody believes it has already been handled.
Ask either side about confidentiality and you will be told there is a clause, it was signed, and the matter is dealt with. The clause exists, it is usually competent, and it does real work. It also addresses the smaller half of what is actually at stake.
What a consultant takes with him at the end of an engagement is not primarily in files.
Because the whole purpose of the engagement was to make somebody understand your situation, and understanding is what he leaves with.
How your margins actually behave. Which of your managers can be relied on and which cannot. Where your process genuinely breaks as against where you believe it does. What your customers say when you are not in the room. Which of your stated priorities are real.
Documents can be returned, deleted and audited. Understanding cannot be returned, deleted or audited, and no clause ever drafted has managed it.
Place Order — no advance, pay after work
And here is the fact that follows from it, which nobody says out loud during the pitch.
You engaged a specialist precisely because he knows your industry. The reason he knows it is that he has spent years working for businesses like yours. The reason he will still know it next year is that he will continue to. His next client is quite likely to be somebody you compete with, possibly within the same quarter.
That is not duplicity and it is not a failure of loyalty. It is the arithmetic of specialisation, and a consultant who only ever worked for one business in a sector would not have been worth engaging in the first place.
Which makes it a question that ought to be discussed openly at the start, and almost never is, for reasons that are human rather than commercial.
The client feels that raising it sounds distrustful before the work has begun. The consultant feels that raising it invites a restriction he would rather not have. So both sides sign a confidentiality clause that addresses documents, feel that something has been dealt with, and move on.
What any restriction on future clients can say, how wide it may be, and whether it will hold are questions of law rather than of drafting taste. We raise the subject, we record what the parties decide, and the detailed answer belongs with an advocate — as does the confidentiality document itself where the position is anything other than straightforward.
What we will say commercially, to both sides, is that a wide restriction destroys a specialist’s practice, so a consultant who signs one without reading it carefully has usually not understood what he has agreed, and a client who obtains one has sometimes bought a person who can no longer afford to be useful.
Now a practical matter about what should actually be written down, and it begins with an observation about memory.
An engagement is mostly conversation. Meetings, site visits, questions asked in corridors, the explanation given to a manager who was sceptical, the reasoning behind a recommendation that was rejected and a second one that was not.
Within eighteen months, none of that exists for anybody. People leave, memories soften, and what survives is a document.
Place Order — no advance, pay after work
Which has consequences for both sides, and they pull in the same direction.
If the recommendation is later questioned, the reasoning behind it exists only where it was written down. If the client later says a risk was never flagged, the flag exists only if somebody recorded it. If the consultant later says the brief changed halfway, the brief as it stood at each point exists only on paper.
Three years on, the file is not a record of the engagement. The file is the engagement, as far as anybody is able to establish.
Which argues for recording rather more than busy people naturally do, and in particular for recording the things that feel too obvious to write at the time.
None of that is defensive paperwork. It is the material that makes the work usable by somebody who was not present, which within a couple of years is everybody.
There is a converse, and it is a genuine judgement rather than a rule, which is why we raise it rather than decide it.
Some things a consultant observes are about individuals, and committing them to a document that will circulate changes what they do. A written assessment of a named manager’s capability has a long life and no context, and it will be read by people it was not written for.
The conventional handling is to put the structural finding in writing and keep the personal one in conversation. Whether that is right in a given engagement is for the parties, and it is better decided deliberately at the start than improvised by a consultant on a deadline at the end.
Place Order — no advance, pay after work
And the useful framing is that this material is not held by one side against the other.
A consultant with a full file can show what he was asked, what he said and what he warned about. A client with the same file can show what was promised, what was delivered and what was agreed to follow. The same pages serve both, which is why the file should be built jointly as the engagement runs rather than reconstructed afterwards by whoever is worried.
Keeping the engagement’s papers together in one organised place is unglamorous and it is the difference between a dispute that resolves in a fortnight and one that does not resolve at all.
A word about where most of the agreements we are shown actually came from, because it explains nearly every defect in them.
Somebody had a document. It was for a vendor supply, or a development project, or a service arrangement with uptime commitments, and it worked well enough for that. The nouns were changed and the structure was kept.
So the draft arrives with delivery and acceptance machinery, sometimes with service levels borrowed from a performance arrangement, occasionally with provisions about goods. Every one of those was answering a question this engagement does not ask, and their presence crowds out the three or four clauses that would have been worth having.
One further inheritance from those templates is worth watching for because it is harmless-looking. Borrowed documents frequently carry liability provisions calibrated to a supply of goods, expressed as a multiple of fees or a fixed sum, and nobody revisits the number. What amount is appropriate where the subject is advice, and how that interacts with everything else in the draft, is something an advocate settles and a drafting service does not, so we raise it instead of quietly carrying the old number over.
