You finished the audit. You know which of your skills survive the move, and you have found the one that sits at the intersection of your craft and your domain, the thing you are unusually good at that also happens to be worth money to a specific kind of buyer.
Then you open a laptop to set the thing up, and you stall. Not on the logo or the website. On a question that sounds like semantics and isn’t: are you going freelance, or are you going into consulting?
Most advice treats these as the same word wearing different clothes. Freelancing sounds younger and scrappier; consulting sounds older and more expensive, so you pick the one that flatters you and move on. That choice is quietly load-bearing. Two people can leave the same company with the same twenty-five years behind them, quote the same hourly number, and end up in businesses that share almost nothing — different clients, different sales conversations, different ceilings, and very different Tuesdays.
Here is the version nobody says out loud. A large share of experienced professionals who intend to consult end up freelancing without noticing, at consulting prices for about six months, until the market corrects them. They do not fail because they lacked expertise. They fail because they sold the wrong thing with the right résumé.
This post is about telling the two apart before the market does it for you.
What is the difference between consulting and freelancing?
A freelancer is hired to execute a defined piece of work; a consultant is hired to decide what the work should be. The freelancer’s client arrives with the problem already named and the solution already chosen, and buys capacity. The consultant’s client arrives with a symptom, and buys judgment about what it means.
Everything else — rate, title, contract length, whether you call it a “practice” or a “shop” — follows from that split rather than defining it.
It helps to notice that the word “consultant” has been stretched until it stopped meaning anything. Job boards use it for staffing placements. Agencies use it for contract designers. Plenty of people who describe themselves as consultants are, by any working definition, freelancing on longer contracts. None of that is dishonest; the language is just loose. But loose language costs you money when you are pricing yourself, because the two models have different economics and you cannot charge for one while delivering the other for long. Naming the arrangement is upstream of naming the number, and what to charge for your expertise picks it up from there.
The useful distinction is not about status. It is about where the decision gets made.
The request that used to arrive on a Thursday afternoon was always some version of “can you put together a quick deck?” It was never quick, and it was never really about the deck. Somebody had already decided what the answer was and needed it in a form that could be circulated. A great deal of freelance work arrives in exactly that shape: the problem comes pre-named, and the invoice is for the formatting.
Why experienced people drift into freelancing by accident
If you spent decades inside organizations, you were trained — thoroughly, and by people who meant well — to be handed problems.
Someone above you decided what mattered this quarter. Someone in another function decided the constraint. Your job, and the thing you got promoted for, was executing beautifully inside a scope that arrived pre-shaped. Twenty-five years of that builds a genuine and valuable reflex: you become the person who can be trusted with a defined problem. It also builds a habit that does not serve you on your first sales call, which is waiting to be told what the problem is.
So the call happens, and a prospective client describes something messy, and you do the thing that has always worked. You listen for the scope. You find the piece you can definitely deliver, you name it, and you quote it. The client is relieved, because you just made their vague worry into a line item. You have also just converted a consulting engagement into a freelance one, and you did it in about ninety seconds, out of professional courtesy.
There is a second force underneath this, and it is the one Michael Polanyi pointed at when he wrote that we know more than we can tell. The judgment that makes you worth hiring is largely tacit. You can list your deliverables easily — a model, a plan, a migration, a campaign — because deliverables are visible. The thing that actually took twenty-five years to build is the pattern recognition that tells you which deliverable is the right one, and that is nearly invisible to you, precisely because it is so automatic. This is the expert blind spot working against you commercially: the more fluent your judgment, the less you notice you are using it, and the less likely you are to charge for it.
Freelancing is what happens when you sell the visible part and give away the invisible part for free.
I sat in a great many meetings that were described as investment meetings. A fair number of them were not. They were meetings about who would be named if the thing went wrong, and everyone in the room could feel the difference even though nobody said it out loud.
That is the line the next section draws. Not what the work is called. Who is holding it when it fails.
The Accountability Line
Every engagement has a line running through it. On one side, the client owns the decision and you own the execution. On the other, you own the recommendation and the client owns the outcome. That line is the whole distinction, and unlike “seniority” or “rate,” you can locate it precisely.
Four tests. Run any engagement — including one you already have — through all four, because the answers do not always agree, and the disagreements are where the money leaks.
1. Who names the problem?
Ask who decided what this project is. If the client arrived with “we need our data warehouse migrated,” they named it. If they arrived with “our reporting is slow and I don’t know whether it’s the warehouse, the queries, or the org chart,” and you determined which, you named it.
Naming the problem is the single most valuable thing you do, and it is almost never on the invoice.