Gathered together, the beliefs that arrive unexamined and do the damage.
| The assumption | What is actually so |
|---|---|
| The report is the deliverable | The report is a container; the judgement inside it cannot be inspected |
| Tying the fee to results creates accountability | It buys a consultant who recommends what will be accepted |
| Scope is protected by a variations clause | Scope here grows by conversation, which no variations clause catches |
| Confidentiality covers the exposure | It covers documents; understanding leaves with the person and cannot be recalled |
And if you are reading an existing draft rather than commissioning a new one, these are the places to look first.
The question is described in a sentence so broad that both sides could argue anything falls inside it. There is no named individual. There is no end date, or an end date that has already passed and been ignored. There is nothing at all about what follows the recommendation. And there is an acceptance clause borrowed from a supply contract that nobody intends to use.
Four of those five are absences rather than errors, which is why reading a draft for what it says rarely finds them.
Where a client sends us something already in existence, our reading runs in a particular order and it is not front to back.
First, what is being bought and whether the document knows. Second, who the person is. Third, whether anything is said about after. Fourth, what triggers a new engagement and who must say so. Fifth, how it ends and what survives the ending. Only then the ordinary clauses, which in our experience are usually competent, because they were inherited from a document that had been used before.
It is a short list and it finds most of what matters. The defects in these agreements are remarkably consistent, which is the main reason we were able to write a page like this one.
Three situations, and placing yourself among them redirects where the care should go.
| The situation | Where the care is needed |
|---|---|
| A business engaging an outside expert for the first time | Defining the question narrowly, and what follows the recommendation |
| A consultant who has been burnt once | Scope by conversation, the named individual, and how it ends |
| A long arrangement begun on an email, now being papered | What the arrangement has actually become, before anybody writes it down |
The third is the hardest and the one we most often see, usually because somebody senior has finally asked to look at the paperwork and discovered there is none.
Place Order — no advance, pay after work
Before drafting a line, and the half hour spent on them decides the quality of what follows more than any amount of careful wording afterwards.
A client who answers those five properly gets a materially better document than one who sends a template and asks us to improve it, and the difference is not in our drafting. It is that the five answers are the agreement, and everything else is wording.
One refusal, stated plainly so nobody is waiting for it.
We will not draft a clause making payment conditional on the outcome of the advice, and if a client asks for one we will explain why before anything else is discussed. The grounds are laid out at length above and none of them are legal ones — the objection is that the clause alters what a consultant is prepared to say, and what he is willing to say is the entire thing being purchased.
If a client still wants it after that conversation, then we are not the right people to prepare this particular document, and we would rather say so than draft something we have just finished arguing against.
One person takes the instructions, asks the five questions, writes the draft and handles the amendments.
On a document this short that might seem a small point, and it is not. Most of the value here is in what was discussed before the drafting, and that does not survive being handed on. A second person working from the draft alone has the wording and not the reasons, and will amend away the clause that mattered because it looks unusual.
Where the engagement needs a company’s internal approval recorded alongside it, a resolution is a separate document and we will say so rather than folding it in.
Place Order — no advance, pay after work
Set down together here, since a boundary mentioned halfway through a long guide has stopped being remembered by the time anybody acts on it.
We do not promise an outcome in any drafting we prepare. We do not advise on what a restriction affecting a consultant’s future clients may say or whether it will hold. We give no view on how a relationship should be characterised for tax or employment. We do not value anybody’s advice, assess whether it was competent, or opine on whether a fee was earned. And we do not act once there is a dispute.
If an engagement has already gone wrong — a fee withheld, a recommendation blamed, a restriction said to have been breached — the matter has left documentation territory, and our response is to name that and withdraw rather than draft our way into it. Court work is for your advocate, whose fee is engaged and paid by you directly; we do not quote, collect or share it.
Plainly, in the manner this site quotes everything.
Our fee for a Consultancy Agreement is ₹900, with the usual turnaround running to 1 – 2 days, the clock starting once your answers to the five questions reach us and not when the first email does.
Nothing else is folded into it. Stamping, registration or notarisation, where any of those apply to your arrangement, are charged by whoever performs them and are not collected by us. A second round of amendments after the other side has commented is included; a redraft following a change in what the engagement is for is a fresh piece of work and we will say so before starting it.
And if you send us something that is already adequate, we will tell you it is adequate. We have returned drafts unchanged more than once with a note about the two things worth adding, and we would rather do that than earn a fee by rewriting competent wording into our own house style.