2. What is actually on the invoice?
Look at what you are being paid for, not how it is billed. Hours and deliverables sit on the freelance side. Decisions, recommendations, and outcomes sit on the consulting side. An hourly rate does not make you a freelancer — plenty of consultants bill hourly — but an invoice where every line is a thing you produced rather than a call you made is telling you something.
The test question: if you had delivered the same recommendation in a two-page memo after one week instead of a full build after three months, would the client have paid you? If yes, they are buying your judgment. If no, they are buying your hands.
3. Who carries the risk if it turns out wrong?
This one is uncomfortable, which is why it is diagnostic. When the approach fails, whose call was it?
Freelancers are protected here in a way that is easy to undervalue. You built what was asked, to spec, and if the spec was wrong that is a difficult conversation but not your liability. Consultants have no such shelter. You said do this, they did it, and if it did not work the failure has your fingerprints on it. That exposure is exactly what the premium is for. If you are not carrying any of it, you are unlikely to be paid as though you are.
4. What survives after you leave?
When the engagement ends and you walk out, what remains?
If the answer is a body of finished work, the value left with the work. If the answer is that the organization now thinks about the problem differently, makes a different decision next quarter, or has stopped doing something expensive that everyone had assumed was necessary, then the value left in the client’s head, and it keeps compounding after you stop billing.
Read the four together. Most people over 40 find they are split: they name the problem and carry real risk, but the invoice is all deliverables and nothing survives their exit. That is the specific, common, expensive position — doing consulting work under a freelance contract. It is also the easiest one to fix, because you are already doing the hard part.
Not sure which of your skills belongs on which side of the line? The Blueprint has a worksheet for exactly this.
→ Download the free Midlife Reinvention Blueprint
The Debrief: score your last three engagements
The four questions are a diagnosis, and a diagnosis you only ever run in your head is one you will keep getting wrong in your own favor. So run it on paper, against work you have already finished, where the answers are matters of record rather than intention.
Take your last three engagements. If you are still employed, use the last three pieces of work substantial enough to have had a proper beginning and a proper end — internal work counts, and for most people it is the more honest sample, because nobody was performing for a prospect. For each engagement, answer the four questions and give yourself one mark for every consulting answer.
Mark one — who named the problem? A mark if you did. If the client or your director arrived with the problem already worded, no mark, however hard the work turned out to be.
Mark two — what was actually being bought? A mark if the thing of value was a recommendation. If you were paid for a build, a deck, a migration, or a set of hours, no mark, even if the recommendation was in there somewhere for free.
Mark three — who carried the risk? A mark if a wrong call would have been yours. If you could have said you built exactly what was specified, no mark. That sentence is the freelancer’s shelter, and it is worth real money, which is why giving it up is worth more.
Mark four — what survived after you left? A mark if what remained was a changed decision. If what remained was a finished artifact, no mark.
You now have three numbers between zero and four, and the temptation is to average them. Do not. The average is the least informative thing available here, and averaging is precisely the move people have been using to avoid this answer for years.
Read the spread instead.
- All three at zero or one. You are a freelancer. This is not a demotion and there is nothing to fix if the model suits you, but it does mean your rate is capacity-bound and your website should stop implying otherwise.
- All three at three or four. You have been consulting for a while without charging for it. The work is already judgment work; the invoice has simply not caught up. This is the most common result among people who have spent twenty years senior somewhere.
- Scattered — a four, then a one, then a three. This is the drift, and now you can see it. You are selling judgment when the conversation goes well and falling back to execution when it stalls, which teaches every client to buy the cheaper version.
The scatter is the finding. A consistent one is a business. A consistent four is an underpriced business. A four, a one and a three is a person who has not decided, and the market has been deciding on their behalf.
Two things worth keeping straight before you run it. The Empty Room Test in productizing your knowledge also scores, and it scores something else entirely: whether the thing you built works while you are absent. The Debrief asks whether you were hired for your judgment at all. You can score four here and zero there, which describes a consultant who has built nothing transferable, and the reverse turns up just as often.
And run it in an hour when you can still be honest, which is not the same as an hour when you happen to be free. The hours that can hold a real decision are a much smaller set than your calendar implies — that arithmetic is its own subject, and a debrief conducted at eleven at night reliably returns the flattering answer.
What the Debrief looks like when it is honest
Neil is not a real person; he is stitched together from a conversation that keeps recurring among people who left a senior operations role and hung out a shingle. He had been calling himself an independent operations consultant for fourteen months, and by his own account it was going adequately and he could not work out why it did not feel like it was going anywhere.
His last three engagements, scored.