One thing to put in a diary, because it is the step that prevents the slowest problem on this page.
Read the agreement again at the end of its first term, whether or not the arrangement is continuing. Not for the wording — for whether the description still matches what is happening. Is the question still the question? Is it still that person? Has the work moved into areas the paper does not mention? Has an end date passed that nobody noticed?
Ten minutes, once a year, and it catches both the scope drift and the long-engagement drift before either becomes anybody’s problem. Where the arrangement has genuinely become something else — a shared venture, a revenue share, an ongoing supply relationship — it wants its own document rather than an extension of this one, whether that is a joint venture, a revenue arrangement or a supply agreement. And where a consultant is being engaged for something closer to delivery than to advice, the freelance side of this subject covers ground — ownership of what is produced, and the position on delayed payment — that this page deliberately leaves alone.
We start with five questions, because the answers are the agreement and everything after them is wording: what question is being answered, whose judgement is being bought, who receives and decides on the recommendation and by when, what counts as the work having become something new and who must say so, and what happens at the end. Then we draft around those — effort, access, time, the named individual, the paragraph about what follows the recommendation that almost no agreement contains, and a confidentiality position that both sides have actually discussed rather than assumed. We will not write a clause making payment turn on results, and we will tell you why. We give no view on restrictions affecting future clients, none on how the relationship should be characterised, and we do not act once there is a dispute.
The omissions here are chosen — this is the account of them
You will not find on this page: any section or statutory provision; any decided case; any specimen clause or suggested wording; any stamp duty figure or registration requirement; any fee other than ours; any statement of what a restriction on a consultant’s future clients may contain or whether it is enforceable; any view on how a consultancy relationship ought to be characterised for tax or employment; and any position on who owns what is produced during the engagement. For a guide about a contract, that is a conspicuous amount of contract law to leave out.
It rests on three grounds, and the opening one belongs to this subject alone. No specimen wording appears because a consultancy agreement is the one commercial document where a borrowed clause is actively dangerous — this entire guide rests on the claim that such agreements fail precisely because they are lifted from contracts written for other things, and a website supplying clauses would be adding to the pile. Restrictions on future clients, characterisation of the relationship, and ownership of output are all out because each is a question of law decided on facts, each is already properly owned by a profession that is not ours, and a paraphrase on a website is the form in which such answers do the most harm. Charges and stamping are absent for the familiar reason: such numbers shift, and one parked on a web page keeps directing choices long past the point of being accurate.
Strip every one of those away and what remains does not depend on any provision anywhere. That a consultancy engagement buys judgement rather than an object, so there is nothing to inspect and no acceptance to give. That a contract promising a result therefore misleads whichever side reads it hopefully, and changes what the consultant is willing to say. That effort, access and time are the three things actually within his control. That the dispute which arrives is almost never about bad advice but about nothing having happened, and that one plain paragraph about who does what afterwards is what prevents it. That scope here grows by questions rather than requests, so the duty to notice must sit with the consultant. That a person was hired while a firm was contracted with. And that understanding leaves with the individual and cannot be recalled by any clause.
If a single practice from this page outlasted it, let it be the five questions before drafting. What question is being answered; whose judgement; who receives and decides on the recommendation and by when; what counts as this having become something new and who must say so; and what happens at the end. Those five take half an hour over a telephone and they are the agreement. Everything a lawyer does afterwards is the expression of five answers — and an agreement built on five vague ones cannot be rescued by good drafting.
Where to take the rest of it. To an advocate for anything restricting whom a consultant may work for next, and for who owns what the engagement produces, before either side signs rather than afterwards. To an advocate and an accountant together if a consultancy arrangement has run for years and acquired a desk, an internal email address and a place in the hierarchy. To the client’s own management, not to any lawyer, for the question of who will actually act on the recommendation — that one cannot be outsourced. And to the consultant himself, at the start, for the awkward conversation about his next client, which is more easily had before the engagement than after it.
Two doors, both free. Clients search a factual directory of enrolled advocates. Advocates apply to be listed on it — no fee, no commission, nothing paid in either direction.
Search Bar Council enrolled advocates by what your matter is about, by court, or by city. Searching and sending a request are both free.
Enrolled advocates anywhere in India can apply to be listed. Your entry is published only after we verify your enrolment number with your State Bar Council.
Would rather not pick anybody? Tell us the matter instead — one form goes to every advocate at once, and the first to take it up calls you.
This directory carries no ratings, no reviews, no rankings and no fees — only the factual particulars the Bar Council of India permits, published at each advocate's own request. Browse the network · Terms for Advocates