The first: two marks. A manufacturer asked him to redesign their inbound goods process. They had named the problem, so no mark there, and the invoice was for a documented process and two weeks of training, so no mark there either. But when it went wrong the fault would have been his design rather than their brief, and what survived him was a changed decision about where to hold inventory. Two.
The second: zero. Six weeks of interim cover for a maternity leave, at a good day rate. He had enjoyed it. Nothing about it was consulting, and he had privately been counting it as his best month.
The third: four. A logistics firm had asked him to quote for a warehouse management system selection. He told them, in the first meeting and without charging for it, that their problem was not the system but that two directors were measuring the same operation with different definitions of on-time. They paid him to sort that out instead. He named it, they bought the recommendation, the call was his, and what survived was a single agreed definition that changed how the whole business argued.
Four, zero, two. The scatter is not subtle once it is written down, and Neil’s reaction to seeing it was the reaction most people have: mild irritation, because he already knew.
What he did next is the useful part, and it was not a repositioning exercise. He noticed that the four had begun with an unpaid opinion offered early, and the zero had begun with him being asked whether he was available. Availability is what you sell when nobody has heard you say anything. So he changed one habit: in every first conversation, before any discussion of scope, he said what he thought the actual problem was, out loud, at the risk of being wrong. That is the accountability line itself, reduced to something he could do on a Tuesday.
His next three scored three, four and two. The two still bothers him. He also stopped taking interim cover, which cost him a comfortable month and bought him the year that followed.
Three ways to actually start
There is no universally correct side of the line. There is a correct side for your skill, your appetite for exposure, and how much runway you have. What follows are three real routes, not a ladder — plenty of people stop happily at the first.
1. Freelance deliberately, not by default
If your Signature skill is genuinely execution — you are the person who can build the thing, and the market is short of people who can build the thing — freelancing is not the lesser option. It is faster to sell, it needs no reputation-building, and the sales conversation is refreshingly concrete. You can be earning inside a month.
The trap is drifting into it while telling yourself and everyone else that you are consulting, which produces a rate you cannot defend and a story that does not match the invoice. Choose it on purpose, price it like the capacity business it is, and keep your utilization honest. Your income is hours times rate, so protect both.
2. Consult from the first conversation
If your value is that you have seen this specific situation forty times and know which of the plausible explanations is usually the real one, then start on the consulting side and do not apologize for it.
Practically, this means resisting the reflex from earlier. When a prospect describes a mess, do not immediately shape it into a quotable scope. Ask what they have already tried, what they think is going on, and what happens if nothing changes. Then say what you actually think — including when your honest read is that their named problem is not their real problem. The moment you tell a client their diagnosis is wrong and you turn out to be right, you are no longer competing on rate with anyone.
Sales cycles here are longer and start colder. Budget for that.
3. The hybrid ramp
The most common real-world path, and the one most likely to fit if you have left a long career recently: take defined execution work to cover your costs, and use every engagement to practice naming problems.
The mechanic is simple. Inside a freelance project, you will constantly notice things adjacent to your scope that are wrong. Normally you keep quiet, because it is not what you were hired for. Instead, say them — deliberately, once per engagement, without invoicing for it. “You asked me to fix the reporting. The reporting will be fine. The reason it broke is that two teams own the same definition of a customer, and it will break again in a year.”
Some clients will nod and ignore it. One will ask you to come back and deal with that instead. That client is your first consulting engagement, and they arrived because you demonstrated judgment rather than claiming it on a website. That is the whole game of a personal brand after 40 — showing judgment in public until the right client finds you.
How to test this without quitting anything
- Take the Signature skill you identified in the transferable skills audit and write two one-paragraph offers for it — one framed as execution, one framed as judgment. The second will be much harder to write. That difficulty is the point, and it is the actual work.
- Run your last three pieces of work through the four tests. Even the ones you did as an employee. Most people discover they have been consulting internally for years without the title.
- Have five conversations, sell nothing. Ask former colleagues and their networks what problem in their world is expensive and nobody owns. You are listening for the same complaint from three different people.
- Price both models on paper before you need to. A capacity business and a judgment business have different math, and doing that math under time pressure on a live call is how people end up quoting a number they resent for the next six months.
- Give one recommendation away, on purpose. Then watch whether anything happens. This is the cheapest possible market test of whether your judgment is legible to buyers, and it costs you a coffee.
Common mistakes
- Charging consulting rates for freelance work. The market corrects this quickly and unkindly. If the invoice is all deliverables, the rate has a ceiling no résumé will lift.
- Waiting to feel qualified enough to advise. The expert blind spot guarantees you will underrate your own judgment. Qualification is not a feeling that arrives; it is a client acting on something you said and it working.
- Selling the twenty-five years instead of the specific problem. “Three decades in financial services” is not an offer. “I can tell you within two weeks whether your ops problem is a systems problem or a headcount problem” is.
- Treating the first client’s request as the shape of the business. Your first client bought whatever you happened to describe first. That is not market research.
- Building the website before the fifth conversation. The website is a place to put language you do not have yet. Have the conversations, steal the words your prospects actually use, then build.
- Assuming the hybrid is temporary. For a lot of people it is the destination, not the ramp — steady execution work funding a smaller number of higher-leverage engagements. That is a good business, not a failure to graduate.
Where that first client actually comes from is a separate question, and it has its own answer in how to get your first client after 40 — which starts not with strangers but with the people who already know what you can do.
Frequently asked questions
Is consulting or freelancing better after 40?
Neither is better, but they reward different assets, and after 40 your accumulated judgment is usually the rarer one. If your value is in knowing which approach is right rather than in executing a known approach faster than the alternatives, consulting will pay you more for the same effort. If you genuinely love the craft and want to keep doing it, freelancing is not a consolation prize.
Can I do both at the same time?
Yes, and most people do at the start. The thing that causes trouble is not mixing the two. It is mixing them inside a single engagement without pricing for it, which is how you end up giving away the valuable half. Keep them as separate offers with separate numbers, even if the same client eventually buys both.
How much should I charge when I’m starting out?
The honest answer is that your first number is a hypothesis, not a valuation. Work out what your annual target requires given a realistic number of billable days, because new solo practitioners consistently overestimate how many of those exist. If you are pricing a judgment offer, the anchor is what the decision is worth to the client, not what your hour costs, and if you want to work back from an income target rather than guess, the Freedom Number is the piece that makes that math concrete.
Do I need to specialize in one industry?
Specializing makes the sale much easier, because “I fix this specific problem for this specific kind of company” is a sentence a prospect can act on and refer. That said, the specialization that works is usually narrower than an industry — it is a recurring situation, and it may show up across three industries at once. Let the pattern in your last five conversations tell you what it is instead of choosing it in advance.
Isn’t AI going to eat consulting work anyway?
AI is genuinely good at the execution layer, which is precisely why the accountability line matters more now, not less. What a model cannot do is sit in a room, read what is not being said, and take responsibility for a call in a political and human context. The digital leverage side of this is worth understanding, because the same tools that compress freelance work also let one experienced person deliver what used to take a small team. Either way you are still paid for hours, which is why the freedom asset ladder treats consulting as a rung to build from rather than a place to land.
What if I’ve only ever been an employee and never sold anything?
Then you have almost certainly sold constantly and called it something else — getting a budget approved, talking a skeptical team into an approach, persuading a peer to change a roadmap. The uncomfortable new part is asking for money, not persuasion. It gets easier faster than you expect, mostly because the first client says yes and the story you were telling yourself stops being convincing.
What if my last three engagements score zero and I do not want to be a freelancer?
Then you have a clean starting point rather than a problem, because the fastest route to a consulting mark is not a rebrand — it is offering one unpaid opinion, early, in a conversation you are already having. Score your next three rather than agonizing over the last three, and if a mark has not appeared by then, the constraint is which conversations you are getting into rather than how you describe yourself.
Get The Midlife Reinvention Blueprint
The free 20-page workbook for redesigning your career, money, and purpose after 40 — the four-pillar framework, a personal audit, the five defining questions, and a 90-day roadmap, including the worksheets for your freedom and coast numbers.
The question underneath the question
The reason this choice feels heavier than a business-model decision is that it is one. Freelancing asks what you can do. Consulting asks what you think — and after a couple of decades of being paid to execute other people’s judgment well, saying what you think, in public, with a price attached, is a genuinely different act.
Most people find the answer is not permanent. You start where you can sell, you build the reputation that lets you sell the other thing, and the mix shifts over a few years without any single dramatic moment. What matters is that you know which one you are doing this week, and that the invoice agrees with you. If you are still working out which of your skills should be carrying this, the career reinvention map is the wider view.
Once the model is chosen, the next question is what it has to earn — not to make you rich, but to buy back the specific hours you left a career to reclaim. That is a number, and it is knowable. Start with your Freedom Number, and if you are still working out which direction any of this should point, the whole map is here.
Most of the frameworks here come with a scored instrument: a worksheet you fill in rather than read. Reading one is quick. Filling one in is the part that changes something.
Open the pack →
